One hundred fourteen farmers participated in the second phase of the study. Almost all were married and most were male. Of the females who participated, 60 percent were widowed or divorced and acted as heads of the household. The rest represented their husbands or were acting farm managers for their husbands who had salaried jobs and were mostly absent during the week. The average percentage of female headed households was 19 percent in the comparison group and nine percent in the tree expert group. The actual percentage of female headed households in the region is more accurately reflected in percentages in the comparison group, since the farmers in this sample were chosen randomly, whereas the tree expert group was selected by the ranking game.
Age and years of farming experience: The average age of the tree experts (57 years) was higher than the comparison farmers (45 years). The maximum age of both groups did not differ much, however the minimum age of comparison farmers was 22, but 34 in the expert group. Tree experts had farmed their present land for an average of 33 years, comparison farmers for an average of 22 years. This indicates expertise can be attributed, in part, to greater experience. Almost all farmers have always lived on the land they presently cultivate. The average consultant established his/her farm unit between the age of 22 to 25.
Number of children: The average tree expert is both older than the comparison farmer and has more children. The average tree expert household has 7.3 children, of whom 5.2 are living at home. The average comparison farmer household has 5.1 children, with 4.2 living at home.,
There is a significant difference in the number of boys who have moved outside their parents' district. There is little future for many young men as land becomes scarcer due to equal division of land among sons, so many have left the area to seek land or employment (1.2 sons of comparison farmers and 2.5 sons of tree experts). In spite of the generally large farm sizes, migration is particularly high in Kibingo, where sons have left 20 percent of comparison farmer households and 45.5 percent of tree expert households.
Division of land takes place when sons get married, not when the male head of the family dies. From the average age of farmers and the mean age at which sons establish their own farms, it is apparent that farm sizes will start to decrease when the male household head is between 45 and 50. The total amount of farm land owned may decrease even though there still may be a need to cultivate all of the land to feed the remaining family members8. The dynamic of farm sizes engendered by this inheritance system may have important implications for agroforestry development.
About 70 percent of the male heads of household in both groups have an average of four years of primary education, slightly higher than average for the area. A number of farmers, especially in Maraba and Simbi, reported completing about three years at vocational schools.
Education of women and girls in Rwanda is generally much lower than for men and boys. Only two of 14 female heads of household among the comparison farmers and one of four in the tree expert group had received primary education.
Fifty-one percent of the wives of comparison farmers and 31 percent of the wives of tree experts attended primary school, much higher than for female heads of household.
Although salaried employment and non-paid community functions offer stature in the community and access to additional resources that help overcome dependency on farming, there is a near- absence of employment in the rural areas to absorb the surplus labour available. Five farmers (all men) said they also had regular salaried jobs at the time of the research. Five comparison farmers and ten tree experts indicated they had held jobs in the past. The number with non-paid community functions was higher (five comparison farmers, eight tree experts), although the number of community function holders among tree experts has decreased as older farmers have retired from active involvement in the community.
Families also acquire additional resources through the salaried jobs of their children, although children with jobs do not always help their parents. Nearly one-third of tree experts reported that they had at least one child with a salaried job but only 5.7 percent of the comparison farmers. Although the farmers in the comparison group were generally poorer, younger and had smaller holdings, three out of four said their employed children helped. In the expert group, only half the farmers said their employed children provided help. Often, these children were married and had their own families, plus a large number had moved outside their parents' districts, making them feel less obliged to help.
The question of whether farmers work for other farmers resulted in a mirror image of the question about hiring labour, with 37.7 percent of comparison farmers but only 9.1 percent of tree experts stating that they worked as labourers for other farmers.
Almost all of the farmers (98 percent) said they had sufficient labour to carry out all agricultural tasks. Whether or not this was actually the case, or whether farmers did not understand the question is not clear, because 34 percent of the tree experts sometimes hire farm labourers. A smaller number (8.5 percent) in the comparison group also hire labour.
The results of the ranking game indicate that farm size is an important determinant for separating farmers according to their level of knowledge about trees and tree cultivation.
Farm size: There is a significant difference in the farm sizes of comparison farmers (0.18 ha average) and tree experts (1.27 ha average). In both groups, farm sizes were largest in Kibingo (0.24 ha and 2.17 ha in the comparison farmer and tree expert groups respectively). In Maraba and Simbi, the maximum did not exceed 1.0 ha, even among tree experts. The smallest farms encountered were around 200 square meters, which is barely enough to build a house and have a small kitchen garden, let alone cultivate enough food to feed a family.
The larger farm sizes in Kibingo need to be qualified. Many Karama farmers obtained land grants at the time of Independence, which were usually larger in size than the average farm at that time. The largest farms were mainly in the northern part of Kibingo, where much of the land is not suitable for cultivation. Many of these farms had not been subdivided among sons for their inheritance at the time of this study and retained their original size. Thus, many of the Kibingo farmers with two or more hectares only cultivate a small portion and leave the remainder in pasture or in woodlots. In contrast, comparison farmers and the majority of tree experts in Maraba and Simbi obtained their land through inheritance and much of it has been subdivided since they started farming it.
Access to valley bottom land, which can be cultivated year-round, is especially important for small farmers, for whom a harvest of sweet potatoes is a welcome addition during the dry season. Less than two-thirds of the comparison farmers cultivate valley land compared to 91 percent of tree experts. Almost all indicated they owned the val ley land they cultivated. Only one comparison farmer and five tree experts lease the valley land. Comparison farmers own and cultivate smaller pieces than tree experts.
Note: 1 are = 0.01 hectare
SOURCE: den Biggelaar 1994
Land tenure: Almost all comparison farmers, 95.7 percent, obtained land through inheritance compared to 61.4 percent in the tree expert group. However, 18 tree experts (43.2 percent) received their land as a gift from the authorities compared to five in the comparison group. Two farmers reported having purchased their land.
The low number of land purchases is due to the fact that all land belongs to the State. Although families/clans are granted permanent usufruct rights which resemble ownership, it is not legal and people do not have official title to the land.
Land leasing: In contrast to the near absence of a market to buy and sell land, there
is an active market in leasing land. Leasing allows farmers access to land in different micro-ecological niches not found on their own farms. Few comparison farmers (10 percent) lease land to other farmers, as most have small farms. Among the tree experts, 27.3 percent rent land to other farmers. In both groups, about one-third reported renting land from other farmers.
SOURCE: den Biggelaar 1994
Fragmentation of holdings: Each farm is divided into fields at various distances from the home compound and not necessarily contiguous. A field is defined as a unit of land with distinct management characteristics cultivated for a particular purpose, i.e. food crops, bananas, coffee, permanent pasture, fallow and woodlots. The banana fields are planted with fruit and medicinal trees, condiments, tobacco and foodcrops, but are so-named because bananas form the dominant overstory. There are also isolated banana trees on the food crop fields. Mixed cropping is practised on both types of fields. Banana fields are located closest to the home compound, while woodlots are the farthest. The difference between fallow and permanent pasture is only a matter of the length of time in which a field is left to regenerate. Fallows often last only one season, particularly on the smallest farms, and the vegetation on the fallows consists of weeds and low quality grasses.
Comparison farmers have an average of 12.7 fields averaging 142 square meters. The large number of fields indicates both farm fragmentation and intensive micro-management of the land to take advantage of variations in soil conditions, slope and aspect. Tree experts have an average of 17.2 fields averaging 738.4 square meters. Almost all farmers in this group indicated having fields with food crops, bananas, coffee, and trees.
In both groups, pastures and fallows were being shortened and/or abandoned altogether, because as farm sizes diminish, all land is needed for crop production. Converting fallow and pasture land to crop land also means less fodder which leads to a decreasing manure supply. Soil fertility thus suffers from shortened or abandoned fallows and pastures as well as from diminishing manure supplies.
Soils: Two soils could be specified for each field type. The names of the soils that farmers gave were the Kinyarwanda names. The table below provides an overview of the main characteristics of the soils found in the three districts of this study.
Farmers soil preferences for particular crops are not absolute because not all farmers have fields with desired soil types. In general, urusenyi was the most preferred soil type, followed by mugugu; both these soils were used for food crops, bananas and coffee. Mugugu soils and less desirable soils such as inombe, umusenga and igishonyi, were the most common soils under pasture, fallow and woodlots. Very few people were cultivating urunombe soils.
SOURCE: ISAR 1991
The choice of soil for a particular crop depends on location, slope, aspect and distance from the home compound. For example, people do not plant bananas at the fringe of their land even if the soil is optimal because of the risk of theft if they are out of sight. Such land was used for woodlots. The matter is even more complicated because there is much variation within each soil type, even though it still is called by the same name (the data for each soil given in Table 7 are averages calculated from a number of samples of each soil type taken in a Département d'Etudes de Milieu et des Systèmes de Production (EMSP) of ISAR study. Also, due to the mountainous nature of the Rwandan landscape, soils can vary greatly even over a short distance.
These soil variations explain in part the active market in leasing land in order to gain access to plots of land with slightly different conditions. The high number of fields and their seemingly fragmented nature provide access to an array of sites with diverse soil and ecological conditions which can be managed (as much as possible) as homogenous units.
Animals, especially cattle, are an important part of Rwandan farming systems. Traditionally, cows not only had symbolic value, they were a sign of wealth. Today, cattle are still a sign of wealth but are also a source of manure. Having cattle means being able to maintain soil fertility and, thus, good harvests. In the ranking games, animals were mentioned more than farm size to differentiate good and highly-knowledgeable farmers from others.
The results of the socio-economic survey confirmed the findings of the ranking games. Fewer comparison farmers possess animals (71.4 percent of comparison farmers compared to 90.9 percent of tree experts), and even if they have animals they have fewer than the tree experts.
Only 20 percent of comparison farmers have cattle, on average one cow per farm, compared to 63.6 percent of tree experts with an average of slightly more than two cows. Pigs are the most common animal, followed by goats. Sheep are the least popular of the larger domestic animals but are kept as companions for cows, since farmers believe sheep make cows more docile and easier to handle. They are not kept for meat since, according to local preferences, it does not taste as good as goat or beef.9
Chickens and rabbits are not widespread, even though they are not land dependent. Bee keeping is practised by a small number of farmers, principally for honey. Honey gets a good price on the market and is also used to produce mead and to make sorghum and banana beer sweeter and more alcoholic.
Animals' presence on the farm does not necessarily mean that farmers own them. As in the case of land, animals can be leased. A farmer is obliged to take care of a leased animal but has the benefit of manure and milk. Animal owners without sufficient pasture land can maintain ownership while leasing their animals to other farmers who feed and take care of them.
8 The total area cultivated, however, may be enhanced by renting or borrowing land from other farmers to make up for the lost amount of owned land. As is discussed later in this chapter, renting and borrowing land is common in the study areas.
9 In contrast, people in the Buberuka Highlands in northern Rwanda rear and eat sheep on a larger scale than farmers in southern Rwanda.