4. SUMMARIES OF TECHNICAL PRESENTATIONS

4.2 Fodder production technologies

Survey on agroforestry technologies used for fodder production in Meru District, Kenya
(P.M. Murithi, H.J.C. Thijssen, D.N. Mugendi, J.N. Mwangi, M.K. O'Neill and O.Z. Nyaata)

Fodder shortage especially during the dry season was identified to be one of the key factors limiting dairy production during a diagnosis and design survey carried out in the coffee-based land-use system (LUS) of the Central and Eastern Provinces of Kenya (Minae and Nyamai, 1988). An investigation into how farmers are currently managing the fodder shortage was felt to be important for the agroforestry (AF) research project to identify farmers' priorities and preferences. This would help focus the project activities at solving the farmers' priority problems with appropriate interventions through the introduction of improved AF and/or non-AF technologies. It was for this reason that this survey was conducted in Meru District during May, 1992 with the following objectives: a) to identify the types and management practices of trees used by farmers as sources of fodder; b) to identify the non-agroforestry options farmers use to alleviate fodder shortages; and c) to assess the farmers' preferences in using agroforestry and non-agroforestry options in alleviating fodder shortages.

Central Imenti Division was selected for the survey, as it is one of the major coffee growing divisions in the district. Fifty-seven farmers, randomly selected from lists of coffee producers kept by coffee factories in the three coffee growing agro-ecological zones (AEZ), were interviewed. Informal discussions and questionnaires were used to gather the required information from key informants and farmers. The altitude of the study area ranged from 1210 m to 1890 m above sea level (masl) with a mean of 1567 masl. The equator passes through the centre of the survey area. Ninety-three percent of the respondents were either the owners of the land or their wives. A relatively large proportion (32%) of the respondents were women although the majority of the farms (97%) were registered under the names of men. Fifty-one percent of the farmers belonged to registered dairy co-operative societies indicating the importance of milk sales as a source of farm income.

Farm sizes ranged from 0.26 ha to 7.7 ha with a mean of 2.2 ha. The majority (77%) of the farmers owned less than 3 ha of land. Ninety-eight percent of the farmers had coffee. Only 14% of them, all found in the upper zone, grew tea. Fifty-four percent of the farmers sold food crops. Maize was sold by most (65%) of the farmers, followed by beans. Cattle was owned by 98% of the farmers with the majority (90%) having improved cows. Poultry was the second most popular species kept. The semi-zero grazing system of keeping animals was practised by most (73%) farmers. The highest amount of milk produced ranged from two to 23 litres per animal per day with a mean of seven litres. The lowest amount produced ranged from one to nine litres per animal per day with a mean of three litres. The majority (84%) of the farmers who sold milk sold it to their neighbours. Other farmers (45%) sold the milk through the milk cooperative societies.

Seventy-five percent of the farmers indicated they experienced different types of problems in keeping dairy animals. Eighty-three percent of those able to identify the problems indicated that the lack of adequate feeds especially during the dry seasons was the main constraint; followed by a general lack of sufficient cash to buy feeds.

Napier grass (Pennisetum purpureum) was being used by nearly 98% of the farmers with cattle. Other dairy feeds included crop residues, naturally growing pastures, commercial concentrates and tree leaves/branches. Eighteen percent of the farmers with cattle fed tree leaves/branches to the dairy animals. Fifty-seven of them indicated they had at one time or another given commercial concentrates to the dairy animals mainly during the milking time. This indicates that farmers appreciated the importance of supplementing the dairy animals with concentrates for higher milk yields. However, only 25% of them used commercial concentrates as a main source of dairy feed on a routine basis. Forty percent of the farmers indicated they sometimes encountered problems while feeding concentrates to dairy animals. Lack of money was mentioned by the majority (70%) of them.

A wide range of indigenous trees were fed to the animals mainly during the dry seasons. Cordia africana was used by most (73%) farmers, followed by Grevillea robusta which was being used by 66% of them. Most farmers, however, indicated that they did not necessarily feed these trees to increase milk yields but to have something to prevent animals from starving given the scarcity of other feeds during those periods. The commonly recommended fodder shrubs such as Calliandra calothyrsus, Gliricidia sepium, Leucaena spp. and Sesbania sesban were absent on most farms. The trial-and-error method was used by most farmers in knowing which tree species were preferred by the animals. Most farmers indicated they started feeding fodder trees at the time they started keeping dairy animals. Farmers in the survey area had been in dairy production for an average of 16 years. This indicates that the introduction and adoption of appropriate fodder trees in the area would be relatively easy since farmers already appreciate the role trees can play as a source of dairy cattle feeds.

Analysis of leaf samples collected from popular trees mentioned during this survey and the ethnobotanical survey carried out in Embu (Thijssen et al., 1993) showed there were some indigenous trees which were of high nutritive quality. Some of these trees compared favourably with the normally recommended exotic fodder shrubs. However, some of the trees such as Grevillea robusta, Cordia africana etc, though high in dry matter content and being fed to animals, are low in other qualities such as organic matter digestibility and protein content. Also other trees such as Lantana camara are noxious weeds.

Eighty-seven percent of the farmers thought trees had some potential of alleviating their fodder shortage problems. The majority (51%) of them, however, indicated either they did not know the trees they would need to plant or they would plant any tree they are advised to be good for fodder. The majority of the farmers who mentioned trees that could be planted for fodder identified Cordia africana (32%) and Grevillea robusta (26%) as the main trees they would plant. It seemed that farmers were identifying trees they were familiar with and not necessarily due to their high fodder qualities. Boundaries were mentioned by the majority of farmers as the main niches to plant fodder trees but there was no particular reason why they were preferred except probably that these are the relatively under-utilized portions of the farm.

Farmers' nurseries, collection of tree wildings, the Forest department and women groups were the main sources of the trees planted by farmers. There were, however, a few (9%) who did not know where they could obtain the trees.

In conclusion, it was observed that farmers in the area appreciated the role trees play in providing additional fodder for their dairy animals, especially during the dry season when other alternative feeds are scarce. The majority of the trees fed were indigenous and some of them were of relatively low nutritive quality. The project, together with the extension personnel could, therefore, expose the farmers to other fodder trees which are of higher quality than those presently being used. Some indigenous trees used for fodder seemed to be of high quality and therefore need to be further evaluated. Management strategies need to be developed for some of these trees so as to have a wide range of fodder trees to select from, especially when exotic ones are attacked by new or unknown pests or diseases. Farmers need to be invited to visit the project site and other farmers with fodder trees to see some of the tree species that can be planted. They can also be provided with seedlings to plant so they can make their own assessment.

Farmers identified various niches where they could plant trees for various purposes but there was no obvious reason for any preference. This aspect and others could best be accurately determined through diagnostic tree planting studies. In general, the majority of farmers were not specific on what qualities they would consider to be important in choosing particular tree species for fodder. The general observation was that they would try any trees recommended to them.

References

Minae, S and D. Nyamai (1988). Agroforestry Research Proposal for the Coffee-Based System in the Bimodal Highlands, Central and Eastern Provinces, Kenya. AFRENA Report N°16. International Centre for Research in Agroforestry, ICRAF. Nairobi, Kenya.

Thijssen, H.J.C., P.M. Murithi, O.Z. Nyaata, J.N. Mwangi, I.O.O. Aiyelaagbe, and D.N. Mugendi. 1993. Report of an Ethno-botanical Survey of Woody Perennials in the Coffee Zone of Embu District, Kenya. AFRENA Report N°62. International Centre for Research in Agroforestry, ICRAF. Nairobi, Kenya.

Spatial variation in growth of Calliandra calothyrsus and Leucaena diversifolia on terrace benches in Southwest Uganda
(D. Peden)

Two non-replicated fodder banks were established at Kachwekano in October 1990, on a single bench terrace. The first one, consisting of Calliandra calothyrsus, was established with contour rows spaced 1 m apart and an intra-row spacing of 25 cm. The second fodder bank, composed of Leucaena diversifolia, was planted in two different inter-row spacings (1m and 2 m apart, giving 8 and 5 contour hedges, respectively on the bench) and an intra-row spacing of 25 cm. The fodder banks serve primarily as a supply as surplus fodder and mulch. They also serve as an observation of MPT behaviour on the steeply terraced lands of southwest Uganda. The fodder banks are pruned periodically to supply fodder and mulch. The Leucaena block is producing large quantities of seed, and it has not been pruned since November 1991. The Calliandra was last pruned in November 1992.

During on-going monitoring of the fodder banks, stem heights were measured prior to cutting in November of 1991 and 1992. In 1992, five Calliandra plants were selected at random from 10 m of row, and the heights of all stems were measured and averaged. In 1991, ten Calliandra were selected from a 20 m length of row. With Leucaena, 10 trees were sampled from a row length of 5 m in both years. In 1994, the trees were not pruned, but measurements of plant height and root collar diameter were made. Ten trees were selected at random from the rows, but exposed guard trees at the ends of the rows were excluded. Stem counts for each selected plant were made.

During the first year of growth, both Calliandra and Leucaena demonstrated variable growth patterns, with heights being greater on the lower sides of the bench. In the early years, the trees tended to be chlorotic. Inspections of the roots failed to reveal any nodulation. However, after two or three years, both Calliandra and Leucaena developed nodules. Differences in height among rows decreased, and the foliage developed a consistent rich green colour.

During the early years, stem counts and root collar diameters were not measured. However, in 1994, the data indicate that trees from the central and upper rows were smaller than in the lower rows, in spite of the small differences in height. The highest root collar diameters and stem counts were found on the lowest row. It is probable that much of these observed differences results from the residual effect of the initial low growth of the central and upper trees. It is also likely that the trees in the lower row suffer the least amount of competition for light. The downward side of this lower row is exposed to considerably more light than any of the other rows. The Leucaena with 2 m spacing between rows demonstrated less suppression of growth in the central and upper rows than the Leucaena with 1 m spacing. This also suggests that competition may account for the observed differences among rows.

A thick layer of litter accumulated during the three year period under both Calliandra and Leucaena. Within the fodder blocks, each row demonstrated the establishment of mini-terraces (approximately 10 cm high). It is likely that soil loss was reduced to nearly zero under the fodder banks.

Table 13. Growth characteristics by row of Calliandra and Leucaena growth on terrace benches at Kachwekano, Kabale, Uganda

These observations suggest that a solid block of leguminous trees can overcome the observed effect of terrace scouring, conserve soil, and possibly, run-off. The changes over three years indicate that eventually, both Calliandra and Leucaena can pick up sufficient Rhizobia to permit nitrogen fixation. The results also suggest that after three years of intensive cultivation, the soil micro-flora was depleted. Rehabilitation of the degraded and steep croplands may be possible with the use of fodder banks or improved leguminous fallow.

Palatability and intake of 8 woody species suitable to the Crete Zaire-Nil region
(J. Ugeziwe, A. Niang, E. Styger and A. Gahamanyi)

From both the climatic (frequent hail, limited sunny period, etc...) and edaphic (very acid soils with higher aluminium contents) characteristics prevailing at the Crete Zaire-Nil region, many fodder shrub species which are disseminated through development projects are not adapted to the region due to those particular conditions. Therefore, it was necessary to look into alternatives from a wide range of species tested by ICRAF/ISAR project in the case of its general species screening work and which are revealed to be more suitable to the region.

Eight woody species have been therefore tested in 2 trials comprising Acacia koaia, Chamaecytisus palmensis, Acacia koa, Hagenia abyssinica and Setaria splendida for the first trial and Mimosa scabrella, Acacia melanoxylon, Acacia mearnsii and Alnus acuminata for the second.

The experimental design is that of a split plot with 5 treatments which are in fact 5 samples of animals with 3 weaned local breed young goats and aged 5-6 months with a mean live weight of 10.4 kg.

A daily meal of 4 kg Setaria + 2 kg woody forage has been given at each animal (i.e 12 kg of Setaria + 6 kg woody forage) at 2 feeding times (9 h and 15 h).

Concerning the 1st trial, the more digestible woody species are: Acacia koaia, Acacia koaia and Chamaecytisus palmensis with a refusal rate of 40%, 50 and 51% respectively. This rate being 45% for Setaria splendida. For the 2nd trial, the more palatable species is Mimosa scabrella with a refusal rate of 50%, and for Setaria splendida 51%.

Complementing Setaria splendida with each of those 8 species has improved total dry matter intake in relation to Setaria alone. More important increases were obtained with Mimosa scabrella (90%), Acacia koaia (88%), Acacia koa (85%) and Chamaecytisus palmensis (78%). The lowest increase was obtained with Alnus acuminata with 45%.

The chemical composition of different forages indicate that higher crude protein contents are achieved by Mimosa scabrella and Chamaecytisus palmensis with 24.9 and 23.3% respectively.

Crude protein contents of other species Acacia koaia, Acacia koa, Hagenia abyssinica, Acacia mearnsii, Alnus acuminata, Acacia melanoxylon and Acacia koaia varied between 15.6% and 16.53% while that of Setaria splendida is 6.97%. Quality analysis of tannins show higher contents for Acacia mearnsii and Hagenia abyssinica and lower contents for Acacia koala and Mimosa scabrella.

The contribution of trees and agroforestry products to the cash budgets of households in Western Kenya
(S. David and L. Otika)

The income and expenditure patterns of 30 households during 1992 were reconstructed, in order to evaluate where the spending priorities of resource-poor households lay, and to determine whether they have sufficient income to invest in agroforestry technologies. In addition, the study provides baseline information on household cash flows for future assessment of improvements in household welfare, as a result of agroforestry interventions. To cross-check the reliability of the 1992 data, budget data were also collected from 17 of the same households during April 1993.

The sample was selected from among households involved in technology testing with the Maseno On-farm Agroforestry Research Project. These households are considered representative of the wider population, in terms of household demographic characteristics, farm size and resources. A random sample was not used because of the sensitivity of the information being collected and the need for close rapport with informants. The households surveyed were purposively selected to represent three wealth categories (average, below average and above average) and the three districts covered by the project (Kisumu, Siaya and Vihiga). Most of the 30 households (67 percent) were headed by resident men and 43 percent were classified by interviewers as below average in wealth. The average household size is 7 persons and the mean farm size (farmers' estimates) is 1.3 hectares.

Median cash income in 1992 for the households surveyed was KShs 15,810 ($506)¹. As table 14 shows, most household income (72%) came from off-farm sources, namely: employment and income generating activities (38%) and gifts and remittances (34%). After crops, the next most important on-farm sources of income were livestock sales (8%) followed by trees (3%), milk (1%) and wood products (1%). Although the sale of trees for poles and timber accounting for only 2 percent of mean income, 3 percent of heads of household indicated that the sale of trees was their most important source of income in 1992. The sale of trees was ranked as the second and third most important sources of income by 11 and 12 percent of household heads. Charcoal and firewood sales provided an insignificant proportion of household income in 1992 and were ranked as third most important sources of income by 4 percent of household heads and 18 percent of partners. The low contribution of tree sales to the household economy in 1992 does not appear to be atypical, since between 1990 and 1992, half or less of the households surveyed sold trees for use as poles, timber or firewood. An average of 37 stems were sold between 1990 and 1992.

Table 14. Cash income sources, 1992 (medians in KShs)

Of the 18 heads of household who reported ever selling trees between 1990 and 1992, 72 percent consider trees as a very important source of income. Nearly a quarter of heads of households and their partners respectively considered firewood to be a very important source of income generally. It is significant that wives regarded trees as a less important source of income in 1992 and generally, reflecting male control over this resource. Among those wives reporting ever selling trees (N=13), 54 percent considered trees as a very important source of income. The difference between farmers' perception of trees as important sources of cash income and the low amounts of cash income most received from trees in 1992 highlights the importance of trees as a store of capital which farmers can fall back on when cash is urgently needed. More commonly, to obtain large sums of cash for investment and to alleviate periodic cash shortages, farmers rely on loans from friends and relatives.

Sampled households spent a median of KShs 32,683 (US$ 1,047) during 1992, a figure which exceeds median income by over 100 percent. Although higher cash outflow relative to cash inflow reflects farmers' tendency to under-report income, it also indicates the insignificant amount of savings available to resource-poor households. Approximately 89 percent of household cash income was spent on nonfood necessities (the most important categories being household items, school-related expenses and cultural and social obligations) and food. The low input nature of farming among smallholders is evident by the small proportion of total expenditure (7%) spent on farm inputs such as hired labour, fertilizer, manure, tools, tree seedlings and livestock maintenance (feed and health costs). Average household expenditure on poles and timber accounted for less than 1 percent of total cash expenditures. Low expenditure on wood products is not surprising since such purchases are only made periodically, and because a significant proportion of households in Western Kenya obtain at least some of their wood supplies from their own farms (table 15).

Answers to the hypothetical question "what would you do if you had more money" reveal interesting patterns about farmers' investment priorities. Judging by the priority given to investment in business activities and livestock (especially dairy cattle), farmers' paramount budgetary strategy involves ensuring a reliable cash inflow within a relatively short period of time. Few farmers indicated willingness to invest in agriculture because of the low expected returns, the risks involved and farmers' subsistence mentality toward farming. Because of farmers' concerns to ensure a steady flow of income throughout the year to meet basic needs, the role of trees in household budgetary strategies is seen from a long-term investment perspective whereby trees are seen as a store of wealth.

Table 15. Areas of cash expenditure, 1992 (medians in KShs)

The picture that emerges from quantification of household income and expenditure patterns in Western Kenya is one of significant cash shortages in most households. This raises questions about the adoptability of agroforestry technologies which require cash outlays for planting material, labour, and fertilizer, among other inputs. Additionally, in assessing the adoption potential of agroforestry technologies, researchers need to consider the contribution of trees and livestock to meeting the cash needs of farming households within the context of the various budgetary strategies pursued by households. Further analysis of the data is underway.

Effect of different pruning techniques on fodder biomass production of Leucaena hedges
(R. Thijssen)

Tree fodder production is receiving more attention as dairy production in Embu is taking up and coffee revenues are rapidly decreasing. One way to produce tree fodder is through hedge systems. Farmers in Embu District, Kenya, indicated a keen interest in planting fodder hedges as a fence around their compounds as well as along the entrance road to the compound and in some cases on the boundary of there farms (Thijssen et al. AFRENA Report N°65).

Management regimes of tree hedges have received much attention since pruning of leafy material affects the root/shoot ratio and the photosynthetic capacity of the plant. These factors are important to productivity, and thus, a better understanding of the effects of pruning on biomass productivity can lead to improved production systems.

Cutting height and interval of subsequent cuttings are the most studied management factors so far. The level of reduction of above ground biomass at a fixed cutting height showed to be another interesting management factor. Trees in some hedge experiments at the KEFRI/KARI/ICRAF Agroforestry station in Maseno, Kenya, as well as in the KARI/ILCA project at Mtwapa, Kenya, are super-pruned, i.e. the trees are stripped of all side branches. At ICRAF's Research Station in Machakos, experimental hedges are pruned the way (aesthetical) fences around compounds or gardens are normally done. These different ways of management do not only have an effect on productivity but have even implication for, for instance, the quality of living fences. Furthermore, an interesting finding from Indonesia is that several woody species produce more fine, superficial roots that can be classified as competitive, when more above ground biomass is removed.

In an experiment at the KARI/KEFRI/ICRAF Agroforestry research site in Embu, Kenya, the two above-mentioned pruning techniques were imposed on a one-year old Leucaena leucocephala fence, starting July 1992.

A third pruning technique was introduced, and can be described as the way hedges are pruned on the Machakos Research Station, while remaining leaves are all stripped manually from the branches. This treatment was considered to be somewhere intermediate between the two other pruning treatments as far as the level of reduction of above ground biomass is concerned. Cutting height in all treatments was 1 m and length of hedges in the experimental plots was 10 m.

After four prunings in the months of July, October, December and March, the results show a relatively low leafy biomass production from the hedges that are super-pruned. This technique results in a very open living fence too. The pruning technique, where remaining leaves are stripped from the branches, is labour intensive but results in^the highest percentage of leafy biomass.

Mean dry yield of leaves (g/10 m)

Cutting management of Calliandra calothyrsus for diy season fodder production
(E. Akyeampong)

The lack of fodder in adequate quantities during the dry season was a common problem encountered in the D&D exercises in the East and Central Africa highlands. This trial was designed to investigate the cutting management of Calliandra during the wet season that will maximise leafy biomass production in the dry season and to assess the effect of cutting times on annual biomass production.

Calliandra calothyrsus was allowed to grow unmanaged for 20 months after which they were cut back to 1 m in October 1991. Subsequently, a treatment was harvested each month from January to May. Finally, all treatments were harvested in August at the peak of the dry season. A plot was 5 m long and 4 m wide with 4 rows of Calliandra spaced at 1 m apart. Within row spacing was 50 cm. A randomised complete block design was used and was replicated four times.

In both years, the highest dry season yields were obtained when the wet season harvest was done 6 months earlier (in February)(figure 5). Beyond six months leaf yield declined because of senescence.

Fig 5. The effect of the length of interval between wet and dry season season harvests on the leaf yield of Calliandra calothyrsus

In 1991/92 highest total (wet + dry season) leaf production was obtained when wet season cutting was in May and in February in 1992/93 (figure 6). Only two harvests one at the beginning of the rainy season and the other at the peak of the dry season resulted in the lowest leaf yield.

Fig 6. Influence of time of cutting during the wet season on total
(wet + dry season) leaf yield of Calliandra calothyrsus

Effect of time of first harvest on the biomass production of Calliandra calothyrsus and Leucaena pallida
(E. Akyeampong)

Calliandra calothyrsus is proving to be one of the best yielding species at all sites in the zone. As fodder species, Calliandra is often cut in the first year or soon thereafter. We are not aware of any systematic study to look the age at first cut during the first year or two of growth on fodder production. Leucaena pallida has never been grown in Burundi. It was included in this trial to assess its potential as a fodder species.

The experiment was established in Mashitsi, Burundi, in March, 1991. The species were cut at 3, 6, 9, 12 and 15 months after transplanting. Subsequent to the first cut, pruning continued every three months.

For Calliandra, the later the first time of harvest, the greater the quantity of total biomass obtained compared to treatments which were cut a few more times beginning at a younger stage (table 16). For example, from one lopping at 15 months after planting, 12.9 t/ha of biomass was obtained whereas for the treatment cut five times beginning at three months after planting yield was 7.0 t/ha. Calliandra was in three distinct groups, those cut during the first six months yielded about 7 t/ha those cut at nine or 12 MAP yielded about 10 t/ha and those cut at 15 months, 13 t/ha. Leaf production of Calliandra was not significantly affected by the time of first cutting.

Table 16. Cumulative biomass production in the first cycle

Total and leafy biomass of Leucaena were not significantly affected by the time of first cutting. Regrowth in the second year was not affected by the time of first cut (table 17).

Table 17. Regrowth in the second year as influenced by time of first cut

Fodder trees in agroforestty: Availability of information and research methodology
(R. Thijssen)

Projected increases in human population throughout the developing world and the limited availability of land for increased food and forage production suggest that agricultural production will need to be intensified considerably if current human and animal nutritional standards are to be maintained. In many of the ruminant-production systems in the developing world, grass (pastures and feed gardens) make up the bulk of feed. The low quality and seasonal nature of the forage supply, together with the low intake by animals and the poor digestibility of the forage, are major factors contributing to the low productivity of ruminant animals.

The principle aim of any forage-improvement program should be to maximize the use of basic feed resources, which, in most situations, will be crop residues or low quality roughages. One suggested method is the addition of leguminous forages that will act as a nitrogen supplement to the low-quality roughage materials.

Many leguminous and non-leguminous tree and shrub species have been documented as useful animal fodders. These range from short-stature shrubs such as Desmodium discolor to tall upperstorey trees. In some countries shrub and tree fodders have been used as livestock feed since ancient times. In Nepal fodders from more than 100 different kinds of trees and shrubs are used as animal feed. Several authors have pointed out that most trees lopped for fodder do not fix nitrogen.

A fodder database has been developed using Micro CDS-ISIS software. Literature information regarding fodder potentials of grasses, herbs, shrubs, and tree species is collected and stored in this database. Searches can be made on species names, botanical family names, countries, and specific animal group. Fields of information include fodder analyses, palatability and digestibility, phenological data on the plant species, experimental results with animals, reported ill-effects, the ICRAF library code, and an abstract from the publication (see example). At the begin of September 1993 the database contained some 600 entries, representing about 300 species.

A four step approach for fodder research with the aim to acquire more information on the usefulness of tree and shrub fodders is suggested:

1. Identification of potentially valuable fodder species. Sources of information are farmers' experience, literature, and databases such as ICRAF's MPTS database and the Fodder database.

2. Propagation of selected species. In order to produce a substantial amount of fodder on many farms a simple way of generating planting material should be found.

3. Management practices on the plants for a sufficient and lasting production of fodder and integrated into the farm design.

4. Animal production/interaction/reaction on feeding fodder material from a specific woody species.

Few quantitative evaluations have been made of the contributions that tree and shrub fodder materials make to animal production. There are problems associated with such evaluations: e.g., the long growth period to maturity for the tree and the difficulty of deciding on experimental conditions such as planting density, cutting height, and cutting frequency. For a formal experiment, many animals are required and, consequently, a large fodder production area. Assessing the feeding value of such fodders is difficult, where the proportion in the diet can vary markedly. In all evaluations, there should be a clear definition of the purpose for which the fodder is to be used. The requirements for a sole feed may be quite different from those of a supplement to low-quality roughages. And the choice of experimental animals is important: is it possible to use chicks or rabbits as "guinea pigs" in fodder research?