P. Nyathi
Dept. of Research and Specialist Services,
Makoholi Research Station, P/Bag 9182, Masvingo, Zimbabwe
Introduction
Soils
Towards a defination of views
Dambo distribution and morphology
Dambo vegetation
Significance of dambos in peasant farming communities
Conclusion
References
Workshop summary and recommendations
Overstocking has been identified in most agricultural literature in Zimbabwe as the main factor that has led to degradation, reduced grass growth, and the replacement of decreaser grass species by increasers and/or invaders. The latter are mostly unpalatable to livestock. Invader grass species although prevalent in degraded dryland grazing areas are also found in "dambos". Problems of dambo utilisation hinge on the ability of dambos to support large livestock numbers during the dry months of the year when grass elsewhere in the grazing area is low in nutritive value and quantities are also low. That livestock numbers have either been maintained or decreased during drought years, and increased during years of copious rainfall raises the question of whether overstocking is the cause of the problem. It may be argued that dambos have formed a buffer between starvation and death, and therefore their management, development and utilisation is the key to increased livestock production rather than reduced livestock numbers as suggested by proponents of overstocking as the cause of the problem.
This paper highlights the soil types, the extent of dambo distribution in Zimbabwe, problem of definition, morphology, vegetation and significance of dambos in peasant agriculture.
At the end of the paper, preliminary data to show trends in dambo herbage production is presented.
There are many different soil types in Zimbabwe. There are soils derived from granite (most common rock), predominantly that named sandveld (with little clay content) which drains well, dries quickly and has low organic matter (OM) and phosphorus (P); then there is paragneiss - a soil which is a result of the weathering of an unusual rock type reformed from granite (brown, slightly red sandy loams) (Bishop, 1975).
Black turf soils (very fertile, heavy clay, hard when dry, very sticky when wet) are derived from a less common rock, basalt. Dolorite weathers to rich red clay soils scattered throughout Zimbabwe. Kalahari sandstone weathers to Kalahari sands (very fine grained soils of low fertility found in the drier regions of the west of the country. Serpentine, the most common rock on the "Great Dyke", weathers to shallow greyish brown sand loam soils unsuitable for crop production because of presence of nickel and chromium salts that are harmful to plant growth. Norite is a rock found in some sections of the Great Dyke and forms clay soils more like black turf soils (north and south of Selous).
Mopani soils are found in dry areas, are composed of heavy clay and are very alkaline. Two main types of dambo soils occur in Zimbabwe (Whitlow, 1985) and have been characterised thus;
a. Calcic hydromorphic soils comprise dark grey or black clays with a high base status. The dominant clay is montmorillonite, hence these soils resemble vertisols in their behaviour with respect to expansion and contraction during wetting and drying phases. Whitlow (1985) further stated that carbonate concretions sometimes occur in subsoils within zones of fluctuating water table. Calcic hydromorphic soils occur in broad depressions on mafic rocks in areas receiving over 600 mm rainfall per year. The mafic rocks comprise dark humic soils overlying light coloured sandy or sandy clay soils which generally have a low base status, with kaolinite being the prevalent clay mineral. Whitlow (1985) found that mottling and iron concretions are common in subsoils within dambo margins.
b. Non-calcic hydromophic soils originate from siliceous parent materials and are the most extensive of the dambo soils in Zimbabwe covering in excess of one million hectares of dambo land (Whitlow, 1984a).
Peaty dambo soils, comparable to those found in parts of Zambia (Brammer, 1973) are quoted by Whitlow (1985) as representing a third type of hydromorphic soil. Three main pedological process occur in the non-calcic hydromorphic soils:
i) organic matter accumulation
ii) gleying and
iii) clay translocation
It should be noted here that organic matter accumulation is an advantage to the establishment of pasture legumes on dambo sites. This is reinforced by further findings (Whitlow, 1985) that dambo soils are characterised by well defined organic surface horizons which increase in depth from dambo margins to wetter sites, and that the OM is generally well decomposed on drier areas (margins) but becomes progressively more fibrous and peaty towards low-lying waterlogged sites. The latter fact supports the idea of introducing pasture legumes on the periphery of dambos with a view to improvement of the quality of forage.
Gleying effects relate to both permanently and periodically saturated soil horizons. The permanently wet horizons are typically pale grey to white in colour due to a decrease in free iron content from the dambo margin to its eye where a reduction of iron compounds into a ferrous state occurs (Whitlow, 1985). Microbial activity is limited by the acid, waterlogged conditions (Whitlow, 1985) but Purves (1976) points out that microbial activity under anaerobic conditions assists in the process of iron reduction thereby enabling its removal by lateral drainage. Removal of iron and production of bleached or pallid subsoils is a process regarded as ferrolysis (Brinkman, 1985) and would seem to occur in some dambos in Zambia as well as in Zimbabwe (Whitlow, 1985). Waston (1964) described ferricrete formation at a depth of 2.5 metres within a dambo near Harare and explained that, that marked the lower limits of the dry season water table level. This observation suggests that the soil above 1.5 metre depth should be moist enough to support both grass and pasture legume growth during the dry periods of the year and is part of the basis of the hypothesis.
Locally known as "vleis" in Zimbabwe, and dambos in central and southern Africa, these seasonally wet areas are not adequately defined. The problem of dambo definitions in Zimbabwe has been alluded to by Rattray et al. (1953), Thompson (1972) and Whitlow (1980, 1984b). The definition given by Ivy (1981) is accepted officially for the purposes of legislation in the Natural Resources Act.
|
Author and year |
|
Definition |
|
Ivy |
1981 |
Land that is saturated to within 15 cm of the surface for the major part of a rainfall season of average or above average rainfall and which may exhibit one or more of the following characteristics: i. the presence of mottles or rust-like stains in root channels within 15 cm of the surface. ii. a black topsoil horizon very rich in OM overlying pale leached sands; iii. a dark grey or black heavy clay showing considerable surface cracking when dry. |
|
Thompson |
1972 |
A vlei is a depression at the head of, or flanking, a water course in which the soils are saturated during the rainy season, and which remains saturated to within 50 cm or less from the surface for some considerable period thereafter. |
|
Rattray et al |
1953 |
A low-lying, gently sloping, treeless tract of country which is seasonally waterlogged by seepage from the surrounding high ground assisted by rainfall, and which contains the natural drainage channel for the removal of excess run-off from this surrounding ground. |
All of these definitions do not account for those areas that are waterlogged throughout the year with visible fluctuations in the water level as seasons change into the driest part of the year. All three definitions recognise seasonal waterlogging during the rainy season but do not indicate when this water logging ceases during the year. This raises my contention that, if to a great extent these lands are waterlogged during the rainy season, this is so because their use at that time is not necessary as grazing elsewhere on topland would be at its best. It is during the dry periods of the year that these waterlogged lands play their role as fodder banks (not grazed or sparingly grazed during the rainy season, and fully utilised during the dry months).
Dambos cover an area of about 1.228 million hectares or 3.6% of the total land area in Zimbabwe (35.6 million hectares) (Whitlow, 1984a) (Table 1). Over one million hectares or nearly 90% of dambos in Zimbabwe occur on gneisses or intrusive granites. The former are mainly tonalitic in composition, with sodic feldspars being common, and the latter are dominated by adamellites with potash feldspars being common (Whitlow, 1979). Dambos are confined mainly to the headwater regions of rivers draining the central plateau of Zimbabwe, and account for up to 30% of the land areas of granitic terrain (Whitlow, 1980).
Morphologically, they vary from broad, lobate depressions through to narrow linear features extending several kilometres down valleys. They also vary greatly in their materials. Whitlow (1985) sound a warning that this variable character of dambos makes it difficult and dangerous to generalise about the effects of human activities on dambos (each situation is unique in some way). As earlier noted, some dambos are wetter and for longer periods than others and it is this feature which should be used to advantage for pasture legume introduction into dambo vegetation. The wetness is an advantage in that it provides abundant moisture for plants; ensuring a reasonable harvest even during the drought seasons (Mazambani, 1982). The disadvantage is that waterlogged soils are typically badly aerated necessitating removal of excess water to enable proper root development (Whitlow, 1983). Butzer (1976) noted that vleis in Africa are best developed on plateaux where for about five months rainfall ranges from 800 to 1300 mm per annum. Bond (1963, 1967) advances the idea that dambo processes may have definite climatic limits. In Zimbabwe, dambos certainly are prone to drying out during the long dry season, perhaps more so than those in higher rainfall areas of between 700 to 1000 mm per annum, and are certainly prone to drying out during the long dry season, perhaps more so than those in higher rainfall regions in Zambia (1000 mm per annum) and Malawi (Whitlow, 1985). Butzer (1976) suggested that some dambos are inactive in Zimbabwe but did not indicate how he reached that conclusion. However, Goudie (1983) explained that there is likelihood that, with changing environmental conditions during the late Quaternary period, rainfall regimes have at some stage been more favourable for dambo development than those which now prevail. In general, one might regard climatic conditions in Zimbabwe as being more marginal for the development and maintenance of dambos than elsewhere in south-central Africa (Whitlow, 1985).
Table 1. Dambo distribution according to variation in average slope in Zimbabwe.
|
Slope under 2° |
2° to 4° |
Over 4° |
Total area |
|
38.4% |
45.6% |
16.0% |
100% |
|
(492,334 ha) |
(584,646 ha) |
(205,139 ha) |
(1,228,120 ha) |
Source: Whitlow, (1984b)
Most dambos (84% of total dambo area) occur in land areas with slopes of 4° or less. This derives from the fact that 60% of land in Zimbabwe has average slopes of 4° or less (Table 1). The 84% of total dambo area (1,076,980 ha) is land that has potential for intensive pasture development. Cormack (1972) has argued that the dambo water should be put to more productive use in growing crops or improved pastures within dambos, providing there is no damage to soils or stream flow, but to the contrary Whitlow (1985) argued that some farmers in Zimbabwe have undertaken maize production and introduction of legumes in dambos with seemingly little effect on stream flow. The most extensive dambos occur on the relatively flat terrain of the central watershed areas above 1200 m above sea level.
They occur in a rather irregular arc varying in width from about 20 to over 80 km and around 300 km long. Localised islands of dambos occur on plateau remnants outside this main arc (Whitlow, 1985). The distribution of dambos in Zimbabwe as elsewhere in south-central Africa is influenced by factors such as relative relief, regolith (bedrock) characteristics, and climatic conditions (Mackel, 1974; Whitlow, 1984b). To date, no detailed analysis of the relationships between dambos and these environmental factors has been carried out. However, Whitlow (1985) pointed out that low relief and gentle rivers with increasing distance from central watershed results in dambos becoming more localised. Eventually, in the lower reaches of the main river systems below 900 m altitude, the dambos are unable to persist. My view is that these dambos that are unable to persist should be the areas with sufficient residual moisture for sustenance of pasture legumes introduced into the dambo vegetation after the rainy season has terminated. The more complex a plant community in terms of species diversity, the more likely that ecosystem stability is achieved.
Henkel (1931) was the first to describe and classify dambos in Zimbabwe as valley grasslands, and Whitlow (1985) noted that sedges and herbs are common and locally dominant constituents. The absence of trees and shrubs in Zimbabwe dambos was attributed to the inhibiting effects of seasonal waterlogging, periodic frosts, and occasional but intensive fires (Rattray, 1957). However, in many respects vegetation in Zimbabwe dambos is comparable to that of dambos elsewhere in central Africa (Fanshawe, 1969; Werger and Coetzee, 1978). The vegetation generally comprises a mosaic of plant communities which changes in character from the margins to the central portions of the dambo dependent upon the degree and duration of waterlogging (Whitlow, 1985). This pattern of vegetation as described above has prompted my hypothesis that there must be different moisture levels supporting the growth of the different communities forming the mosaic, the moisture levels decreasing progressively away from the centre of the dambo and in effect allowing plants of different moisture requirements to grow within the limits of their tolerance to waterlogging. For this reason, a screening trial of forage legume species for selection of the best-bet types has been proposed on seasonally waterlogged lands. This is validated by the realisation by many researchers that these seasonally waterlogged lands are key grazing resources for livestock during the dry part of the year (June to October) (Whitlow, 1983; Scoones, 1987; Rattray, 1957). It is the availability of these key grazing resources (vleis browse, river banks, and drainage lines) and the facility of flexible utilisation that has sustained high livestock populations in the communal areas for a long period (Scoones, 1987).
In view of the foregoing, it becomes apparent that research aimed at improvement and development of dambo pasture is long overdue. If the carrying capacity is to be increased the key resources mentioned above are the components that research should focuss on, as they directly determine carrying capacity levels. Improvements in the extensive grazing land, although beneficial, will have less direct effect on carrying capacity and, because this is a far larger area than the key total land resource area, any attempts at intervention will be financially and operationally more difficult than a focussed approach (Scoones, 1987). With draught power provision being a primary objective for communal area livestock, development oriented research on key resource areas that provide dry season fodder is vital. Selective use of the key resources opens up the possibilities for selective feeding of priority stock (draught oxen; milking cows), reserved dry season grazing and their development as fodder banks.
The moist conditions which prevail in dambos for most of the year provide a favourable environment for plant growth. However, there appears to have been a preoccupation with the hazards of soil erosion and drying out of dambos in Zimbabwe (Jennings, 1923; Rattray et al; 1953) such that legal restrictions have prevented realization of the potential of these seasonally waterlogged lands through various prohibitive measures enshrined in the legislature.
The most common form of dambo utilisation is for livestock grazing. In commercial farms the dambos are burned from late August to October. The residual soil moisture is adequate to support new growth of herbaceous cover, providing valuable grazing at a time when other sources of feed are at a low level (Whitlow, 1985). In peasant farming areas dambos are used as part of the communal grazing lands. Dambo cultivation has required drainage of excess water. A traditional method developed by peasant farmers is to use ridges and furrows trending down the slope. This system is ideally suited to small-scale hoe cultivation but did not meet the needs of mechanical farming on commercial scale (Whitlow, 1983).
The prevalence of dambo cultivation with rice and tsenza (Cleus esculentus) being important crops amongst peasant farmers was reported nearly a century ago (Bishop, 1975; Whitlow, 1985). Remnants of ancient ridges and furrows occur in many dambos in the wetter parts of Zimbabwe and may date back some 250 to 300 years (Whitlow, 1983). Dambo cultivation is a well established tradition amongst peasant communities in Zimbabwe as elsewhere in central Africa (Walker, 1966; Russell, 1971). Dambo gardens provide a good regular supply of crops for home consumption and for sale in urban centres. They are especially important during drought years when dryland crops are poor, but wet conditions in dambos are still adequate to yield reasonable harvests (Mazambani, 1982). Only one detailed study on the importance of dambos in the agrarian economics of the peasant farming sector of Zimbabwe has been carried out (Thiessen, 1975). This confirmed that there was a positive relationship between the area of dambo cultivated by individual families and their socioeconomic well-being.
The legislative restrictions on dambo cultivation have retarded implementation of organisational reforms and have contributed to the deterioration of man-land relationships in these areas (Whitlow, 1979). Whitlow's (1985) view is that there is a need to reassess the role of dambos in peasant farming particularly since overgrazing and trampling by livestock have probably caused more serious erosion than cultivation. I also find it important that research into improvement of dambo management, pasture development and utilisation has to be carried out with the aim of improving and sustaining livestock condition through the dry part of the year, and improvement of quality of forage at that time.
Studies have been initiated aimed at quantifying grazing management effects on dambo resources (Tables 2 and 3).
Table 2. Seasonal changes in dry matter (DM) yields of vlei sites compared to adjacent top land in Zimbabwe (1987/88). Preliminary Data.
|
Grazing schemes: Kowoyo A. Province: Mashonaland East (1987/88) | |||||
|
Rainfall: 800 mm+/annum. |
kgDM/ha |
|
|
| |
|
Pad. |
Sampling site locations |
Dec '87 |
Feb '88 |
May '88 |
Change in DM yield Dec-May |
|
A |
Vlei margin |
550 G |
466 R |
1099 R |
549 |
|
B |
Vlei stream bank |
834 G |
364 G |
742 G |
-92 |
|
C |
Vlei stream bank |
158 G |
440 G |
749 G |
591 |
|
D |
Vlei topland |
229 G |
229 G |
631 G |
401 |
G = Grazed R= Rested at time of sampling
Source: Mupangwa and Nyathi, 1988.
Observations:- There was a decline in DM yield in Pad A (15%) and B (56%) from early to mid-season due to stock concentration in these two paddocks during that period (records of stock movement). The vlei stream bank sites had higher dry matter yields than topland sites at the end of season (May) despite having been continuously grazed throughout (December 1987 to May 1988), attributed to higher moisture levels in the former than the latter. The vlei margins (Pad A) had a partial rest from mid (Fete) to end of season (May) resulting in increased DM production, almost twice as much as the vlei stream bank, and the wooded topland site. Continuous grazing of the vlei stream bank (Pad B) throughout the summer season seems to have negative effects on grass growth and herbage production (92 kg/ha by May).
Table 3. Seasonal changes in dry matter (DM) yields of vlei sites compared to adjacent top land in Zimbabwe (1987/88). Preliminary data.
|
Grazing yields: Chikowore. Province: Mashonaland Central | |||||
|
Rainfall: 800mm+/annum |
|
|
|
| |
|
Pad |
Number/location |
Dec '88 |
Feb '88 |
May '88 |
Change in DM yield Dec-May |
|
1 |
Topland-Mt slope thick bush |
196 G |
577 R |
668 |
472 |
|
2 |
Previously cultivated Mt slope |
142 G |
613 R |
823 |
681 |
|
3 |
Vlei stream bank |
137 G |
615 R |
941 |
804 |
|
4 |
Mountain foot sparse tree distribution |
177 R |
995 G |
338, |
161 |
G = Grazed; R = Rested at sampling time
Source: Nyathi, 1987; 1988
Observation: In all sites change in DM yield is greater between Dec-Feb than Feb-May due to vigorous grass growth and higher rainfall in the early summer season period than the late season. The vlei stream bank site seems to produce higher DM yields than the previously cultivated Mt. slope and topland when all three sites had mid-season rest while the mountain foot area was grazed.
Dambo utilisation is seemingly hampered by waterlogged conditions which prevent normal microbial activity' root development' and organic matter degradation. This should not be so prohibitive since indications from the literature alluded to in the text are that dambos have gradations of moisture resulting the vegetation zonation according to decreasing moisture levels away from the eye of the dambo (wettest spot). Furthermore, the pattern of organic matter accumulation and degradation follows the same trend with higher levels of undecomposed organic matter in the centre of the dambo (eye) than in the dambo margin (drier part). Soil depth increases towards the dambo centre. At the dambo margin, however, conditions appear to be ideal for crop production and the peasant farmer has evolved ridges and furrows over the years for production of rice and other crops.
An area of 1.2 million hectares occupied by dambos in Zimbabwe is indeed large enough to warrant investigations into its proper management, utilisation and development for sustained livestock production. Preliminary results of comparisons of dambo herbage production with topland sites seem to show that dambos are consistent sources of forage. With the overwhelming evidence that dambos are used for livestock grazing both by the large-scale commercial farmer and the communal small-scale farmer, it becomes imperative for agricultural scientists to start development-oriented agricultural research programmes that will help to reveal the productive potential of dambos.
Bishop, J.W.S. 1975. The environment of the Rhodesian people. Agriculture 4: M.O. Collins (Pvt) ltd., Salisbury. pp. 90-117.
Bond, G. 1963. Pleistocene environments in Southern Africa In: F.C. Howell and F. Bourlieve teds), African ecology and human evolution. Aldine Press, Chicago. pp. 308-334.
Bond, G. 1967. River valley morphology, stratigraphy and paleoclimatology in Southern Africa. In: W.W. Bishop and J.D. Clark (eds), Background to evolution in Africa. University of Chicago Press, Chicago. pp. 303-312.
Brammer, H. 1973. Soils of Zambia. Soil Surv. Rep. II, Ministry of Rural Development, Lusaka.
Brinkman, R. 1969. Ferrolysis, a hydromorphic soil forming process. Geoderma 3:199-206.
Butzer, K.W. 1976. Geomorphology from the earth. Harper and Row, New York.
Cormack, J.M. 1972. Efficient utilisation of water through land management. Rho. Agric. J. 9(1):11-16.
Fanshawe, D.B. 1969. The vegetation of Zambia. For Res. Bull. 7. Ministry of Rural Development, Lusaka.
Goudie, A.S. 1983. Environmental change. Clarendon Press, Oxford.
Henkel, J.S. 1931. Types of vegetation in Southern Rhodesia. S. Afr. J. Sci. 20:1-24.
Ivy, P. 1981. A guide to soil coding and land capability classification for land use planners. Dept. of Agricultural Technical and Extension Services, Harare, Zimbabwe.
Jennings, A.C. 1923. Erosion, especially surface washing in Southern Rhodesia. S. Afr. J. Sci. 20:204-207.
Mackel, R. 1974. Dambos: a study in morphodynamic activity on the plateau regions of Zambia. Catena 1:327-365.
Mazambani, D. 1982. Peri-urban cultivation within Greater Harare. Zimbabwe Science News 16(6):134-138.
Mupanga, J.F. and Nyathi, P. 1988. End of season (1987/88) veld trend monitoring report. Department of Agricultural Technical and Extension services, Harare, Zimbabwe.
Nyathi, P. 1987. End of season (May-1987) veld trend monitoring report on 7 grazing schemes. Department of Agricultural Technical and Extension Services, Harare, Zimbabwe.
Nyathi, P. 1988. Mid-season veld trend monitoring report on seven grazings scheme in 7 provinces. Department of Agricultural Technical and Extension Services, Harare, Zimbabwe.
Purves, W.D. 1976. A detailed investigation into the genesis of granite derived soils. Ph.D. thesis, Univ. of Rhodesia.
Rattray, J.M. 1957. The grass and grass associations of Southern Rhodesia. Rhod. Agric. J. 54(3):197-234.
Rattray, J.M., Cormack, R.M.M. and Staples, R.R. 1953. The vlei areas of S. Rhodesia and their uses. Rhod. Agric. J. 50:465-483.
Russell, T.P.J. 1971. Dambo utilisation survey. Lilongwe Land Development Programme, Lilongwe.
Scoones, I. 1987. Economic and ecological carrying capacity. Implications for livestock development in the dryland communal areas of Zimbabwe. Paper presented at a Seminar, Dept. of Biological Science, University of Zimbabwe, 1987, Harare.
Thiessen, R.J. 1975. Development in rural communities. Zambezia 4(2):93-98.
Thompson, J.G. 1972. What is a Vlei? Rhod. Agric. J. Tech. Bull. 15:153-154.
Walker, R.O. 1966. The dambo system of vegetable growing in the Western Province. Farming in Zambia (July) pp. 22-23.
Waston, J.P. 1964. A soil catena on granite in Southern Rhodesia. 1 - field observations II - analytical data J. Soil Sci. 15:238-257.
Werger, M.J.A. and Coetzee, B.J. 1978. The Sudano - Zambezian Region. In: M.J.A. Werger (ed.) Biogeography and ecology of southern Africa - Dr. W. Junk, the Hague. pp. 301-462.
Whitlow, J.R. 1979. Bornhardt terrain on granitic rocks in Zimbabwe: a preliminary assessment. Zam. Geog. J. 34:75-94.
Whitlow, J.R. 1980. The morphology of two different vleis on the highveld of Zimbabwe-Rhodesia. Zim-Rhodesia Agric. J. (2):71-80.
Whitlow, J.R. 1983. Vlei cultivation in Zimbabwe reflections on the past. Zim. Agric. J. 80(3):123-135.
Whitlow, J.R. 1984a. A survey of dambos in Zimbabwe. Zim. Agric. J. 81(4):129-138.
Whitlow, J.R. 1984b. Some morphological characteristics of dambo features in Zimbabwe. Trans. Zim. Sci. Assoc. 62(1):1-15.
Whitlow, J.R. 1985. Dambos in Zimbabwe (a review): Z. Geomorph. N.F. suppl. Bd. 52. Berlin-Stuttgart. pp. 115-146.
The workshop was attended by nearly 90 delegates from twenty-two countries of East, North, Southern, Central and West Africa with 52 papers presented in four sessions. The content of these sessions was somewhat varied ranging from specifically forage to specifically agricultural by-products but the majority of papers addressed both forages and by-products together. This makes it difficult to draw a line of distinction between the people and subject matter with respect to their affiliation with either PANESA or ARNAB networks.
In relation to by-products, there appeared from this meeting that there was a move away from such treatments as alkali/ammonification as being very unrealistic from a smallholder farmer's point of view to conserving to reduce trampling spoilage and earlier harvesting of residues to reduce spoilage by exposure. There is evidence of a growing realisation that forage materials, particularly legumes, have a very important supplementary role to play.
It is apparent from the number of presentations and from the field trip to visit a smallholder rural dairy project outside Lilongwe city, that grass forages play a very significant role in meeting the nutritional demands of upgraded livestock for milk production. Napier grass (Pennisetum purpureum) was perhaps the most important resource in this respect though Rhodes grass (Chloris gayana) was also widely used on fallow lands. The questions of long-term maintenance of soil fertility was not squarely addressed. However, there is the danger of running into competition for the fertility inputs which come largely from manure, between the demands of crops and forages. In many cases this was not being considered, except in one paper on the transfer of technology related to Rhizobium and legume-nitrogen economy, to the likely detriment of productivity and longevity of the fodder resource and possibly also crop production.
While grass forages clearly have a role to play in their capacity to produce bulk, which is often in short supply during the wet season because of the intensity of land use for crop production, there is the need for an overall strategy of quality forage production using legumes to help maintain soil fertility and during the dry season to supplement the quality of the otherwise low-quality crop residues and agro-industrial by-products.
Then there was the question of lack of or minimal adoption of research results due to socio-economic and poor research-extension-farmer linkage.
The resolutions derived from the four working group sessions on the last day of the workshop were:
1. Despite the long history of research in both pasture and agricultural by-products there has been little impact to the small-scale farming systems. There are two possible reasons for this: some of this research might have been inappropriate; where research has been appropriate the resultant technology had not been transferred to the farmers, mainly due to weak extension service.
2. There is a need for continued monitoring and evaluation of the relevancy and effectiveness of research programmes if we are to deliver the goods to the farmers.
3. The extension services need to be strengthened if progress is to be made in farmer adoption of research results. This may be done through inter alia, improved linkages between researchers and extension worker (through joint research), involvement of networks in training extension staff and improved university training in extension.
4. Land tenure systems in most countries militate against improved pasture and animal management. There is, therefore, a need to involve and advise the policy makers. The workshop also highlighted the need for studies on ecosystems, grazing studies and degradation of the environment, and integration of soil conservation and forage production if sustainable land-use systems are to be designed.
5. The availability of pasture seed was identified as a constraint. This problem could be alleviated by importation of seed in the short term, but ideally efforts should be made to produce seed locally and at affordable prices. It will be necessary to train farmers in seed production.
6. The important role of networks in co-ordinating research programmes within and between our countries was acknowledged. This may be facilitated by increased publication of Network Newsletters collaborative research and manpower development programmes.
7. There is a need for the authorities to provide conducive pricing policies and credit facilities in form of inputs in order to facilitate adoption of improved technologies by farmers.
8. It is imperative that the nutritive characteristics of available crop residues, by-products and pasture forages are measured and information placed in data banks for use by all interested parties. An inventory of the products and their accessibility is necessary. The need for standardization of terminology in describing food resources was also highlighted.
9. There is a need to develop appropriate experimental methodologies for on-farm research so that both the scientific merit and application of research results can be accommodated.
10. Economic appraisal should form an integral part of all research programmes.
11. Studies on feed resource budgeting and allocation should be encouraged in order to facilitate design of correct feeding strategies.
12. Reviews on topical subjects should be commissioned by the Steering Committee of the two networks in order to improve access to information which may be scattered in numerous publications
13. Commercial feed manufacturers should be encouraged to produce stockfeeds from crop residues and agricultural by-products at prices which are affordable to the small-scale farmers. It may, however, be necessary to first investigate the suitability of machinery.
14. There is a need to training in oral presentation of research results by scientists, for instance, this may be offered as part of training programmes put by Networks or ILCA.
15. Recommendations from previous workshops should be referred to as most remain topical.
