African Crop Science Journal African Crop Science Society ISSN: 1021-9730 EISSN: 2072-6589 Vol. 4, Num. 3, 1996, pp. 365-372

SHORT COMMUNICATION: on-farm integration of forage legumes into cereal cropping systems in the Ethiopian highlands

N. Z. LUPWAYI*, ABATE TEDLA and HAILU REGASSA^1

International Livestock Research Institute (ILRI), P.O. Box 5689, Addis Ababa, Ethiopia
^1 Institute of Agricultural Research (IAR), P.O. Box 2003, Addis Ababa, Ethiopia

(Received 8 January, 1996; accepted 26 September, 1996)

Code Number: CS96077
Sizes of Files:
    Text: 19.2K
    Graphics: Line drawings (gif) - 18.6K 
              Tables (jpg) - 250K  

ABSTRACT

An on-farm trial, in which a wheat-clover mixed cropping treatment was compared with un-fertilised (control) and N-fertilised (23 kg N ha^-1) pure wheat treatments, was conducted at Ginchi (2200 metres above sea level (m asl)) and Deneba (2600 m asl) in the Ethiopian highlands. Compared with the pure wheat control, inclusion of clover in the wheat at Deneba produced significantly more fodder dry matter (DM) (5.20 vs 3.16 t ha^-1), with higher N yield (36.35 vs 12.68 kg N ha^-1) and in vitro DM digestibility (51.7 vs 45.2 %), without reducing wheat grain yield significantly (1.93 vs 1.98 t ha^-1). At Ginchi, even more fodder (9.58 vs 3.12 t ha^-1) with higher N content (117.05 vs 15.03 kg N ha^-1) and DM digestibility (48.6 vs 34.9 %) was produced under mixed cropping, but this occurred at the expense of wheat grain yield (0.91 vs 1.39 t ha^-1). The N-fertilised wheat produced the highest grain yield at each site. Clover DM, nodule number, nodule size, nodule activity (internal colour) and nitrogen yield were all higher at Ginchi than at Deneba. In the second season, no residual effect of clover on wheat yield was detected. However, on the basis of the increase in feed quantity and quality, with little effect on wheat grain yield in the first season, it can be concluded that integration of wheat and clover production has potential to increase fodder availability for dry season feeding of livestock in the Ethiopian highlands.

Key Words: Clover, mixed cropping, on-farm, vertisols, wheat

RESUME

Un essai la ferme, dans lequel les trefles du ble melangees au traitement de la culture etait compae avec un controe non fertilise et N-fertilise ( 23 kg N par hectare) des traitements du ble de qualite pure, ete mene a Ginchi ( 2200 m au dessus asl) dans les regions montagneuses ethiopiennes. Comparativement au conte du meilleur ble, l'inclusion des trefles dans le ble abondamment produit Deneba a donne plus de matieres seches du fourrage (MSF) (5.20 contre 3.16 tonnes par hectare), avec un abondant rendement N (36.35 contre 12.68 kg N par hectare) et une digestibilite DM in vitro (51.7 contre 45.2%), sans reduction drendement abondant des grains du ble (1.39 contre 1.98 tonnes par hectare). A Ginchi, il y a meme plus de fourrages (9.58 contre 3.12 tonnes par hectare) avec le plus haut contenu de N (117.05 contre 15.03 kg par hectare) et la digestibilite DM (48.6 contre 34.9% etait produit sous une culture mixte, mais cela eu lieu aux depens du rendement des graines du ble N fertilise a produit le plus grand rendement de grains dans chaque centre. La trfle DM, le nombre de nodule, la dimension du nodule, l'activivite du nodule (couleur interne) et le rendement du nitrogne etaient tous plus eleves Ginchi que Deneba. Au cours de la seconde saison, let non residuel de la trefle sur le rendement du ble etait detecte. Cependant, c'etait sur la base de l'augmentation de la quantite et la qualite, avec moins det sur le rendement de grain du ble pendant la premire saison, on peut conclure que la production du ble a le potentiel dmenter la disponibilite du fourrage pendant la saison seche pour nourrir le betail en regions montagneuses ethiopiennes.

Mots Cles: Trefle, culture mixte, sur terrain, vertisols, du ble

INTRODUCTION

Throughout sub-Saharan Africa, one of the major constraints to livestock production is the limited amount and poor quality of feed available during the dry season. This affects not only meat and milk production, but also draught output as animals are often in poor condition at the end of the dry season when they are required for ploughing (ILCA, 1993).

One way of increasing the quantity and quality of dry-season feed for livestock is to include a forage legume in the crop mixture. This would be particularly attractive to farmers if cereal grain yields were not adversely affected by inclusion of the legume. Integration of forage legumes into cereal cropping systems is widely practised in Australia (Puckridge and French, 1983), where the residual effects of the legumes reduce the fertilizer requirements of the cereals grown subsequently. In the Ethiopian highlands, work conducted on research stations has shown that intercropping forage legumes with cereals is possible. Kahurananga (1991) found that in two seasons, intercropping clover with wheat had no significant effect on wheat grain yield. Abate et al. (1992) also found that the presence of clover in wheat stands did not affect wheat grain yields significantly across locations and seasons. Similar results were obtained when maize was undersown with Desmodium uncinatum and Chloris gayana over a five-year period (Alemu et al., 1987). The objective of this work was to test this technology on-farm and assess the role of legumes in improving feed quantity and quality for livestock production and supplying N to the soil for use by subsequent crops.

MATERIALS AND METHODS

The work was conducted in two seasons. The effect of intercropping wheat with clover was investigated in the first season, and the residual effect of clover on subsequent wheat was studied in the second season. Intercropping wheat with clover The study was conducted on drained Vertisols in farmerselds at two sites: Ginchi (9 degrees 0' N, 38 degrees 15' E) at medium altitude (2200 m asl) and Deneba (9o50' N, 39 degrees 20 at high altitude (2600 m asl). The soils were drained by preparing the seed beds with a broadbed maker (BBM), a modified traditional plough which makes broadbeds 80 cm wide, separated by furrows 40 cm wide. The furrows serve as ditches to drain excess water from the beds. The soil properties and meteorological data for the sites are presented in Table 1 and Figure 1, respectively. Temperature at Deneba was not recorded. Instead, 1993 data for Debre Birhan station is reported in Figure 1.

Figure 1

Table 1

Wheat (Triticum aestivum, cultivar ET-13).
Wheat + clovers (Trifolium tembense and Trifolium
quartinianum).
* Wheat + N fertilizer (23 kg N ha^-1 as urea).

Each plot was 35 x 14.4 m in size, consisting of 12 broadbeds and furrows, each 35 m long. Sowing was done in mid-June, 1992. The wheat and mixture of clovers were broadcast on the same day at seed rates of 150 and 10 (i.e., 5 + 5) kg ha^-1, respectively. A basal dressing of diammonium phosphate (DAP, 20:18:0 N:P:K) at 100 kg ha^-1 was applied to all treatments at planting time. The urea fertilizer in treatment 3 was applied four weeks later. All the cultural practices were carried out by the farmers. Intercropped plants were sampled at 50% flowering stage of clover growth. A 25 x 25 cm quadrant was used for sampling. The quadrant was randomly placed on three spots in each wheat-clover plot. At each spot, six plants were randomly selected for nodulation assessment in the field, following CIAT (1988) guidelines for forage legumes. The plants from all the three spots per plot were bulked and taken to the laboratory, where they were separated into roots and shoots. They were dried at 70 C for 48 hr and weighed to determine DM.

The plants were harvested in November at Ginchi and in December at Deneba. In each plot, the plants were harvested from six 1 x 1 m quadrants selected at random. Wheat grain was harvested first. The stubble was then cut at ground level and, in the wheat-clover treatment, separated into wheat straw and clover. The grain and crop residues were weighed in the field and samples taken to the laboratory for DM determination. The nitrogen content and in vitro DM digestibility of the crop residue separates and mixture were determined (Mebrahtu and Tenaye, 1995).

For comparison of results from all the three treatments at each site, analysis of variance was carried out for a randomised complete block design with ten replicates (farmers). Nitrogen yield and in vitro DM digestibility data were analysed similarly. Since nodulation and clover DM data were obtained from one treatment only per farmer, T-tests were used to compare the data from the two sites. In all cases, MSTAT-C computer software (Michigan State University, 1988) was used for the analyses.

Residual effects of clover on wheat. In the 1993 growing season, the residual effects of clover and fertilizer N were assessed by growing wheat on the first-season plots. This was done on seven farms at each of the two sites. The three first-season (cropping system) treatments were split for three rates of N: 0, 30 and 60 kg N ha^-1. The N source was urea and no other fertilizer was applied. Wheat cultivar ET-13 was sown in all the plots in June at the same rate as for the first season. It was harvested in November and December at Ginchi and Deneba, respectively. The grain and straw yield data were analysed as split-plot design, with the preceding cropping system as main plots and N rates as sub-plots.

RESULTS AND DISCUSSION

At mid-flowering stage, clover plants at Ginchi had significantly more abundant, bigger and more effective nodules than those at Deneba (Table 2). However, the distribution of nodules on the roots was similar at both sites. Table 2 also shows that the shoot and root DM of the clover plants at Ginchi were more than double those at Deneba.

Table 2

Figure 2

Figure 3

One reason for the difference in clover DM and N yields between the two sites may be the higher fertility level of the soils at Ginchi, particularly with regard to P and, to a less extent, K contents. Table 1 shows that the available P levels at Ginchi and Deneba were 31.07 and 4.57 g g-1 soil, respectively. The high response of clovers and other forage legumes to P in the Ethiopian highlands has been demonstrated before (Kahurananga and Tsehay, 1984; Nnadi et al., 1993; Haque et al., 1996). The high vegetative growth of clover at Ginchi apparently reduced wheat grain yield (Fig. 2a). This may have occurred due to competition for water and nutrients between wheat and clover during wheat maturation period. The higher temperatures at Ginchi may have been more favourable to clover growth than the cooler weather at Deneba (Fig. 1). In contrast, wheat is best suited to cool temperatures (14-16 C mean daily temperatures) during early growth (Choudhri, 1987) and the higher temperatures at Ginchi may have been less favourable for wheat grain formation.

In the second season, each increment of N rate significantly (P < 0.10) increased wheat grain and straw yields at Deneba (Tables 3 and 4). However, the preceding cropping system had no significant effect (P > 0.10) on wheat grain yield (Table 3), and the mixed wheat-clover treatment resulted in a significantly lower wheat straw yield than the other two treatments (Table 4). At Ginchi, fertilizer N at 30 kg N ha^-1 significantly increased wheat grain and straw yields over the control, but 60 kg N ha^-1 did not increase yield further (Tables 5 and 6). Preceding cropping systems had no significant effect on wheat grain and straw yields at this site either (Tables 5 and 6), implying that there were no residual effects of clover or fertilizer N.

Table 3

Table 4

Table 5

Table 6

Wheat-clover mixed cropping may seem to be better suited to Deneba than Ginchi since, compared to wheat only, more fodder (Fig. 2b) with higher N (or protein) content and digestibility (Figs. 3a and 3b) was produced without reducing wheat grain yield at Deneba (Fig. 2a). However, the fodder produced at Ginchi under mixed cropping was much higher both in quantity (Fig. 2b) and in N content (Fig. 3a) than that produced at Deneba. Therefore, the loss in wheat grain yield could be compensated for by an increase in livestock output from the fodder. This means that farmers' priorities are important, i.e., whether the farmer' priority is on wheat production or livestock production. These on-farm results confirmed on-station results (cited in the introduction), indicating that forage and cereal production can be integrated in the Ethiopian highlands, and possibly in the whole East African highland region, with little effect on cereal yields. Intercropping will increase fodder quantity and quality for dry season feeding by livestock.

ACKNOWLEDGEMENTS

We thank Ato Girma Alemayhu, Ato Degefa Beru and Ato Daniel Habtewold for their assistance in the implementation of the experiments. We are also grateful to Ato Mebrahtu Ogbai for plant N and in vitro digestibility analyses.

REFERENCES

Abate, T., Tekalign, M. and Getinet, G. 1992. Integration of forage legumes into cereal cropping systems in Vertisols of the Ethiopian highlands. Tropical Agriculture 69:68-72.

Alemu, T., Taylor, M.S. and Tadesse, T. 1987. Intercropping of maize with forages. Ethiopian Journal of Agricultural Science 9:15-24.

CIAT (Centro Internacional de Agricultura Tropical). 1988. Legume-Rhizobium Symbiosis: Methods Manual for Evaluation, Selection and Agronomic Management.

CIAT, Cali. Choudhri, M.B. 1987. Wheat Production Potential in Africa. FAO, Rome.

Haque, I. 1992. Use of legume biological nitrogen fixation in crop/livestock production systems. In: Biological Nitrogen Fixation and Sustainability of Tropical Agriculture. Mulongoy, K., Gueye, M. and Spencer, D.S.C. (Eds.), pp. 423-438. John Wiley and Sons, Chichester.

Haque, I., Lupwayi, N.Z. and Luyindula, N. 1996. Inoculation and phosphorus effects on Desmodium intortum and Sesbania sesban in the Ethiopian highlands. Agriculture, Ecosystems and Environment 56:165-172.

ILCA (International Livestock Centre for Africa). 1993. Sustainable Production from Livestock: ILCAedium-Term Plan, 1994-1998. ILCA, Addis Ababa.

Kahurananga, J. 1991. Intercropping Ethiopian Trifolium species with wheat. Experimental Agriculture 27:385-390.

Kahurananga, J. and Tsehay, A. 1984. Preliminary assessment of some annual Ethiopian Trifolium species for hay production. Tropical Grasslands 18:215-217.

Mebrahtu, O. and Tenaye, S. B. 1995. Analytical Methods for Feeds, Animal Excretions and Animal Tissues. ILCA, Addis Ababa.

Michigan State University. 1988. MSTAT-C, A Microcomputer Programme for the Design, Management and Analysis of Agronomic Research Experiments. Michigan State University, Michigan.

Nnadi, L.A., Haque, I. and Mugwira, L.M. 1993. Phosphorus response and mineral composition of Ethiopian highland Trifolium (clover) species. Communications in Soil Science and Plant Analysis 24:641-656.

Puckridge, D.W. and French, R.J. 1983. The annual legume pasture in cereal-ley farming systems in Southern Australia - a review. Agricultural Ecosystems and Environment 9:229-267.

Tanner, D.G., Amanuel, G. and Asefa, T. 1993. Fertilizer effects on sustainability in the wheat-based small-holder farming systems of southeastern Ethiopia. Field Crops Research 33:235-248.

Copyright 1996 The African Crop Science Society

The following images related to this document are available:

Photo images

Line drawing images