African Crop Science Journal African Crop Science Society ISSN: 1021-9730 EISSN: 2072-6589 Vol. 10, Num. 1, 2002, pp. 81-97

(Received 21 April, 2001; accepted 2 February, 2002)

Code Number: cs02008

ABSTRACT

The study investigated the effect of intercropping cereals (maize, sorghum and wheat) with forage legumes (lablab and clover), planting methods and manure application on cereal grain and fodder dry matter yield and fodder  nutritive value. Data collected over a period of three years, indicated that intercropping significantly (P<0.001) yielded more fodder (27%) but slightly depressed grain yields compared to sole cereal cropping. Row planting significantly (P<0.001) yielded more fodder DM (5%) and  more cereal grain (21%) than broadcast planting. The nutritive value (CP, NDF and dry matter degradability) of the fodder was not affected (P>0.05) by planting method. However, intercropping forage legumes with cereals generally resulted in fodder with higher fodder CP concentration, lower NDF and higher dry matter degradability than fodder from sole cereals. Manure application into cereal + forage legume cropping systems significantly (P<0.05) yielded higher grain and fodder DM (range: 2.9-3.6 t ha^-1 and 8.2-9.3 t ha^-1, respectively) than inorganic fertiliser (3.2 and 8.8 t ha^-1). These yields were significantly (P<0.001)  higher than the of no manure and no diammonium phosphate (DAP) urea fertiliser  control, which yielded 2.3 and 7.1 t ha^-1 of grain and fodder DM, respectively.  Fodder DM yield gains of manure over control in sole crops averaged 46.9% for maize, 41.4% for sorghum and 64.2% for wheat, while yield gains in the intercrop averaged 34.5% for maize + lablab, 39.1% for sorghum + lablab and 37.1% for wheat + clover. Cereal grain yield gains caused by manure over the control in sole crops averaged 58.3% for maize, 22.7% for sorghum and 28.6% for wheat, while yield gains for the intercrops averaged 70.4% for maize + lablab, 55.6% for sorghum + lablab and 14.3% for wheat + clover.  Cattle manure application  yielded as much or even more grain and fodder as chemical fertiliser (DAP/urea).  For optimum cereal grain yield and fodder production from cereal and forage legume cropping systems, cattle manure should be applied at a rate of 13.7 t DM ha^-1. However, the rate of 2.5 t DM manure  ha^-1 year ha^-1, which had a yield advantage of 15.5% over the control in sole stands, was recommended for tropical smallholder crop/livestock farming systems.

Key Words: Clover, fertiliser, lablab, maize, manure, urea

RÉSUMÉ

Les effets des cultures intercalées des céréales (mais, sorgho et blé) avec fourrages (lablab et trèfle), méthodes de planter et l'application des fumiers sur les grains des céréales et les rendements en fourrages secs et leurs valeurs nutritives ont été étudié. Les données récoltées pendant trois ans ont montré que les rendements en fourrages ont significativement augmenté (27%) mais légérement diminué les rendements en grains (21%) par rapport aux céréales en monoculture. La disposition en lignes a produit un rendement élévé en fourrages secs (5%) et en grains des céréales que les céréales plantés aux hazard. La valeur nutritive  (CP, NDF et degradabilité de la matière sèches) des fourrages n'a pas été affectée (P>0.05) par la méthode de planter. Cependant, la culture des fourrages et  les  céréales intercalés a généré des fourrages a avec concentrations en CP et une degradabilité de la matière sèches élévées et de faibles NDF. L'application des fumiers dans les céréales + les systèmes culturales des fourrages a augmentée  significativement  les  rendemenrs en grains et  matière sèches  des  fourrages  (2,9-3,6 t ha ^-1 et 8,2 - 9,3 t ha^-1 ; respectivement) que les fertilisants inorganiques (3,2 et 8,8 t ha^-1). Ces rendements  étaient significativement (P<0.001) élévés  que ceux des non  fumiers, non  DAP et  l'urée qui ont produit 2,3  et 7,1 t ha ^-1 des grains et fourrages secs; respectivement. Les gains en fourrages secs générés par les fumiers par rapport au contrôle des céréales en monocultures avoisinés 46,9% pour les maïs, 41, 4% pour le sorgho et 64,2 % pour le blé, tandis que les gains en cultures intercalées avoisinés 34,5 % pour les maïs+lablab, 39,1% pour le sorgho et 37,1% pour le blé + trèfle. Les rendements en grains des céréales causés par l'application du fumier sur le contrôle (plantes en monocultures) avoisinés 58,3% pour les maïs, 22,7% pour le sorgho et 28,6% pour le blé, tandis qu'ils avoisinés 70,4% pour les maïs+lablab, 55,6% pour le sorgho + lablab et 14,3 % pour le blé + trèfle. L'application du fumier a produit au moins la même quantité de la matière végétale et des grains que fertilisants chimiques (DAP/urée). Pour une production optimale, des grains et la matière vegetales des céréales et des fourrages les fumiers devra être appliqué à un taux de 13,7 t ha ^-1) des matières sèches. Cependant, le taux de 2,5 t ha ^-1) qui avait produit un gain de 15,5% sur le contrôle (monoculture) était le plus recommendé aux petits fermiers et agriculteurs des régions tropicales.

INTRODUCTION

Ruminant production in the tropical and subtropical regions of the world is largely dependent on natural pastures and crop residues. Seasonal shortages and low nutritional value of these feed resources constitute the most wide-spread technical constraints to livestock production in smallholder crop-livestock systems.  Growing of forage legumes in association with food crops to improve the feeding value of crop residues is one option that has shown promise for low-resource farmers to feed their animals better (Abate et al., 1992; 1994; Umunna et al., 1997) while contributing to soil fertility.  However, farmers are likely to adopt the technology of growing forage legumes in association with food crops (cereals) only if this results in increased fodder production and does not depress grain yields (Abate et al., 1992).  

On the other hand, a major problem of cropping systems in the tropics is the reduction in soil productivity that accompanies most systems of continuous cultivation (Haque and Jutzi, 1984). While intercropping usually includes a legume which fixes nitrogen, applied nitrogen may still confer some benefits to the system, because the cereal component depends heavily on nitrogen for maximum yield (Ofori and Stern, 1986; Moreira, 1989; Cochran and Schlentner, 1995).  Under smallholder systems, maintenance of soil fertility is poor.  Fertiliser prices have tripled in recent years, and coupled with unavailability, their use is limited (Brumby, 1991).  Therefore, there is need for alternative soil ameliorating agents for sustainable crop and livestock production.  Animal manure has become an important factor in maintaining land productivity, particularly in areas where the cultivation density is high (Powell, 1986).  However, data on the influence of applied animal manure in cereal and forage legume intercropping systems are presently scarce.

This study was, therefore, undertaken to evaluate the effects of different methods of intercropping forage legumes like lablab (Lablab purpureus Dolichos lablab) and clover  (Trifolium steudnerii) with cereal crops like maize (Zea mays), sorghum (Sorghum bicolor) and wheat (Triticum aestivum L.) and source of nutrients (cattle manure or commercial fertiliser) on fodder production and cereal grain yield.

MATERIALS AND METHODS

Study area. The experiments were carried out at the International Livestock Research Institute (ILRI) Debre Zeit Research Station in Ethiopia for three consecutive years starting from 1995.  The station is located 50 km southeast of Addis Ababa in the Ethiopian highlands (8⁰ 44' N and 38⁰ 58' E) at an altitude 1850 m above sea level.  The rainfall pattern is bimodal with the short rainy season occurring between February and April and the long one  between mid-June and mid-September (Astatke et al., 1995).  The mean annual rainfall is 865 mm and mean minimum and maximum temperatures are 10.6⁰ C and 25.0⁰ C, respectively.  The short rains are highly erratic and very little cropping occurs in this season.  During the study period, about 83, 76 and 85% of the total annual rainfall were received in 1995/96, 1996/97 and 1997/98 growing seasons, respectively.  The major soil type of the area is a chromic vertisol covering more than 90% of the area and a small proportion of alfisol [Nitosol, FAO-UNESCO system (Berhanu, 1985)].  Field studies were carried out on both soil types for three cropping seasons from 1995 to 1998 and the results presented in this paper are from the two soil types.

Experimental design and cultural practices. The experimental design was a split plot with four replicates: manure/chemical fertiliser comprised the main plots, while cropping systems (sole crops and intercrops of cereals and forage legumes planted by row and broadcast methods) were the sub-plots.  The main plot treatments  comprised of four graded levels of cattle manure (M1=2.5, M2=7.5, M3=12.5 and M4=17.5 t dry matter (DM) ha^-1, one level (FC) of chemical fertiliser [diammonium phosphate (DAP)/Urea] and a control of no fertilisation (M0).  The subplots each measured (3 m x 7 m) and comprised of sole cereal crops (maize, sorghum, wheat); 6 plots of cereal + forage legume intercropped mixtures (maize + lablab (ML), sorghum + lablab (SL) and wheat + clover (WC)), each planted by row and broadcast methods and 2 plots of sole forage legumes (lablab and clover), each row-planted.

The crop varieties used were maize hybrid BH 660, sorghum variety Gambela 1107, wheat  variety K-6785-4A, lablab variety Rongai and clover. The chemical fertiliser was applied at the recommended rate for Ethiopian highlands (18 kg ha^-1 N and 19.4 kg ha^-1 P₂O₅ applied  at  planting  and 23  kg ha^-1 of N top-dressed, six weeks after planting). The manure used was decomposed cattle manure collected from the station cattle barns and its physical properties and chemical composition are presented in Table 1. Manure was analysed at every planting season for total nitrogen by the Kjeltec System, available phosphorous by Bray II method, organic carbon by the Walkley and Black method and exchangeable bases by ammonium acetate method (Tekalign et al., 1991). The chemical fertiliser DAP contained 18% N and 19.4% P as P₂O₅ and urea contained 46%N.

The seedbed was disc ploughed and harrowed using a  a tractor.   Each crop or crop combination was grown successively in the same plot for the three consecutive years of the study. All the crops in the sole and intercrop stands were planted on the same day for the three cropping seasons.  Uncultivated gaps of 3 m and 2 m separated the main plots and subplots, respectively, while drainage trenches were dug around the main plots to avoid nutrient spill over effect from one main plot to another by runoff water. Guard rows of sole maize were planted around the experimental plots.

Planting methods, spacing and seed rates. Maize and lablab seeds in the sole and intercrop treatments were planted at two seeds per hill and thinned to one per hill two weeks after emergence.  In sole stands, maize and lablab rows were 0.75 m apart,  while, the within-row plant spacing was 0.25 m.  In the intercrops, single rows of lablab were planted in between the spaces of maize rows planted at a row spacing similar to the sole treatments. For the broadcast intercrop treatment, maize and lablab seeds were mixed at a ratio of 3:2 by weight and planted at a population density of 10% above that used in their respective sole stands.  Two weeks after emergence, maize and lablab, in both the sole and intercrop broadcast stands, were thinned to a density approximately equal to that used in their respective sole stands.

Sorghum in both sole and intercrop treatments was planted at a seed rate of 6 kg ha^-1.  In sole row treatments, sorghum was planted in single rows (75 cm apart), while for the row intercrop treatments, lablab was interplanted in single rows into spaces between the sorghum rows at a row and within row spacing similar to that in the maize + lablab intercropping treatments.  Four weeks after emergence, sorghum was thinned to a population density of 87,000 plants (seed rate = 4 kg ha^-1) by maintaining a within row spacing of 0.25 m between sorghum plants in each row.  For the broadcast intercropping treatments, sorghum and lablab seeds were combined in a proportion of 2:5 by weight and applied at a rate of 25 kg ha^-1 of the mixture.  After emergency, sorghum and lablab were thinned to approximately same plant population as was in their respective sole row treatments. Thus, the final mixture of sorghum and lablab broadcast intercropping treatments was at a seed rate of 17 kg ha^-1.

In wheat and clover cropping systems, the seeds were planted at a rate of 125 and 14 kg ha^-1 for wheat and clover respectively.  For the sole cropping, wheat rows were 0.30 m apart, while in wheat + clover row intercropping treatments, clover was planted in single alternate rows in spaces between wheat rows at the same spacing like in the sole stands; thus, the distance between adjacent wheat and clover rows was 0.15 m.  For the intercropping treatments, wheat and clover seeds were combined in a proportion of 8.5:1 by weight and broadcasted at a rate of 150 kg ha^-1 of the mixture.  At first weeding, the crops in the intercrop treatments were thinned to approximately the same plant population as in their respective sole stands.

Harvesting, measurements and chemical analyses. Cereal grain cobs or heads were harvested at full maturity and shelled after sun drying. Cereal grain and fodder dry matter yields were determined on a whole plot basis from a sampling area of 12 m² in the middle of each sub- plot; leaving the outer plants (0.50 m) as guard rows on either side of the sub-plot. Cereal grain weight was taken after sun drying to about 14% moisture content. After harvesting the cereal cobs or heads for grain, the remaining above ground plant material (crop residue) was cut down using machetes and sickles, and weighed immediately in the field. Two representative samples were immediately taken from each harvested plot, chopped and dried in paper bags to determine fodder dry matter yield (DMY) and chemical composition.  The samples for DMY were oven dried at 100⁰ C to constant weight (about 48 hr) while the samples for chemical analysis were dried at 60⁰ C for 72 hr in an air-circulating electric oven.  For the intercrop plots, after the two samples were taken, all the harvested material from each sample area was separated into the individual cereal and forage legume components to determine the proportion of the legume in the mixture. A sample of each component was then taken and oven dried at 100⁰ C for 48 hr for DM determination.

After drying, samples for chemical analysis were each ground to pass through a 3-mm screen.  Each ground sample was thoroughly mixed and a sub-sample taken and ground again to pass a 1-mm screen. Each sample was stored in sealed plastic sample cups (50 g) for chemical analyses.  The remaining 3-mm ground samples were stored in plastic bags and used for nylon bag DM degradability studies.  Samples were analysed for DM, ash and total nitrogen (crude protein) by standard procedures of AOAC (1990).  Neutral detergent fibre (NDF) and acid detergent fibre (ADF) were determined by the procedures of Van Soest and Robertson (1985).

In sacco DM degradation. Fodder DM degradation was determined using the nylon bag technique (Orskov et al., 1980).  Air dried samples of each forage were ground to pass through a 3- mm screen and 3 g was weighed into 6 cm x 12 cm nylon bags of pore size 41 µm (Polymon, Switzerland).  The bags were incubated in duplicate for 24 and 48 hr in the rumen of three fistulated crossbred (Bos taurus x Bos indicus) steers (average liveweight 300 kg).  The steers were housed individually and had water and mineral salts ad lib.  The steers were fed ad lib, a basal diet of grass hay (88.5% DM, 90.3% OM, 6.2% CP and 66.2% NDF) supplemented with concentrate (1:1 mixture of cottonseed cake and wheat middlings) at 3 kg head^-1 day^-1. Duplicate bags of each forage sample were added sequentially into the rumen of each steer. After incubation, the bags were removed at the same time and immediately washed in a washing machine (Tefal Alternatic, Finland) for 30 minutes (5 cycles) and later rinsed under running cold tap water until the rinsed water was clear. The bags were then dried in an oven at 60-70º C for 48 hr and weighed after cooling in a desiccator. Dry matter disappearance was calculated as described by Osuji et al. (1993).

Statistical analysis. Statistical analysis was performed on grain yield and fodder (crop residue) DM yield, chemical components of the fodder (CP, NDF, OM and ash), and dry matter disappearance (DMD) from nylon bags after 24 and 48 hr of incubation.  The analysis was done using General Linear Model (GLM) procedures for split plot design  (SAS, 1989). Treatment sum of squares were partially partitioned into linear orthogonal contrasts: (1) sole vs intercropping; (2) row vs broadcast planting; (3) sole maize vs maize + lablab intercrop; (4) sole sorghum vs sorghum + sorghum + lablab intercrop; sole wheat + clover intercrop. From the main plot treatment sum of squares, two non-orthogonal contrasts were examined: M0 vs M1-M4 and FC vs M1-M4.  Manure treatments were also tested for both linear and quadratic relationships after excluding the chemical fertiliser (FC) and the control treatment (M0).  Regression analysis using RSREG procedure in  (SAS, 1989) was used to determine the optimum level of manure application in cereal + forage legume cropping systems.  

RESULTS

Fodder DM and cereal grain yields. Results for the effect of cropping system on fodder DMY  and cereal grain yields are presented in Tables 2 and 3.  Year x main plot (manure), site x main plot and year x site x main plot interactions were not significant. However, year x cropping system, main plot (fertiliser) x cropping system and year x fertiliser x cropping system interactions were highly significant (P<0.001). Overall means indicated that row planting significantly (P<0.001) yielded more fodder DM (9.1 vs 8.7 t ha^-1) and 21% more cereal grains (3.4 vs 2.8 t ha^-1) than broadcast planting.  Comparing the cropping systems, the results showed that cereal + forage legumes intercropping significantly (P<0.001) yielded  more (27%) fodder DM (10.5 vs 7.2 t ha^-1) but slightly depressed cereal grain yield by 8% (3.3 vs 3.6t ha^-1), compared to sole cereal cropping.

In maize and/or lablab cropping systems, there was no significant (>0.05) difference between row and broadcast planting, but in sole stands, maize planted in rows significantly (P<0.05) yielded more fodder DM than sole maize broadcast (Table 2).  Fodder DM yield of maize + lablab (ML) intercrop, was 1.5 times higher (53.2% increase) than in sole maize and the proportion of lablab in the mixtures was 39.8  and 41.3% for the row and broadcast ML intercrops, respectively.  In sole stands, maize planted in rows yielded more grain (1.2 times) than broadcast planting; while for the mixtures, maize + lablab intercrop row planted yielded 1.3 times (31.2% increase) more maize grain than ML broadcast intercrop.  However, regardless of planting method, intercropping of maize and lablab reduced maize grain yield by 20% (5296 vs 4210 kg ha^-1)  compared with grain yield from sole maize stands.

For sorghum and/or lablab cropping systems, there was no significant (P>0.05) difference in fodder DM yield between sole row and broadcast planting methods.  However, for the intercrops, sorghum + lablab (SL) row planting significantly (P<0.01) yielded more fodder DM than sorghum + lablab broadcast.  Sorghum grain yield in both sole and intercrop stands, was significantly (P<0.01) higher for the row than the broadcast planting method.  In sole stands, row-planted sorghum grain yield was 1.1 times higher than sorghum broadcast-planted (10.2% increase),  while for the intercrops, sorghum grain yield from SL row was 2.23 times (123% increase) higher than SL broadcast.  Fodder DM yield from SL intercropping was 1.6 times higher (63.1% increase) than sole sorghum, but the SL mixture contained more lablab than sorghum.

For wheat, planting in  row or broadcast planting did not significantly (P>0.05) affect fodder DM yield in both sole wheat and wheat + clover intercrop stands.  Although statistically not significant, the results for wheat also showed that row planting yielded more fodder than the broadcast method in both sole wheat and wheat + clover intercrops.  Fodder DM production in wheat + clover (WC) intercrop was 1.2 times (21% increase) higher than in sole wheat stand.  The percentage clover in the mixtures was 23 and 20.5 for WC row and WC broadcast, respectively.  The results for wheat grain yield in the wheat + clover intercrops, there was significant (P<0.01) wheat grain yield reduction of 7.8% compared to sole wheat.  However, planting in rows had no significant (>0.05) difference in wheat grain yield between wheat + clover row intercrop over sole wheat planted in rows. The results for wheat and/or clover cropping systems also showed that row planting minimised wheat grain yield reduction (6.2%) in the intercrops as was the case with maize or sorghum and lablab intercrops.

Effect of cropping systems on chemical composition of fodder. Results of the effect of cropping system on crude protein yield (CPY), chemical composition and DM degradability of the fodder from cereal + forage legume intercropping are presented in Table 3.  Within the sole crops and intercrops, CPY was similar for row and broadcast planting methods.  However, intercropping cereals with forage legumes led to higher fodder CPY than sole cereal stands.  Pooled means indicated that the CPY of the intercrops was 4.2 times  for maize + lablab, 3.9 times  for  sorghum + lablab and 2.4 times for wheat + clover mixtures higher than in their respective sole stands. This was attributed to the higher (P<0.001) crude protein concentration of the mixtures compared to forages from sole cereal stands (Table 3).

There were nutritive value treatment differences for  fodder  among cropping systems.  In all  cropping systems, planting method had no significant (P>0.05) effect on the concentration of crude protein (CP), neutral detergent fibre (NDF) and dry matter degradability (DMD) of the fodder. However, intercropping cereals with forage legumes significantly P<0.001) improved the chemical composition of the forages  compared with forages from sole cereal stands.  Crude protein content was 2.7 times in ML, 2.4 times  in SL and 2.2 times in WC mixtures higher than in the respective sole crops.  Considering sole stands  irrespective of planting method, fodder N content was 0.5% for sole maize, 0.7% for sole sorghum and 0.6% for sole wheat.  Among the intercropped forages, mean CP content was highest in sorghum + lablab (105 g kg^-1), followed by maize + lablab (85 g kg^-1) and was lowest in wheat + clover (75 g kg^-1) (Table 3).

Intercropping significantly (P<0.001) reduced the NDF content of the forages, with intercropped forages having lower fibre content than their respective sole stands.  There were also significant (P<0.01) differences in NDF content among the intercropped forages.  Sorghum/lablab forage had the lowest NDF content (523 g kg^-1), followed by maize/lablab (572 g kg^-1) and wheat/clover (634 g kg^-1).  Furthermore, the DM degradabilities of the intercropped forages were significantly (P<0.001) higher than the sole cropping treatments.  Among the intercropped forages, DMD followed the same trend as NDF, with sorghum/lablab having the highest DMD (674 g kg^-1) followed by maize/lablab (665 g kg^-1) and wheat/clover (523 g kg^-1).

Effect of manure and cropping systems on fodder dry matter and cereal grain yields. The results for manure effect on cereal and forage legume cropping systems on fodder dry matter yields are presented in Table 4. The general trend indicated that increasing levels of manure significantly (P<0.001) increased fodder DM yield from 7.1 t ha^-1  for the  control treatment to 9.8  t ha^-1 at 17. 5 t DM manure ha^-1. Mean fodder DM yield for the manure (9.03 t ha^-1) was significantly (P<0.05) higher than the mean for the commercial fertiliser (8.8 t ha^-1); and was significantly (P<0.001) higher than the fodder DM yield of 7.08  t ha^-1 from the control treatment (M0). The highest fodder DM yields in both sole and intercrop stands were obtained at the rate of 12.5 t DM manure ha^-1.  At the highest level of manure application (M4), fodder DM yield was significantly (P<0.05) reduced. Yield advantages over the control treatment were 28% for manure (M4) and 24% for DAP/urea fertiliser. The results also indicated that for both manure and chemical fertiliser, fodder DM yield response to fertilisation was greater in sole cereal treatments than in their respective intercrops with forage legumes. Fodder DM yield gains of manure over control in sole crops averaged 46.9% for maize, 41.4% for sorghum and 64.2% for wheat, while yield gains in the intercrop averaged 34.5% for maize + lablab, 39.1% for sorghum + lablab and 37.1% for wheat + clover.

For maize and/or lablab cropping systems, there was no significant (P>0.05) difference between manure and commercial fertiliser while in sorghum and/or lablab cropping systems, manuring significantly yielded more fodder in sole sorghum row (P<0.01) and sorghum + lablab broadcast intercrop (P<0.05) than with commercial fertiliser (Table 4). For wheat and/or clover cropping systems, manuring yielded significantly (P<0.05) higher fodder DM than commercial fertiliser in both wheat + clover row and wheat + clover broadcast treatments. To separate the beneficial effect of intercropping from the effect of manure or commercial fertiliser, fodder DM yield response advantages of intercrops over those of sole stands for the control treatment (M0) only were computed and compared.  This then gave the following yield advantages for intercropping over sole cereal cropping: 60% for maize + lablab, 56.8% for sorghum + lablab and 29.6% for wheat + clover intercrops.

The results for the effect of manure levels and cropping system on mean cereal grain yield are presented in Table 5.  Overall cropping system means indicated that increasing levels of manure significantly (P<0.001) increased cereal grain yield but there was no significant (P>0.05) difference in cereal grain yield between manuring levels of 12.5  and 17.5 t DM ha^-1. Maize grain yield advantages in response to manure application in maize and lablab cropping systems were 59.3% compared to the control (M0) and 4.1% compared to DAP/urea fertiliser (FC).  For sorghum and/or lablab cropping systems, sorghum grain yield response to manuring was  46.7% greater than M0 and 10% greater than FC, while for wheat and clover cropping systems, the yield advantage was 19% greater than M0.  When compared with DAP/urea, manuring reduced wheat grain yield by 8%.  Cereal grain yield gains of manure over the control (M0) in sole crops averaged 58.3% for maize, 22.7% for sorghum and 28.6% for wheat, while yield gains for intercrops averaged 70.4% for maize + lablab, 55.6% for sorghum + lablab and 14.3% for wheat + clover.

Regression analysis of the results for the overall means of fodder DM yield revealed a significant (P<0.001) quadratic relationship of fodder DMY with increasing levels of manure application (Fig. 1).  The  regression  equation  of the response curve for  fodder  DMY was  Y =  7162  +  360.6X  + -13.3X²; where Y = the predicted fodder DM yield (kg ha^-1) and X = the rate of manure applied (t DM ha^-1).  The maximum stationary point of the response curve was achieved at a fodder DMY of 9798 kg ha^-1 and this corresponded to manure application level of 13.6 t DM ha^-1.  For cereal grain yield, the regression response curve equation was Y = 2358 + 155x - 4.9 X²; where Y = the predicted cereal grain yield (kg ha^-1) and X = the rate of manure applied (t DM ha^-1).  The maximum stationary point of the regression curve was established at cereal grain yield of 3563 kg ha^-1 which corresponded to manure application level of 13.8 t DM ha^-1 (Fig. 2).  Using regression, the biological optimum level for manure application in cereal + forage legumes intercropping systems was achieved at  the  rate  of 13.7 t DM  manure ha^-1 year^-1 for both cereal grain yield and fodder production.

Using the optimum level of manure, yield advantages in response to manure for fodder DM production were 38.4% (9798 vs 7082 kg ha^-1) over the control (M0); and 11.4% (9798 vs 8800 kg ha^-1)  over  DAP/urea fertiliser  at  100/50  kg  ha^-1).  For cereal grain production, yield advantages at optimum level of manure application were 61.6% (3563 vs 2267 kg ha^-1) and 12% (3563 vs 3181 kg ha^-1) over M0 and FC treatments, respectively.

DISCUSSION

Effect of cropping system on fodder DM and cereal grain yields. The general trend in most intercropping experiments is that the grain and stover yields of a given crop in the mixture are less than the yields of the same crop grown alone, but the total productivity per unit of land is usually greater for mixtures than for sole crops (Willey, 1979; Natarajan and Willey, 1980; ILCA, 1989). This was true for cereal + forage legume intercropping results obtained in this study. Intercropping greatly increased total fodder biomass production compared to sole stands,  although there were varying reductions in cereal grain yield for the different cropping systems. The results also showed that row planting was superior to broadcasting for both fodder biomass and cereal grain production, irrespective of the cropping system. Planting in rows has the advantage that more efficient utilisation of light, water resources and soil nutrients can be attained by the spatial arrangements of the crops than in broadcast planting (Andrews and Kassam, 1976; Nyambo et al., 1982; Singh, 1981). In the present study, therefore, planting in rows probably minimised competition in the mixtures and, hence, the higher fodder and grain yields realised than under broadcasting.  Reductions in cereal grain yield of intercrops relative to sole stands in this study, especially with ML and SL intercrops, could be attributed to the reduction in light intensity that was imposed on the cereals by lablab, which reduces the rate of photosynthesis (Whyte et al., 1953).

The higher maize grain yield reductions in the broadcast than in the row planted mixtures of maize + lablab obtained was attributed to yield advantages for the row planted mixtures that accrued basically through a lower inter- and intra-crop competition for space, both aerial and edaphic (Kassam, 1979). Less reductions in wheat grain yield from the intercrops than in the maize or sorghum + lablab intercrops indicated that the competition for light and nutrients between clover and wheat in the mixtures was probably minimal and, therefore, did not affect grain yield as much as in ML and SL intercrops. This could be explained by the findings of Nnadi and Haque (1986) that small seeded forage legumes such as Trifolium and Medicago species, are likely not to compete with cereals if sown at the same time.

The increases in total fodder biomass production and high reductions in sorghum grain yield in the intercrops, compared to sole sorghum, was associated with the above ground competition for light between lablab and sorghum in the mixtures (Willey, 1979; Rees, 1986). This could probably be due to the rapid deep rooting system of the lablab (Nnadi and Haque, 1986; Skerman et al., 1988).  The competition was mainly attributed to the extremely higher growth vigour of lablab than sorghum, which quickly overshadowed the sorghum and suppressed its tiller formation at the establishment phase since both were planted at the same time.  As a result, sorghum plants in the mixtures were single shoots and had smaller heads than sorghum in sole stand which had on average 3 shoots with bigger heads. As such, the quantity of lablab compared to sorghum was very high in the mixtures.  This effect was more pronounced in the broadcast than in row planted mixtures and this was attributed to intercrop competition for space and light (Whyte et al., 1953).

Yield reductions for sorghum grain reported in this study were higher than the acceptable decrease of 10-15% relative to the cereal monocrop reported by Nnadi and Haque (1986). Therefore, it was noted that the technology of sorghum + lablab intercropping would unlikely find favour in subsistence agriculture in areas with similar environmental conditions (humid and subhumid tropics), where this study was conducted.  However, the high DM and quality SL fodder produced was sufficiently promising to suggest that slightly different sorghum + lablab combinations like adjusting the time of planting lablab into the sorghum could be advantageous and minimise the grain yield reductions to acceptable levels for adoption by subsistence farmers.

Effect of cropping system on crude protein yield and nutritive value of fodder. The high crude protein yield of the fodder for the mixtures when compared to the yield from sole stands was attributed to the higher CP concentration of the mixtures than in the forages from sole cereal stands. According to Humphreys (1978), the N content of ingested ruminant feed should be at least 1.1%. Hence, the fodder from sole cereal crops in the present study would not meet the N requirements of ruminant livestock. On the other hand, the CP content of fodder from the cereal + forage legume intercrops was higher than the minimum level (7% CP) required for effective microbial activity in the rumen (Crowder and Chheda, 1982; Van Soest, 1982).  The results of this trial indicated that  fodder from cereals + forage legumes intercropping would be expected to increase intake and utilisation by ruminants because of the higher CP content.

High proportions of lalab in the mixtures resulted in higher CP content in sorghum or maize + lablab mixtures.  This proportion of lablab was higher than the proportion of clover in the wheat + clover mixtures probably because the clovers matured earlier than wheat in the mixtures. The low proportion of clover in the mixture could also have been due to the high degree of dry clover leaf crushing as a result of handling during the process of harvesting and storage (Umunna et al., 1997).  The advantage of lablab over clover in the intercrops is that lablab, which is deep rooted probably utilises moisture reserves below the top soil layer and, thus, remains green for a much longer time than clover during the dry season. The presence of the forage legumes that had lower NDF content resulted into lower NDF content of the mixtures than in sole stands and, thus, the highest dry matter degradability (DMD).  According to Singh and Oosting (1992) any roughage feed with an NDF content of 45-65% is categorised as medium quality feed, while below 45% was categorised as high.  Thus intercropped feeds obtained in this study were classified as medium quality feeds, while forages from sole cereal stands especially wheat were low quality feeds.

The results of this study demonstrated that intercropping forage legumes with cereal crops resulted in crop residues of higher feeding value in terms of improved total biomass yield and crude protein content. These qualities have the potential to contribute to increased production in smallholder crop/livestock farming systems.  The technology of intercropping cereals with forage legumes is of particular importance to resource poor crop-livestock farmers, for it would provide improved fodder production, while maintaining cereal grain yield from the same piece of land. In cases where reduced grain yield is experienced, the production of large amounts of high-quality feed in the intercrops is enough to offset the reduced cereal grain yield through increased animal production (milk and meat).  In addition, farmers might be able to benefit from several trade-offs, such as reduced soil erosion due to better vegetative cover of the soil, nitrogen accumulation in the soil through biological N fixation (Tothill, 1986), and better feed for the animals (Urs et al., 1995).  Any of these attributes, either singly or in combination, may make intercropping cereal crops with forage legumes attractive to farmers; and even where there may be no extra yield advantages, intercropping may still be a normal practice as it is a common practice in tropical farming systems.

Effect of manure and cropping system on fodder dry matter and cereal grain yields. In general, applying fertiliser N to fodder legume cereal intercrops had been found to decrease the yields of the legume and to increase the yield of the cereal (Humphreys, 1978; Venkateswarlu, 1984). The greater fodder and cereal grain yield advantages in response to manure than for DAP/urea fertiliser was attributed to the slowly released organic nutrients, especially N and P in the manure (Powell, 1984; Wolf et al., 1989; Bationo and Mukwunye, 1991), which could have benefited the plants over the growing period. Williams et al. (1995) also noted that, whereas nutrients in fertiliser are in readily soluble form, and therefore,  become rapidly available to crops, nutrients in manure must mineralise. Nutrients in manure stay longer in the soil for the benefit of the growing plants over the growing season.

The decline of fodder DM yield in response to levels of manure higher than 7.5 t DM ha^-1 was in agreement with the fact that with continued application of most forms of nitrogen fertiliser, there is a general decline of herbage yields.  This situation could be explained by various unfavourable soil conditions such as decreased soil pH, possible phosphorous fixation brought about by increased soil acidity, increased exchangeable aluminium and decreased exchangeable calcium and magnesium (Crowder and Chheda, 1982).

The higher fodder and cereal grain yield responses obtained with manure than with DAP/urea, especially in the intercrops, was attributed to the residual effect of the nutrients from the manure (Wigg et al., 1973) and forage legumes (Waghmare and Singh, 1984) from one cropping season to the subsequent cropping seasons.  Thus, in this study, the residual unmineralised N and P from the original manure and additional annual manure applications could have built up soil organic N and P reserves which could have long-term beneficial effects on soil productivity. In addition, Powell (1986) noted that decomposed manure is well mineralised and that its application has the advantage of ameliorating soil productivity  by increasing soil pH, organic carbon, total N, exchangeable P and maintaining the C:N ratio and the cation exchange capacity (CEC) of soils. Manure increases soil organic C, resulting in a greater moisture-holding capacity (Bationo and Mokwunye, 1991) which could in itself has led to increased grain yields by extending the period of water availability during dry periods.

The results in this study further showed that,  although the fodder yield gains of manure over the control were higher in sole than intercrop, for the cereal grain yield, the greatest response to manure fertilisation was in the intercrops  compared to sole stands.  Similar findings were reported by Gibberd (1995) that intercrops gained more from manure application than did sole stands. This showed that, although in the intercrops the forage legumes could have provided more soil nutrient requirements, fertilisation was still required to support higher cereal grain production levels. The significant year x cropping system, fertiliser x cropping system and year x fertiliser x cropping system interactions were attributed to certain cropping systems (sole clover or wheat) responding better to chemical (compared with manure). This could not be explained by the results of this study and further investigation about these interactions is recommended.

Although regression analysis indicated that 13.7 t DM ha^-1 was the optimum manure level for biomass and cereal grain yields, it was noted that this was a very high rate which would be very difficult to collect and transport for smallholder dairy farming systems. However, using the overall means (Tables 2 and 3), yield responses to manure levels were, 15.5, 24, 38  and 31% fodder DM and 26, 39, 52  and 57% cereal grain yield advantage at 2.5, 7.5, 12.5 and 17.5 t DM manure ha^-1 over the control (M0), respectively. This showed that even at the lowest level of 2.5 t DM manure  ha^-1 (M1), a considerable yield advantage of manure over sole cropping was realised and was, therefore, recommended for use in smallholder crop/livestock farming systems. The recommendation was made based on the fact that an average smallholder crop/livestock dairy farmer in sub-Saharan Africa owns 2-3 dairy cattle (Gryseels and Anderson, 1983) and a mature cow averaging 400 kg can produce about 1.3 t DM of manure (Herbert, 1983). However, it was proposed that  economic analysis should be carried out on the results of this study to determine the most economic manure rate for fodder DM and grain production in cereal and forage legumes cropping systems.

CONCLUSION

This study has shown that cereal + forage legume intercrops especially when planted in rows have potential for improved fodder production, while maintaining acceptable cereal grain yields in smallholder crop/livestock farming systems. Fertilisation is required for higher fodder DM and grain yields from cereal + forage legume intercropping systems and therefore, the use of readily available on-farm cattle manure could be the appropriate technology for smallholder subsistence farmers who can not afford to purchase expensive commercial fertilisers.

ACKNOWLEDGEMENTS

The authors express deepest gratitude to the International Livestock Research Institute (ILRI) and the Livestock Services Project (World Bank) in the Ministry of Agriculture, Animal Industry and Fisheries, Uganda for financial support to  this research. We are also grateful to all staff of ILRI (Debre Zeit Research Station and Addis Ababa) for assistance. Mamadou Diedhiou, Zerihun Tadesse, Amare Atale and Aklilu Bogale are also highly appreciated.

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 (Trin.) 69(1):68-72.
• Abate, T., Sherington, J. and Mohamed-Saleem, M.A. 1994. Integration of forage and food crops grown sequentially on Vertisols under rainfed conditions in the mid-altitude Ethiopian highlands. Experimental Agri-culture 30:291-298.
• Andrews, D.J. and Kassam, A. H. 1976. The importance of multiple cropping in increasing world food supplies. In: Multiple cropping: Special publication No. 27. Papendick, R.J., Sanchez, P.A. and Triplett, G.B. (Eds.), pp.1-10. American Society of Agronomy, Madison, Wisconsin.
• AOAC (Association of Official Analytical Chemists) 1990. Official methods of analysis, 15th Edition. AOAC Inc., Arlington, Virginia 22201 USA.
• Astatke, A., Mohamed Saleem, A.M. and El Wakeel, A. 1995. Soil water dynamics under cereal and forage legume mixtures on drained Vertisols in the Ethiopian highlands. Agricultural Water Management 27:17-24.
• Bationo, A. and Mokwunye, A.U. 1991. Role of manures and crop residue in alleviating soil fertility constraints to crop production: With special reference to the Sahelian and Sudanian zones of West Africa. Fertiliser Research 29:117-125.
• Berhanu, D. 1985. The Vertisols of Ethiopia: their properties, classification and management. Fifth Meeting of the Eastern Africa Sub-Committee for Soil Correlation and Land evaluation.  Wad medani, Sudan, 5-10 Dec. 1983.  World Soil Resources Report No. 56, FAO, Rome, pp. 31-35.
• Brumby, P.J. 1991. Livestock, food production and land degradation. In: Evaluation for sustainable land management in the developing world. Volume 2, Technical Papers, Bangkok, Thailand, IBSRAM, Proceedings 12:403-412.
• Cochran, V.L. and Schlentner, S.F. 1995. Intercropped Oat and Faba bean in Alaska: Dry matter production, dinitrogen fixation, nitrogen transfer, and nitrogen fertiliser response. Agronomy Journal 87:420-424.
• Crowder, L.V. and Chheda, H.R. 1982. Tropical Grassland Husbandry, Longman, London and New York. 562 pp.
• Gibberd, V. 1995. Yield response of food crops to animal manure in semi-arid Kenya. Tropical Science 35:418-426.
• Gryseels, G. and Anderson, F.M.  1983.  Research on farm and livestock productivity in the central Ethiopian highlands: Initial results, 1997-1980,  ILCA Research Report No. 4, ILCA, Addis Ababa, Ethiopia.
• Haque, I. and Jutzi, S. 1984. Nitrogen fixation by forage legumes in sub-Saharan Africa: Potential and Limitations. ILCA Bulletin No. 20:2-13.
• Herbert, R.F. 1983. Kraal Manure. The Bulletin of Agricultural Research in Botswana. 1, Department of Agricultural Research, Ministry of Agriculture, Gaborone, Botswana.
• Humphreys, L. R. 1978. Tropical pastures and fodder crops. Longmans Group Ltd. Harlow, Essex England, 135 pp. International Livestock Centre for Africa (ILCA), 1989. ILCA Annual Report 1989, pp. 86-87.
• Kassam, A.H. 1979. Multiple cropping and rainfed crop productivity in Africa. Consultant's Working Paper No. 5. FAO/UNFPA Project INT/75/P13, Land Resources for Populations of the Future. AGLS, FAO, Rome, July 1979.
• Moreira, N. 1989. The effect of feed rate and nitrogen fertiliser on the yield and nutritive value of oats-vetch mixtures. Journal of Agricultural Science (Camb.) 112:57-66.
• Natarajan, M. and Willey, R.W. 1980. Sorghum-pigeon pea intercropping and the effects of plant population density. 1. Growth and yield. Journal of Agricultural Science (Camb.) 95:51-58.
• Nnadi, L.A. and Haque, I. 1986. Forage legume-cereal systems: Improvement of soil fertility and agricultural production with special reference to sub-Saharan Africa. In: Potentials of Forage Legumes in Farming Systems of sub-Saharan Africa. Proceedings of a workshop held at ILCA, Addis Ababa, Ethiopia, 16-19 September 1985. Haque, I. Jutzi, S. and Neate, H.J.P. (Eds.). ILCA, Addis Ababa pp. 330-362.
• Nyambo, D.B., Matimati, T., Komba, A.L. and Jana, R.K. 1982. Influence of plant combinations and planting configurations on three cereals (Maize, Sorghum, Millet) intercropped with two legumes (Soyabean, Green-Gram). In: Intercropping. Proceedings of the Second Symposium on Intercropping in Semi-Arid Areas, held at Morogoro, Tanzania, 4-7 August 1980. Keswani, C.L. and Ndunguru, B.J. (Eds.), pp. 56-62. IDRC-186e, Ottawa, Canada.
• Ofori,    F. and Stern, R. W. 1986. Maize/cowpea intercrop system: Effect of nitrogen fertilizer on productivity and efficiency. Field Crops Research 14:247-261.
• Orskov, E.R. Hovell, F.D and Mould, F. 1980. The use of nylon bag technique for the evaluation of feedstuffs. Tropical Animal Production 5:195-213.
• Osuji, P.O., Nsahlai, I.V. and Khalili, H. 1993. Feed Evaluation, ILCA Manual No. 5. International Livestock Centre for Africa (ILCA), Addis Ababa, 40 pp.
• Powell, J.M. 1984. Assessment of dry matter yield from grain yield in the West African Savanna Zone. Journal of Agricultural Science (Camb.) 103:695-698.
• Powell, J.M. 1986. Manure for cropping: A case study from Central Nigeria. Experimental Agriculture 22:15-24.
• Rees, D.J. 1986. Crop growth, development and yield in semi-arid conditions in Botswana. II. The effects of intercropping Sorghum bicolor with Vigna unguiculata. Experimental Agriculture 22:169-177.
• SAS Institute Inc., 1989. Statistical Analysis Systems Institute Inc., SAS/STAT User's Guide, Version 6, Fourth Edition, Vol., 2, Cary, NC, USA. 846 99.
• Singh, S.P. 1981. Studies on spatial arrangements in sorghum-legume intercropping. Journal of Agricultural Science (Camb.) 97:655-661.
• Singh, G.P. and Oosting, S.J. 1992. A model describing the energy value of straw. Indian Dairyman XLIV pp. 322-327.
• Skerman, J.P., Cameroon, G.D. and Riveros, F. 1988. Tropical forage legumes, Second edition, FAO. Plant Production and Protection Series No. 2, FAO, Rome 692 pp.
• Tekalign, T., Haque, I. and Aduayi, E.A. 1991.  Soil, plant, water, fertilizer, animal manure and compost analysis.  Plant Science Division Working Document No. 13.  ILCA, Addis Ababa, Ethiopia.  260pp.
• Tothill, J.C. 1986. The role of legumes in farming systems of sub-Saharan Africa. In: Potential of Forage Legumes in sub-Saharan Africa. Proceeding of a Workshop held at ILCA, 16-19 September, 1985, Addis Ababa, Ethiopia. Haque, I, Jutzi, S.C. and  Neate, P.J.H. (Eds.), pp. 162-185.
• Umunna, N.N., Osuji, P.O. and Nsahlai, I.V. 1997. Strategic supplementation of crossbred steers fed forages from cereal-legume cropping systems with cowpea hay. Journal of Applied Animal Research 11:169-182.
• Urs, S., Abate, T., Mohamed-Saleem, M.A. and Abdullah, N.S. 1995. Effects of variety, altitude, and undersowing with legumes, on the nutritive value of wheat straw. Experimental Agriculture 31:169-176.
• Van Soest, P.J. 1982. Nutritional ecology of the ruminant, Cornell University Press, Ithaca, NY, USA. 375 pp.
• Van Soest, P.J. and Robertson, J.B. 1985. Analysis of Forage and Fibrous Foods. A Laboratory Manual for Animal Science 613, Cornell University, Ithaca, New York, USA.
• Venkateswarlu, J. 1984. Nutrient management in semi-arid red soils. In: Nutrient management in drylands with special reference to cropping systems and semi-arid red soils. All India coordinated Research program for Dryland Agriculture, Hyderabad, India.  Part 2. pp. 1-56.
• Waghmare, A.B. and Singh, S.P. 1984. Sorghum-legume intercropping and the effects of nitrogen fertilisation.  II. Residual effect on wheat. Experimental Agriculture 20:261-265.
• Whyte, R.O., Nilson, L. and Trumble, H.C. 1953. Legumes in Agriculture, Rome: FAO. pp. 64-5.
• Wigg, P.M., Owen, M.A. and Mukurasi, N.J. 1973. Influence of farm yard manure and nitrogen fertilizers on sown pasture seed yield and quality, of Cenchrus ciliaris L. at Kongwa, Tanzania. East African Agriculture and Forestry Journal 38:367-374.
• Willey, W.R. 1979. Review Article: Intercropping: Its imprortance and research needs.  Part 1.  Competition and yield advantages. Field Crop Abstract 32:1-10.
• Williams, T.O., Powell, J.M. and Fernandez-Rivera, S. 1995. Manure utilisation, drought cycles and herd dynamics in the Sahel: Implications for cropland productivity. In: Livestock and Sustainable Nutrient Cycling in Mixed Farming Systems of sub-Saharan Africa, Proceedings of an International Conference, held in Addis Ababa, Ethiopia, 22-26 November 1993. Powell, J.M., Fernandez-Rivera, S., Williams, T.O. and Renard, C. (Eds.) ILCA (International Livestock Centre for Africa), Addis Ababa, Ethiopia. 568 pp.
• Wolf, J., de Witt, C.T and van Keulen, H. 1989. Modelling long-term crop response to fertiliser and soil nitrogen. I. Model description and application. Plant Soil 120:11-22.

©2002, African Crop Science Society

The following images related to this document are available:

Photo images