The performance of Stylosanthes species in different production systems in Nigeria

Abstract
Introduction
Performance of stylosanthes under different production systems
References

E. Agishi¹ and P.N. de Leeuw^2
1National Animal Production Research Institute, Shika, Ahmadu Bello University, Zaria, Nigeria.
² Kenyan Rangelands Programme, ILCA, Nairobi, Kenya.

Abstract

The different pasture production systems discussed in this paper are: (a) pure legume pastures, (b) mixed legume/grass pastures, (c) mixed legume/cereal crops, (d) legumes/rangeland, (e) legumes for seed production, and (f) legumes in short fallows. Observations are reported on the performance of the three outstanding Stylosanthes species in Nigeria, namely S. guianensis cvs Cook and Schofield, S. hamata cv. Verano and S. humilis, and factors that affect legume performance in these production systems are discussed. Research problems associated with stylosanthes production and areas for future research are enumerated.

Introduction

Among the complex interacting factors that affect livestock production, nutrition is one of the most important. For ruminants, forage is the most important source of feed. In Nigeria, natural pastures do not meet the nutritional needs of grazing ruminants. Most of the liveweight gained in the wet season is lost in the dry season, resulting in low net annual growth.

Pasture research in Nigeria has aimed at improving dry season nutrition using legumes, some of the most outstanding of which are in the genus Stylosanthes. Three species, namely S. guianensis cvs Cook and Schofield, S. hamata cv.

Verano and S. humilis have received most research attention. This paper summarises the performance of these legumes under different production systems in Nigeria, such as in pure legume pastures, mixed pastures, with crops, in rangeland and under seed production.

Performance of stylosanthes under different production systems

Pure legume pastures

Establishment: Stylosanthes species can be established by drilling or broadcasting seed into recent fallow, burned, disced or well-prepared seed beds (Haggar, 1969; Haggar et al, 1971; Agishi, 1971; de Leauw, 1975). Well-prepared seed beds have given the best results and are recommended, particularly for seed crops. The application of dieldrin immediately before seeding at a rate of 200 ml a.i. in a 1% solution increased the plant density of Townsville stylo and Verano stylo more than three-fold and two-fold, respectively, over the control (Roeleveld and de Leeuw, 1978). Applying phosphate increases early establishment of Cook, Townsville and Verano stylos (de Leeuw, 1975, Holm, 1979; Agishi, 1982). A basal N application is beneficial on N-deficient soils, but high levels of N application have been shown to reduce nodulation and N fixation (Fayemi et al, 1970).

Schofield stylo grown under irrigation between March and May developed faster than when sown under rain-fed conditions during August and October (Agishi, unpublished). It was also found that stylo grew faster than gamba (Andropogon gayanus) during the dry season but slower in the wet season. The generally accepted view that stylo grows more slowly during the early stages of its establishment than do grasses may not apply to the later part of the dry season in northern Nigeria, when mean maximum air temperature is 30 C and solar radiation is at its highest.

Herbage dry-matter yield: Applying P and S fertilizers increases the dry-matter yield of S. humilis (Fayemi et al, 1970), and similar results have been reported for Cook, Townsville and Verano stylos (Agishi, 1971; 1982; Holm, 1979). At Shika, applying 20 and 40 kg P/ha to 2-year-old stands of S. humilis and Townsville stylo, respectively, gave the highest dry-matter yields (Agishi, 1971). Applying phosphate fertilizer favoured weed invasion, particularly at higher P levels, but the effect was less in the second-year crop (Table 1). Small applications of phosphate fertilizer have been found to increase the yield of weeds more than that of stylo that had been established on a recent fallow (de Leeuw, 1972).

Table 1. The effect of single superphosphate on dry-matter yields of Schofield and Townsville stylos. kg P/ha

Variety	Year	0	20	40	60
Schofield	1970	1282a	2783	2697	2409
		(53.1)b	(51.3)	(35.0)	(28.3)
	1971	4842	7159	5340	5299
		(69.3)	(70.7)	(58.9)	(59.2)
Townsville	1970	1800	2489	2630	2534
		(92.6)	(60.6)	(58.7)	(52.0)
	1971	2914	5960	7238	6734
		(81.5)	(71.4)	(63.6)	(69.5)

^a. Dry-matter yield of legume component.
^b. Percentage of stylo in total yield.

Light defoliation, either by cutting or by grazing animals, increases the legume content of the pasture (Haggar, 1971) but reduces the total herbage dry-matter yield when compared with an undefoliated pasture.

Stylosanthes/grass mixtures

The proportion of legume in a legume/grass mixture depends on:

The species of legume
The grass species in the mixture
The initial ratio of legume seed to grass seed
The type of seed bed
Time and method of seeding
The type, quantity and timing of fertilizer application
Defoliation practice
Relative palatability of legume to grass and
The age of the pasture.

The salient points of these factors are discussed below.

Species mixtures: In both the rain-forest and the derived savanna zones of Nigeria, Schofield stylo was found to be compatible with Andropogon gayanus and Melinis minutiflora, but incompatible with Cynodon plectostachyus, Digitaria decumbens, Panicum maximum and Pennisetum purpureum (Hedrick, 1961; Adegbola and Onayinka, 1966). In the Guinea savanna, Schofield stylo combined well with Chloris gayana and Panicum maximum (Blair Rains, 1963; Haggar, 1969; 1971), and Cook stylo with buffer grass (Holm, 1979). Verano stylo has been reported to be compatible with A. gayanus, Cenchrus ciliaris, C. gayana, P. maximum var trichoglume, P. maximum cv. Gatton and Urochloa mosambicensis (Shehu, 1980; Agishi, 1982). Townsville stylo combines poorly with C. plectostachyus and Digitaria smutsii (Anon., 1972). At Shika, Onifade (1982) found that herbage dry matter yields of Cook stylo increased with increasing proportions of Cook stylo seed in stylo/Rhodes grass mixtures and that this effect persisted into the second year (Table 2). Total dry-matter yields were similar in all treatments, showing that increases in legume dry-matter yields resulted in proportional decreases in the grass component.

Table 2. Effect of Cook stylo/Rhodes grass seeding ratios on dry-matter yield.

		Legume % in seed mixture
		30	40	50	60	70
1979	DM yield	4290	4210	4650	5010	5200
	% legume	18.9	20.9	25.8	27.9	31.5
1980	DM yield	8240	8440	8390	8280	8610
	% legume	18.1	21.0	23.6	25.7	36.6

Townsville stylo and Verano stylo performed poorly when sown into existing pastures of green panic and Gayndah buffer grass (de Leeuw et al, 1980). However, defoliation increased legume yield, particularly at higher defoliation frequencies (Table 3). Haggar (1971) reported that the proportion of stylo in an unfertilized stylo/Rhodes grass mixture was lower under 4 cuts than 2 cuts per year, but that the proportion of stylo diminished more under the 2 cut than the 4 cut treatment when N was applied (Table 4).

In the Northern Guinea savanna, Schofield, Townsville and Verano stylo can be planted at any time between June and July without having a significant effect on yield (Agishi, 1971; Roeleveld and de Leeuw, 1978). Erratic rainfall experienced during this period, however, reduces Townsville stylo production, but has little effect on Verano stylo yield (Roeleveld and de Leeuw, 1978).

In Shika, Onifade (1982) compared broadcasting Cook stylo and Rhodes grass in alternate rows and mixing the seeds before broadcasting. He found that seeding by broadcast gave the best legume dry-matter yield. Herbage dry-matter yields showed no significant differences between sowing treatments.

Table 3. Total dry matter yield of Townsville stylo and Verano stylo undersown in panic and buffer grass pastures.

Treatments	Green panic	Townsville	Verano	Buffel grass	Townsville	Verano
	Yield (t/ha)
Cut in July and October	4.5	0.56	0.73	6.00	0.25	0.31
Cut in July, August and October	3.9	1.04	1.50	5.00	0.73	0.67
Cut in July, August, September and October	3.6	1.41	1.44	5.2	0.90	1.23
Cut only in October	4.2	0.08	0.46	5.2	0.13	0.21

Table 4. Effect of N application on stylo and crude protein contents of stylo/Rhodes grass pasture under two and four cuts per year.

Nitrogen level (kg/ha)	Percentage stylo		Crude protein (%)
	4 cuts	2 cuts	4 cuts	2 cuts
0	37.7	41.6	11.9	10.2
84	30.4	15.8	13.2	11.3
168	22.2	106.0	13.8	12.5
252	20.9	4.3	14.6	11.0

Fertilizer application: Applying N fertilizer on a legume/grass mixture at Shika reduced the legume component, the amount of reduction being related to N level, especially under light defoliation (Haggar, 1971). It is possible that applying N to frequently cut stylo increases the crude protein content of the stylo (Table 4). Applying nitrochalk (26% N) to a Cook stylo/Rhodes grass mixture at 0, 100 and 200 kg N/ha, gave legume dry-matter yields of 2.5, 2.10 and 1.76 t/ha respectively (Onifade, 1982). Superimposing cutting regime on N treatment increased the legume content only slightly.

Both P and S reduce grass dry-matter yields, particularly in the second year, but increase the yield of a stylo/rhodes grass mixture (Haggar, 1971). Haggar also found that S had a greater effect on stylo yield when applied with up to 67 kg P/ha. The application of micronutrients has been reported to be beneficial to Cook stylo and Verano stylo growth (Mohamed-Saleem, 1984) but detailed work is yet to be done in this field.

Defoliation management: Frequent defoliation has been reported to increase the proportion of Townsville and Verano stylos in mixtures with buffer grass and green panic (de Leeuw et al, 1980). The dry-matter yield of stylo in a stylo/rhodes grass mixture was higher under 4 cuts than under 2 cuts per year (Haggar, 1971). A similar finding was also reported by Onifade (1982) on Cook stylo/rhodes grass mixture. Both Verano stylo and Townsville stylo contributed more to the dry-matter yields of mixtures with buffer grass and green panic under defoliation than when non-defoliated (Shegu et al, 1980).

Observations in the Northern Guinea savanna have shown that both frequent and close defoliation of S. humilis/grass mixtures result in pronounced reduction in the proportion of the sown species and a dominance of such unpalatable weed species as Sporobolus pyramidal is, Brachiaria stigmatisata and Sida spp. Verano stylo, on the other hand, increases its proportion in a mixture with grass when grazed at high stocking rates. For example, the proportion of Verano in mixture with buffer grass, grazed at 1.1, 1.67 and 3.33 heifers/ha, was 10.2, 16.2 and 13.8%, respectively in the first year and increased to corresponding values of 28.3, 44.7 and 60.5% 4 years later (Agishi, 1982).

Stylo is reported to have low palatability to stock in the wet season (Blair Rains, 1963; de Leeuw and Brinckman, 1974), which explains the increase in the proportion of stylo and the decrease in that of grass under grazing (Anon., 1972). Observations at Shika, however, show that under high stocking rates, when selectivity is low, cattle readily graze stylo during the wet season. At high stocking pressure, the proportions of both sown grass and stylo decrease drastically due to increases in volunteer grasses and other weeds, but the decrease is greater in stylo. Verano stylo is readily eaten by cattle in the wet season, but is preferred by goats and sheep at a mature stage (Agishi, 1982).

The proportions of S. humilis, Townsville stylo and Verano in mixtures with grasses increase significantly with time up to a point (de Leeuw, 1972; Adamu, 1977; Shehu, 1981; Onifade, 1982; Agishi, 1982). This increase in legume content with time can be modified by management methods already described above and by leaf fall. After 3 years, stylos tend to disappear from the pasture.

Stylosanthes/crop mixtures

Early experiments at Shika on the establishment of Schofield stylo and Townsville stylo into maize and millet showed that these species could be sown under cover crops provided the canopy was removed 4-7 days after seeding the legumes. The earlier maturing millet with an open canopy was more suitable than maize, which has a thick canopy (de Leeuw, 1975). Legume plant counts in the first and second years of establishment are shown in Table 5. Plant populations for the early cut were comparable to populations observed when these legumes were sown in pure stands. The Table shows the detrimental effect of late cutting of maize (harvested for grain) on legume populations. The survival of Townsville stylo and S. humilis was three and seven times higher, respectively, when the maize was harvested early than when it was harvested for grain. With millet, the planting density and date of harvest of the millet had little effect on the legume regrowth.

In a current experiment at Shika on cereal/legume mixtures involving Cook stylo, maize and sorghum sown in alternate rows, stylo establishment was poor. However, by August of the second year, stylo cover had reached 50-75% and 75-90% in maize and sorghum plots respectively (Mzamane and Agishi unpublished).

Table 5. The effect of plant population and cutting date of maize and millet on survival of undersown stylo and Townsville stylo.

Cereal	Legume	Time of harvest	Cereal population (plants/ha)						Mean
			48,000		71,000		143,000
			1st yr	2nd yr	1st yr	2nd yr	1st yr	2nd yr
Maize	S. humilis	Early	283	132	237	57	332	71	87
		Late	198	13	591	25	156	3	165
	Townsville	Early	229	66	254	97	167	163	109
		Late	280	41	192	14	161	39	31
Millet	S. humilis	Early	79	59	119	46	237	99	68
		Late	251	65	120	69	146	75	70
	Townsville	Early	219	156	195	198	75	338	231
		Late	82	161	220	237	134	138	179

A more detailed study of stylo/crop interactions is currently underway in an ILCA project in the subhumid zone of Nigeria (Mohamed-Saleem, 1984). Results obtained so far show that:

1. Sowing Cook stylo or Verano stylo at the same time as sorghum gave the highest legume herbage yield but drastically reduced sorghum grain yield. Undersowing these legumes 3 and 6 weeks after the sorghum gave a good balance between sorghum grain yield and legume herbage yield and also increased the quality of harvested fodder (Mohamed-Saleem, 1984).

2. Planting stylo and sorghum either on ridges or on flat seed beds had no significant effect on stylo yield but planting on ridges increased sorghum yield.

3. Spraying the existing stylo crop to introduce a crop reduced legume yield, particularly on ridged plots. At Shika, sowing sorghum into a 2-year-old Verano stylo pasture in rows 2 m apart did not affect the dry-matter yield of the stylo. Removing stylo plants within a 30 cm radius of each sorghum plant improved sorghum establishment. Applying N significantly increased sorghum herbage and grain yields (Agishi, 1978).

Stylosanthes in recent fallow

At Shika, introducing S. humilis, Cook stylo and Verano stylo into recent fallows has been found to be the cheapest way of establishing them. In this method, plots to be sown to the stylos are planted to millet or early maturing maize, which are harvested in late August or early September. Immediately after harvest, the land is harrowed and planted to stylo. Alternatively, S. humilis or Verano seeds can be planted directly into the fallow by broadcasting or drilling. The young seedlings make limited growth by the end of the rains but remain alive for the 5-6 months of the dry season. The seedlings resume fast growth with the onset of the rains, competing strongly with associated weeds. The application of 8-16 kg P/ha enhances the dominance of the legume. The method has many advantages:

· The cost of a separate seed bed preparation is reduced or virtually eliminated.

· The phosphate applied to the cereal crop benefits the legume.

· It is easier to manage than cereal/legume mixtures.

· It is cheaper to establish than a pure legume pasture.

· on land that has been infested with noxious weeds or has erosion problem, this method can be used to make it productive again.

· It requires only 2-3 weeks of rain at the end of the growing season for the legumes to establish well.

The main disadvantage with this method is that the legume cannot be grazed in the year of establishment.

Stylosanthes in rangeland

Attempts to establish and grow stylos in the Nigerian rangelands have had mixed results. The successes reported in Australia with establishing Townsville stylo in rangeland have not been matched in trials in Nigeria. The poor performance of Stylosanthes has been attributed to removal of most of the seed by ants and by shading from the existing vegetation (de Leeuw, 1975). Stylo has been successfully established in the southern Guinea savanna at Mokwa by sowing into cultivated strips in rangelands (Agishi, 1971). In the northern Guinea savanna, where strip cultivation was limited to only disc-harrowing, establishment and performance were poor and were only improved when oversowing was repeated in the following year (Haggar et al, 1971). It was also found that applying superphosphate increased stylo establishment but the effect did not persist. Earlier studies by Foster (1961) showed that stylo could be introduced into the range by feeding seed to the grazing cattle, but application of this technique is limited by the high cost of stylo seed.

Numerous efforts by a World Bank team to establish S. humilis and Verano stylo by broadcasting seed onto harrowed strips in degraded grazing reserves in the Sudan zone (Kukar-Jangarai) failed (Perrier, 1982). It was postulated that removal of seed by ants and the hard soil surface reduced germination. A controlled study carried out later on methods of stylo establishment in this reserve indicated that harrowed strips recrusted fast and so provided no improvement in water infiltration (Perrier, 1982).

De Leeuw's (1972) studies on this grazing reserve showed that S. humilis and Townsville stylo established and grew best on burned or cultivated sites, with phosphate application giving an additional benefit. Verano stylo has performed better in the Sudan zone than S. humilis (Perrier, 1982).

Stylosanthes planted in unimproved seed beds has performed best in Nigeria in Benue state. Townsville stylo was first introduced to Yander Agricultural Research Station in 1956 and today it covers many square kilometers of fallow land, roadsides and rangeland in the state. The centres of spread are the cattle routes. Townsville stylo associates well with Imperata cylindrica. Best growth is to be seen in areas receiving repeated defoliation such as cattle routes and sport fields.

Stylosanthes for seed production

S. humilis, Townsville stylo and Verano stylo grown for seed production in Nigeria are managed in one of the following ways:

a. The pasture is left unweeded throughout the year and seeds are harvested when ripe,
b. The pasture is weeded once or twice a year before seed harvesting,
c. The pasture is grazed between June and August, and then the paddock is closed to grazing to allow seeds to form and mature.

Method (a) is not suited to producing seed of any of the stylo species; about 8 weeks after establishment, the pasture is dominated by weeds. Method (b) is suitable for all three species, as competition for light and soil nutrients is almost totally eliminated. Plants grow vigorously and produce a lot of bulk and seed, particularly in the second and third years after establishment. Method (c) favours Verano stylo and Townsville stylo more than S. humilis, which tends to thin-out where grazing pressure is high. However, at high pressure, grazing should be stopped at least 6 weeks before the end of the wet season, in order to allow stylo to recover and set sufficient seed.

Stylos managed for seed production usually regrow from seed at the onset of the rains. In this way even the perennial stylos are managed as annuals. Each year, a new pasture is re-established from seeds, and so the productive life of the pasture seed crop is longer than grazed pastures of the same species. Observations have shown that grasses and other nitrophilous weeds become dominant after 4 to 5 years, even in seed crops, due mainly to the accumulation of N in the soil. In such a situation, hand weeding or grazing by cattle followed by phosphate application may help to increase the legume content of the pasture.

Stylosanthes in citrus orchards

In the humid and sub-humid areas of Nigeria, many citrus orchards have been established. Because of the normally wide spacings between plants, weed infestation is a problem. Weeding large tracts of land under citrus is not feasible, but weeds can be controlled by planting the areas between the trees with a stylo pasture. The areas between the trees are harrowed and seeds of stylosanthes broadcast onto it.

With the break in rainfall during December and January, the Verano stylo or Townsville stylo is cut with sickles and the fallen seed swept from the ground. The cleaned seed is sold, but the cut herbage is left on the farm to be grazed by livestock and to provide mulch for the trees. Difficulties have been experienced in cutting Cook stylo because of the relatively large, woody stems. Though goats and sheep would not touch any of the stylos during the wet season in the first 2 years of the pasture, they now readily consume large quantities. It was also observed that the soil under a 5-year-old orchard, when planted to Townsville stylo, became very friable and had many earthworm casts only 2 years after planting the stylo. A similar observation was made in the soil which had been under Cook stylo for 3 years. Harvesting the stylo and grazing by sheep and goats help to keep down the weeds in the orchard.

Problems and gaps in Stylosanthes research in Nigeria

Lazier (1984), writing on Stylosanthes in West Africa, listed the following problems with stylos:

a. Sensitivity to shading.
b. Low persistence beyond 3 years.
c. Termite attack, causing death of stands.
d. Fire outbreaks on mature Stylosanthes pastures, causing huge losses.
e. Anthracnose, which can cause death of stands.
f. Difficulty in obtaining Stylosanthes germplasm.

In addition to these problems, there is an acute shortage of trained manpower in pasture legume research.

Some of the gaps in Stylosanthes research in Nigeria include:

a. Narrow germplasm base.

b. The nutrient requirements of Stylosanthes on different soil types are not fully known.

c. Little is known on the effects of micronutrients on the performance of Stylosanthes in Nigeria.

d. Stylosanthes spp. are being recommended for inclusion in cereal crops, but the nature of competition in these mixtures is yet to be understood.

e. The performance of Stylosanthes spp. under irrigation has not been properly investigated.

f. Technology for weed control in Stylosanthes pastures needs to be developed.

g. Cutting and grazing studies are very short term, and little is known of the long-term performance of Stylosanthes species.

To fully exploit the great potential of the genus Stylosanthes in improving livestock production in Nigeria, there is need to intensify research to cover to points outlined above.

References

Adamu A M. 1977. The effect of stage of growth on nutrient content and in vitro dry matter digestibility of tropical forages. M.Sc. Thesis, Ahmadu Bello University, Zaria, Nigeria.

Adegbola A A and Onayinka B. 1966. The production and management of grass/legume mixture at Agege. Nig. Agric. J. 3: 84-91.

Agishi E C. 1971a. Use of legumes for livestock in Nigeria. Sam. Agric. Newsl. pp. 115-119.

Agishi E C. 1971b. Effect of sowing dates and superphosphate fertilizer on Stylosanthes guyanensis and S. humilis. Shika Agric. Station. Annual Report.

Agishi E C. 1978. Seed production in Verano stylo. In: Annual Report 1977/78. NAPRI. Shika, Nigeria.

Agishi E C. 1982. Verano stylo- a promising new pasture legume for the Nigerian savannas. National Conference on Beef Production in Nigeria, Kaduna, Nigeria; Mimeo. 19 pp.

Anon. 1972. Yield of mixed Stylosanthes humilis pastures. In: Annual Report 1972. Shika Agricultural Research Station, Zarai, Nigeria.

Blair-Rains A. 1963. Grassland research in Northern Nigeria, 1952-62. Institute of Agricultural Research, Ahmadu Bello University, Zaria, Nigeria. Samaru Misc. Paper No. 1, 70 pp.

Fayemi A A, Odu C T I, Fagbami A. 1970. Nutrient requirement of tropical pasture legumes. I. Influence of soil type and N levels on the growth, nodulation and N-fixation of Centrosema pubescens Bent and Stylosanthes gracilis L. Niger. J. Sci. 4:311-318.

Foster W H. 1961. Notes on the establishment of a legume in rangeland in Northern Nigeria. Emp. J. Exp. Agric. 29: 319-322.

Haggar R J. 1969. A guide to the management and use of stylo (Stylosanthes gracilis). Shika Agricultural Research Station, Zaria, Nigeria, 3 pp.

Haggar R J. 1971. The production and management of Stylosanthes gracilis at Shika, Nigeria. I. In sown pasture J. Agric Sci. (Camb) 77: 427-436.

Haggar R J. de Leeuw P N and Agishi E. 1971. The production and management of Stylosanthes gracilis at Shika, Nigeria. II. In Savanna Grassland. J. Agric. Sci. (Camb) 77: 437-444.

Hedmick D W. 1961. Forage research in western Nigeria. W. Region. MANR.; IDS. Mission to Western Nigeria. Final report of the pasture agronomist. 98pp.

Holm A. 1979. Pasture Investigations (1976/78), in pasture improvement project - A report submitted to the Federal Govt. of Nigeria, by Australian Agricultural Consulting and Management Company Pty. Ltd.

Lazier J R. 1984. Global ventures in Stylosanthes. IV. West Africa. In: The biology and agronomy of Stylosanthes. pp. 503-527.

de Leeuw P N. 1972. Establishment of Stylosanthes humilis. In: Annual Report. Shika Agricultural Research Station, Zaria, Nigeria.

de Leeuw P N. 1975. The establishment of Stylosanthes spp. with cereal companion crops of maize and early millet. In: Annual Report. Shika Agricultural Research Station, Zaria, Nigeria. pp. 49-51.

de Leeuw P N and Addingi C J. 1978. Establishment of Stylosanthes guyanensis under cereal crops. In: Annual Report. NAPRI, Zaria, Nigeria.

de Leeuw P N. Onifade O S and Kera B. 1980. The undersowing of Stylosanthes spp. in sown grass pastures. Annual Report of Research in Animal Production, NAPRI. Zaria, Nigeria. p 83.

de Leeuw P N and Brinckman W L. 1974. Pasture and rangeland improvement in the Northern Guinea and Sudan Zone of Nigeria. In: Loosli J K, Oyenuga VA and Babtunde GM (eds). Animal production in the tropics. Heineman (Nig) Ltd.

Mohamed-Saleem M A. 1984. Crop-forage interactions. Paper presented at the second ILCA/NAPRI symposium on livestock production in the subhumid zone of Nigeria, Kaduna, Nigeria.

Onifade O S. 1982. Effect of seeding ratio, sowing method and nitrogen level on the growth of Rhodes grass (Chloris gayana cv. Callide)/stylo (Stylosanthes guianensis cv. Cook) mixture in Northern Nigeria. M.Sc. Thesis. Ahmadu Bello University, Zaria, Nigeria.

Perrier G. 1982. Effects of ridging and harrowing with or without seeding and fertilization on forage production and infiltration rates on rangeland in Nigeria. A report to Livestock Project Unit, Kaduna. 12pp.

Roeleveld A and de Leeuw P N. 1978. Effect of dieldrin application at seeding on the establishment of Townsville and Verano stylo in savanna.

Shehu Y. 1981. Production of grass/legume swards. M.Sc. Thesis, Ahmadu Bello University, Zaria, Nigeria.

Shehu Y. Agishi E, Akinolo J O. de Leeuw P N. 1980. The performance of pure and mixed Townsvills and Verano stylo pastures sown in shrub savanna. Annual Report of Research in Animal Production. NAPRI. Zaria Nigeria. pp. 92-94.