The agronomy, physiology and use for animal production of Kikuyu grass (Pennisetum clandestinum) - A review.

M. HERRERO, R.H. FAWCETT, G. RUSSELL and J.B. DENT
Institute of Ecology and Resource Management, University of Edinburgh, Scotland, UK

Correspondence: M. Herrero, IERM, School of Agriculture, University of Edinburgh, West Mains Road, Edinburgh EH9 3JG, Scotland, UK.
Abstract

The factors affecting dry matter yields, nutritive value and animal performance of kikuyu grass (Pennisetum clandestinum) are reviewed. These are related to physiological concepts to explain existing management strategies and to develop suitable guidelines for their improvement. Future areas of research are proposed.
Kikuyu grass is a stoloniferous and rhizomatous tropical pasture species which originated from the Kenyan highlands. Now widely distributed throughout the subtropics and tropical highlands, it is regarded as an important pasture species for animal production. It produces high dry matter yields when nitrogenous fertilisers are applied, tolerates low soil moisture conditions, is well adapted to cool temperatures and withstands occasional frosts. In the subtropics, its seasonal growth patterns are largely determined by temperature and it is sometimes associated with temperate grass species and tropical legumes to match the feed requirements throughout the year. In the tropics, it is difficult to manage in associations with legumes due to its aggressive growth habits.
Kikuyu grass can have a high nutritional value if properly managed. This usually consists of hard grazing or mulching to reduce the proportion of stolons, which have a lower nutritive value and mature faster. However, more information is needed about the kinetic properties of the different nutritional fractions and their balance in order to predict intake and animal performance more accurately.
Well fertilised kikuyu swards can withstand heavy stocking rates and produce high animal production per hectare. However, individual animal production from kikuyu grass is only considered moderate. Mean milk production levels of 9-12 kg/cow/d are common while maximum levels of 15-16 kg/cow/d have been found. Typical growth rates are 0.4-0.6 kg/animal/d. To obtain the desired levels of production, swards should be kept in a short, dense and leafy state since animals tend to avoid grazing the stoloniferous material.
It is suggested that directing research efforts on kikuyu grass grazing systems towards a more physiologically and behaviourally-oriented approach would increase our understanding of the role and management of this grass species.
Introduction

Kikuyu grass (Pennisetum clandestinum) is a stoloniferous and rhizomatous C4 tropical pasture species originally from the Kenyan highlands. It has become one of the most important pasture species in subtropical regions of countries like Australia (Minson et al. 1993), New Zealand (Percival 1978; Piggot 1985) and South Africa (Marais et al. 1987), where it provides a substantial amount of the summer and autumn feed in some livestock production systems. In the tropics, its distribution is confined to highland areas of Africa, Central and South America and Asia, where it is also highly regarded for animal production.
Mears (1970) reviewed the available literature on kikuyu grass and concluded that this species had several desirable attributes as a tropical pasture, but that it had received little attention from research workers. Since then, a vast amount of information has been produced regarding its agronomic characteristics, nutritional value, management and animal production potential.
This review provides an update of research on kikuyu pastures since the Mears review was published. It attempts to integrate concepts from the different disciplines, rather than to analyse each area separately, in order to provide some insight into their interactions and to more validly explain some of the results in the literature. Such an approach may help to understand and develop suitable management guidelines to improve grazing systems for this grass species.
Readers requiring detailed information on the botanical description, geographical distribution and diseases of kikuyu grass are referred to Mears (1970).

Genetic variability

From 3 ecotypes (Molo, Kabete and Rongai) in Kenya, which rarely produce seed, 4 free-seeding cultivars (Whittet, Breakwell, Noonan and Crofts) have been bred and released in Australia. Six lines have been identified in New Zealand. Cultivars or lines vary in morphology and differ in dry matter (DM) production because of varying tolerance of cooler conditions.
Mears (1970) stated that 3 ecotypes had been recognised in the Kenyan highlands. These were Molo, Kabete and Rongai and they differed in leaf morphology and flowering behaviour. However, he stated that these differences were difficult to recognise in existing pastures and that in Australia, common kikuyu was a mixture of them. Before 1970, kikuyu pastures were established vegetatively, as male sterility is normally a heterozygous condition in common kikuyu (Piggot and Morgan 1986). A mixture of male sterile but female fertile and fertile bisexual plants was produced resulting in very low or rarely observed seed production.
Breeding of kikuyu therefore began, and is most advanced, in Australia, where 4 cultivars have been released since 1970. These are Whittet (Wilson 1975) released in 1970, Breakwell (Wilson 1975) released in 1972, and Crofts (Anon. 1983a) and Noonan (Anon 1983b). In contrast to common kikuyu, these cultivars produce only (i.e. Whittet and Noonan), or mostly (Crofts and Breakwell), bisexual plants and can produce large quantities of seed.
Whittet and Breakwell varieties were introduced primarily for their ability to produce seed (Wilson 1975; Quinlan et al. 1975) which made the establishment of kikuyu swards easier. Noonan was selected primarily for its resistance to kikuyu yellows - a serious fungal disease - (Anon. 1983b), and Crofts for its adaptation to cooler regions (Anon. 1983a; Pearson et al. 1985).
There are also morphological differences between the cultivars. Whittet kikuyu is taller, has longer internodes, wider leaves and thicker stems than common or Breakwell kikuyu (Wilson 1975). It forms a more open sward and is therefore considered more suitable for associations with legumes (Wilson 1975; Rumball and Lambert 1985). Breakwell is similar in morphology to common kikuyu. It has a prostrate growth habit, shorter internodes and tillers profusely. Crofts is the tallest cultivar but has narrower leaves and thinner stolons than Whittet (Anon. 1983a). These differences are most apparent at cooler temperatures (Anon. 1983a). When immature, Noonan kikuyu cannot be distinguished easily from Whittet, but Noonan plants seem to have shorter leaves and stolons at more advanced stages of growth.
In New Zealand, 6 lines have been identified, of which some are male sterile and others are fertile (Piggot and Morgan 1985). Piggot and Morgan (1985) found with plants grown in pots, that total DM yield was relatively similar in all 6 lines, however there were differences in the proportions of leaf and stem produced. Piggot and Morgan (1985) and Piggot (1991) concluded that these New Zealand lines did not outyield the Australian cultivars nor the high altitude Kenyan accessions they tested (Molo and Njoro). The morphological differences between the Australian cultivars and the New Zealand and Kenyan lines evident in these studies were less obvious under field conditions, since pot experiments with individual plants do not reflect the processes of canopy development and senescence in established swards. As the leaf area index (LAI) of artificial swards is less than in the field, light usually penetrates to the lower leaves thus reducing leaf senescence. As a consequence, the leaf to stem ratio in these pot experiments overestimates the value in field swards. For example, Piggot (1991) reported leaf:stem ratios of 3.0:1-3.3:1 for Whittet kikuyu and the other Australian cultivars. This contrasts with the results of 1.4:1 and 1.8:1 obtained by Hacker and Evans (1992) and Köster et al. (1992) respectively, in established swards. For common kikuyu, Mears and Humphreys (1974b) found that, under field conditions during the summer/autumn period, leaf:stem ratios were close to 1.1:1, and Piggot (1991), in his second trial, reported mean values close to 1.5:1 for the lines, which are selections from common kikuyu.
Differences between cultivars or lines in total DM production seem to depend on their ability to grow at cooler temperatures (Pearson et al. 1985; Rumball and Lambert 1985). The cold-tolerant varieties produce more dry matter, reflecting their ability to extend the growing season during late autumn and early spring in these subtropical regions. Although cultivars and lines vary in leaf size, stolon yield, flowering behaviour and vigour, the effect of these differences on dry matter production and sward composition is slight in comparison with the effect of grazing management and environmental factors such as temperature and rainfall (Rumball and Lambert 1985). N utilisation does not seem to vary between cultivars (Pearson et al. 1985). Nevertheless, trials to separate the genotype effects on morphology and yield under grazing have not been designed.
It has been suggested that all cultivars have similar in vitro DM digestibility and mineral concentrations (Anon. 1983a). However, the differences in nutritive value between cultivars and lines has rarely been studied systematically. Rethman and de Witt (1988) found no differences in crude protein content (CP) between Whittet and a local line, although Quinlan et al. (1975) suggested that Whittet had an above average CP. The interpretation of the results is complicated by possible differences in the unstated proportion of leaf and stem in the material analysed (see section on nutritive value).

Seed production

Experimental production of seed from kikuyu grass has reached 700 kg/ha (Wilson and Rumble 1975), but under commercial conditions, reported seed yields range from 118-500 kg/ha (Wilson 1970; Wilson et al. 1975; Quinlan et al. 1975). Kikuyu grass seed usually has a germination rate of 75-90% (Wilson 1970; Quinlan et al. 1975; Gardener et al. 1993a). Gardener et al. (1993a) reported that the specific gravity of kikuyu seeds was 1.27g/cm3 and found seed dimensions of 2.4, 1.5 and 1.1 mm for length, width and depth respectively. Average seed weights of 2.3-2.6 mg (380-434 thousand seeds/kg) have been found (Pearson and Ison 1987; Piggot 1991), and Pearson and Ison (1987) reported typical sowing rates are 1-6 kg/ha .
Management of kikuyu grass for seed production requires short mowing or grazing (Wilson and Rumble 1975) to stimulate the production of secondary tillers, which flower prolifically and form seeds (Wilson et al. 1975). For a detailed methodology, see Wilson et al. (1975).
If ingested, kikuyu grass seeds survive ruminant digestion to some extent (Gardener et al. 1993a, 1993b), and are spread by cattle (Wilson and Hennessy 1977; Gardener et al. 1993a, 1993b) and sheep (Rethman 1989) via faeces. About 50% of the consumed viable seeds still germinate after digestion, and this seems to be of ecological importance for the species survival and dispersal during wet periods (Rethman 1989, Gardener et al. 1993a). Rethman (1989) also found that the success of establishment of seedlings via sheep faeces was lower than that of seedlings in cattle dung pats due to their lower moisture content.

Factors affecting growth rate and dry matter yields

Soil nutrients and fertilisation

Nitrogen (N) fertilisers. The response of kikuyu grass to N fertilisers is well documented. Annual DM yields on well watered soils have ranged from 4.0 t/ha with modest N applications (<100kg/ha N) (Cross 1979b) to 30-32 t/ha at rates above 850 kg/ha N (Colman 1966; Whitney 1974b; Colman and O'Neill 1978). However, annual DM yields for current N fertilisation practices (< 500 kg/ha N) have ranged from 4.5-17.3 t/ha depending on soil and environmental factors (Whitney 1974a, 1974b; Whitney and Tamimi 1974; Forde et al. 1976b; Taylor et al. 1976b; Cross 1979a; Tainton et al. 1982; Davies and Hunt 1983; Castillo et al. 1983; Brockett and Gray 1984; Cook and Mulder 1984a; Lowe and Bowdler 1984; Consentino et al. 1985; Rethman and de Witt 1988; Rumball 1991). Without N fertilisers, yields are substantially lower and confirm Mears' (1970) conclusion that kikuyu needs high soil fertility to be productive. However, it is difficult to make comparisons even between experiments carried out in similar regions because experimental procedures have not been standardised. For example, cutting height varies between experiments from 3 cm (Colman and O'Neill 1978) to 10 cm (Cook and Mulder 1984a). This is important due to the differences in herbage accumulation and density across the sward's vertical strata. The lower horizons are usually denser and contain most of the dry matter, thus underestimating the total DM yields more than proportionally at higher cutting heights. Murtagh et al. (1980a) found differences of up to 71% in total DM yield when cutting kikuyu swards at heights of 8 or 12 cm above the ground, although leaf yield was unaffected.
The efficiency of response of a crop to applied N fertiliser is largely dependent on the soil organic matter (OM) concentration, which determines the potential of soils to supply N and the slope of the response (Russell 1973), and by temperature (see below). From the analysis of a series of experiments, Mears (1970) concluded that the mean efficiency of response of kikuyu grass to N fertiliser ranged from 13-27 kg DM/kg N. Similar results have been obtained by Kemp (1975), Colman and O'Neill (1978) and Cook and Mulder (1984a) in Australia, Soto et al. (1980) in Colombia, Consentino et al. (1985) in Brazil and Tainton et al. (1982) in South Africa. Higher average responses (> 40 kg DM/kg N) have been reported by Whitney (1974b), Cross (1979a) and Anon. (1987) and can be explained by 2 reasons. First, the higher responses in Whitney's (1974b) trial in Hawaii were probably caused by temperatures nearer the optimum for growth which resulted in a higher potential growth rate in the absence of nutrient limitations. This is in agreement with Mears and Humphreys (1974a), Colman and O'Neill (1978) and Murtagh and Moore (1987) in subtropical Australia, who observed seasonal differences in the efficiency of response to N fertiliser, with responses being higher during the summer-autumn period, where temperatures are higher. The results reported by Cross (1979a) and Anon. (1987) cannot be explained by this factor, since the South African temperature regimes where kikuyu grows are lower. Possibly these higher responses were caused by lower soil OM concentrations, which were enhanced when N was applied. As a general criticism, it is surprising that none of the trials investigating yield responses to N fertiliser stated the basal soil N status and their potential for N mineralisation. Miles (1991) has shown that differences in the magnitude of the growth response to applied N can be due to this factor.
Kikuyu grass grows well on soils with pH levels above 4.8 (Awad et al. 1976). High applications of ammonium sulphate or nitrate may reduce soil pH by reducing the concentration of exchangeable bases with the result that soluble aluminium (Al) and/or manganese (Mn) concentrations can increase to toxic levels (Teitzel et al. 1991). This phenomenon can explain the reduced efficiency of N sometimes seen at moderate-high levels of N fertiliser applications (Awad and Edwards 1977). Although few critical reports exist on the effects of lime on kikuyu DM yields, liming is common practice to balance the acidifying effects of high N fertilisation (Awad and Edwards 1977; Miles et al. 1985; Miles 1991). Awad and Edwards (1977) found substantial increases in DM yields (around 30%) when pastures received 3.7 t lime at N applications of 672 kg/ha/yr. Nutrient uptake depends on rooting density and pH; therefore interactions between N fertilisation, lime and phosphorus uptake are sometimes observed (Miles et al. 1985). The mineral composition of kikuyu grass can also be affected by these interactions (Awad et al. 1979; Miles 1991) (see section on minerals).
Even when soil pH and other nutrients are adequate, it is common for the response to applied N to decline with increasing application rate, and this has important implications when optimal use of N fertilisers is sought. Whitney and Tamimi (1974), Cross (1979a) and Anon. (1987) found reduced growth responses at applications higher than 360 kg/ha/yr N, which agree with the results obtained by Mears and Humphreys (1974a) above 334 kg/ha/yr N, and Miles (1991) at around 400 kg/ha/yr N. These results might explain the low efficiency of response obtained by Kemp (1976) at higher N levels, and suggest that the efficiency of growth response of kikuyu grass to applied N decreases at around 350-400 kg/ha/yr N, depending on environmental (i.e. temperature, rainfall) or management factors (i.e. soil N supply, previous grazing). This would imply that there is a physiological limit to the utilisation of N by the grass crop which is dependent on its concentration in the plants' tissues and their potential growth rate. However, this does not imply that these are the optimum levels of N fertilisation for kikuyu pastures. The optimum levels will depend, not only on the growth responses of the pasture, but on the interactions with defoliation pressure, animal performance and the economic, environmental and productive objectives of the farming enterprise.
Differences between sources of N. Differences in yield response to N fertiliser type have been reported. Gartner (1966), in Australia, found that urea produced lower DM yields at rates of 225 and 450 kg/ha/yr N than ammonium sulphate, ammonium nitrate or sodium nitrate. His results were confirmed by Whitney and Tamimi (1974) in Hawaii, who applied N at 112-336 kg/ha/yr. Responses with urea were 29-90% of those with ammonium sulphate. Volatilisation of ammonia, especially at high application rates, is a well known problem with urea and this was probably the cause of reduced responses. Murtagh (1975) and Cross (1979a) found similar results comparing urea with ammonium nitrate and limestone ammonium nitrate, respectively. However, Castillo et al. (1983) found no differences in DM yields when they repeated the Whitney and Tamimi (1974) trial in Costa Rica at rates between 125-500 kg/ha/yr N. The different results between studies could be due to the cutting intervals, the amount of fertiliser used per dressing or environmental conditions (in particular soil moisture status after application). Whitney and Tamimi (1974) applied the N in 4 dressings/yr, while Castillo et al. (1983) applied the same amount in 12 dressings/yr. In Whitney and Tamimi's (1974) experiment, the increased amount of fertiliser per dressing could have led to a higher rate of volatilisation. Minson et al. (1993) concluded that research required on N fertilisers should aim to improve their economics by reducing volatilisation and oxidation losses. They stated that modelling studies and new techniques to assess N transfer between the different N pools could be used as tools to develop management strategies that reduce N losses. Urease inhibitors could be used to reduce volatilisation losses (Russell 1973), but their cost might not justify a wide on-farm use.
South African results (Hefer and Tainton 1990) suggest that a liquid source of N (urea ammonium nitrate, UAN) could be cheaper but that its effective use would depend on using it with low N concentrations to prevent leaf scorching. More recently, Hefer et al. (1992) found that applications of up to 69 kg/ha N as UAN (8% N) combined with 5% ammonium thiosulphate as a urease inhibitor would not damage the leaves of kikuyu. However, this level of N application is too low to maintain highly productive kikuyu pastures.
Sward structure. The effect of N fertiliser on the proportion of leaf and stem of kikuyu DM yield has been studied by Minson (1973) and Mears and Humphreys (1974a). These authors found that the proportion of stolons in the DM increased slightly with higher N applications. For example, Mears and Humphreys (1974a) report proportions of stolon in DM yield of 49% without N applied, to 54% for N applications of 672 kg/ha/yr. These were possibly due to increases in leaf senescence at higher N levels caused by faster development of the canopy which decreased the light interception of the lower leaves. Management decisions such as frequency, pressure or height of defoliation and their interactions with N fertilisation play a more important role in determining the sward's structure.
Gibberelic acid promotes an elongation of the stems (Lester et al. 1972). Although sward height increases (Whitney 1974a) and higher stem DM yields are obtained, these are associated with lower grazing intakes (Whitney 1976) due to the lower quality of the stem material.

Phosphorus (P) fertilisers. Mears (1970) concluded that little work had been done on the phosphorus requirements of kikuyu grass and the picture has hardly changed since then. In pot experiments, Wilson and Sandland (1976) found increased DM yields when applying 125-1072 kg/ha superphosphate at N fertiliser levels of 17.8-480 kg/ha/yr. Under normal conditions, the additive response of P fertilisers to increased kikuyu DM yields is small when N is applied (Cross 1979a; Annessens 1989; Miles 1991) but greater responses have been observed at higher N fertiliser levels (Cook and Mulder 1984b).
As with other plants, high responses to P fertilisers may be observed in situations where soil P is extremely deficient (Mears 1970). As soil P is relatively immobile, root length and distribution play an important role in its absorption (Wilson and Sandland 1976). Most of the kikuyu grass roots are in the top 70 cm of the soil (Quinlan et al. 1975) and its root mass can be as high as its above-ground DM (Mears and Humphreys 1974a). This would explain high P absorption rates when soils are P deficient and P fertilisers are applied.
Beneficial but marginal effects of P fertilisation on kikuyu DM yields are also observed on acid soils due to the removal of Al toxicity (Awad and Edwards 1977). Although P is more readily available at soil pH 5.5-7.5 (Buckman and Brady 1969), evidence appears equivocal in the studies where lime and P have been applied together. Awad and Edwards (1977) found no increases in kikuyu DM yields when applying both fertilisers. However, they concluded that, in their trial, P per se was not limiting growth of kikuyu, thus explaining the lack of response to additional P. Miles et al. (1985) attributed the increases in DM yield to an increased P uptake when lime was applied. Although they did not relate the results to soil pH or Al toxicity, the increases in soil pH observed in their study would have resulted in a higher availability of P. Nevertheless, P nutrition of kikuyu pastures and its interaction with lime and N should be the subject of further studies.

Potassium (K) and sodium (Na) fertilisers. Relationships involving Na and K cannot be separated as their roles are closely linked. A good explanation of the interactions between the 2 minerals is given by Smith et al. (1980) who postulated that kikuyu grass is a natrophobe plant species. This means that kikuyu does not require Na as a nutrient and that there is some biochemical or physical barrier to the movement of Na into the transpiration stream. Leaves therefore have a low Na:K ratio. This would provide a physiological explanation for the low Na concentrations in kikuyu grass and would explain why Russell (1976) found that kikuyu was tolerant of high soil salt concentrations.
Yield responses to K applications have been observed only when N supply is non-limiting (Cross 1979a; Miles 1991). This may partly be because the trials have taken place on K-rich clay soils or more likely because of the efficient recycling of this mineral through livestock urine and faeces (Cross 1979a). Recent evidence from pot experiments suggests that increases in soil K have positive effects on plant size and tillering rate and may also reduce the rate of senescence of older leaves (Pinkerton and Randall 1993), but the major effect of K fertilisers on kikuyu grass is to change its mineral composition (Miles 1991; Pinkerton and Randall 1993). Luxury uptake of K can occur at high rates of K application (see section on mineral composition).

Other nutrients. Apart from the observations cited by Mears (1970), few responses to other nutrients have been reported, although it is rarely clear whether this is because the plant requirement is low or because the soil can provide an adequate supply. Lipsett (1975) found no response in DM yields to additional molybdenum (Mo) but stated that the soil might have had an adequate Mo status. Provided that N, P and K are supplied in adequate amounts and the soil pH is maintained above the critical level to prevent Al or Mn toxicities and to maintain high concentrations of exchangeable bases, it is unlikely that DM yields will be depressed by low soil micronutrient concentrations in the regions where kikuyu grows. Micronutrients, especially Mo, together with the macrominerals P and sulphur (S), are more important when kikuyu is associated with legumes (Fulkerson et al. 1993).

Responses to light

Despite the fact that photosynthesis is one of the major physiological processes controlling DM yields of higher plants, very little work has been carried out on the response of photosynthesis to environmental factors in kikuyu grass (G. J. Murtagh, personal communication).
Lester et al. (1972) estimated apparent photosynthesis of whole micro-swards but their results are difficult to interpret as LAI was not reported. Weng (1988) estimated single leaf photosynthesis but unfortunately at only one qunatum flux density (1500 *mol/m2/s photosynthetically active radiation, PAR) which is below the point of light saturation.
We recently estimated light response curves for single leaves of well-watered kikuyu grass grown in a glasshouse at 20 oC, which is a representative temperature during the growing season in regions where kikuyu grows. Measurements were carried out at irradiance (I) (corrected for light source) ranging from 0-480 Watts/m2 PAR. A non-rectangular hyperbola (Johnson and Thornley 1984) was fitted to the data (r2 = 0.97), and values of 0.018, 2.3 and 0.869, were obtained for light utilisation efficiency (?, mg CO2/J), light saturated gross photosynthetic rate (PMax, mg CO2/m2/s), and the required dimensionless parameter (?), respectively.
An analysis of solar radiation records from Wollongbar, New South Wales, Australia (lat 28o 50' S)(Murtagh 1990b) and Poás, Costa Rica (lat 10o 01' N) fitted to the 20 oC light response curve (Figure 1) indicated that solar radiation is unlikely to be a major source of variation in photosynthesis between latitudes, because of the small difference in their seasonal irradiance ranges and the small slope of the light response curve as it reaches its asymptotic value at saturating light levels (PMax). This would explain why Murtagh (1988a) found no significant effect of solar radiation (300-3000 nm) on growth rate between 12 and 30 MJ/m2/day. This results have been recently confirmed by Herrero (1995) using a mechanistic kikuyu grass growth and utilisation model.
(Figure 1 about here)
When modelling canopy photosynthesis, the factors affecting LAI (rates of leaf appearance and extension and specific leaf area, (SLA)) and N utilisation play a more important role than photosynthetic rates at the single leaf level. Latitudinal effects such as seasonal temperature fluctuations play a major role.
Irradiance affects the SLA of grasses and legumes (Ludlow and Wilson 1971). In their study, light levels up to full sunlight decreased the SLA of Panicum maximum and Macroptilium atropurpureum and low light levels increased it. The average SLA of kikuyu leaves has been found to be between 21 and 24 m2/kg. (Murtagh 1988a; Herrero et al. unpublished). We found an average SLA of 23.2 m2/kg with values of 21.4, 23.4 and 25.2 m2/kg for growing, first fully expanded and older leaves respectively. This distribution reflects their position through the canopy as older, shaded leaves had higher SLA's. Leaf photosynthesis is commonly found to increase with leaf N concentration. Bolton and Brown (1980), for example, showed this relationship for Panicum maximum. Weng (1988) similarly found that high soluble N concentrations in kikuyu grass leaves were associated with higher photosynthetic rates and Ludlow et al. (1988) showed that the N concentration in kikuyu grass was positively correlated with its chlorophyll content. Their results are related to the high proportion of plant N (> 30%) present as photosynthetic enzymes, and might explain the observations of Henzell and Oxenham (1973) and Murtagh (1988a), that for optimal growth, the first fully expanded leaf of kikuyu should contain at least 3.5% N. The low growth responses at 365 kg N/ha and high irradiances found by Ericksen and Whitney's (1981) could be explained by this threshold not being reached.
The responses to light is one of the fundamental subjects that requires more attention if the physiology of kikuyu grass growth is to be better understood.

Temperature

Temperature is another key determinant of kikuyu grass productivity. It causes the seasonal growth cycle in the subtropics (Lambert et al. 1977; Goold 1979; Murtagh and Moore 1987; Piggot 1988; Rumball 1991; Minson et al. 1993), influences the responses to N fertilisation (Whitney 1974a; Murtagh 1975; Colman and O'Neill 1978; Cook and Mulder 1984a; Pearson et al. 1985) and has marked effects on its physiology.
The optimum 24-hour mean temperature in the field depend, both, on the shape of the response and on the diurnal temperature cycle. Working in growth chambers, Ivory and Whiteman (1978a) reported optimum day and night temperatures for maximum total plant growth of 29.4 and 25.6 oC, respectively. Murtagh et al. (1987) obtained maximum growth in the field at average temperatures of 25 oC.
Ivory and Whiteman (1978a) also suggested that differences existed in the temperatures maximising the growth rate of different plant parts. The optimal temperature for tillering rate was lower (mean 23.4 oC) than for total plant growth, and tillering was less responsive to temperature (except at extreme temperatures). Although leaf size increased with increasing day temperature up to 34.1 oC, it was relatively insensitive to low temperatures which might be explained by its origins (Kenyan highlands), and would confirm its adaptation to the subtropics and cooler regions of the tropics. They also found that kikuyu growth ceased at around 8 oC (Ivory and Whiteman 1978b) and this was confirmed by Colman and O'Neill (1978) in a field experiment. Murtagh et al. (1987) also found that internode thickening and extension of the primary stolons increased up to 30oC.
Leaf appearance rate is markedly affected by temperature and can be one of the most important factors explaining seasonal DM yield at different latitudes. Murtagh (1987) reported leaf appearance rates of 5.5, 3.8, 1.8 and 1.4 days at 15, 20, 25 and 30 oC, respectively. We found similar results, with rates of 5.3, 3.8, 1.9 and 1.4 days (Figure 2) at similar temperatures. Mannetje (1975) found that increasing temperature increased the leaf content of kikuyu grass swards. His results could be explained by an increased leaf appearance rate. These responses to temperature have a marked effect when scaling single-leaf gross photosynthesis up to the whole canopy.
(Figure 2 about here)
Ludlow and Wilson (1971) and Johnson and Thornley (1984) suggested that the temperature effects on single-leaf light-response curves are mediated through changes in the PMax and that there is a linear relationship between the two variables. Herrero (1995) studied this concept in kikuyu light response curves and found that there would be a decrease of 6-8% in the photosynthetic rate of single kikuyu leaves per 1 oC between 20 oC and a basal temperature of 8 oC at irradiances ranging from 140 to 280 W (PAR)/m2. This would equate to a loss of photosynthate of 0.15-0.25 mg CO2/m2/s per oC. This, together with the temperature effects on leaf appearance, may explain to a great extent the differences in DM yield at different latitudes or seasonal differences in the subtropics (Herrero 1995).
Respiration rates are also affected by temperature (Murtagh et al. 1987). In their miniature sward experiments, specific maintenance respiration rates increased from 11 mg/g/d at 15 oC to 37 mg/g/d at 30 oC.
Kikuyu has a good overwintering ability (Quinlan et al. 1975). It can withstand mild frosts (Mears 1970) and, in countries like South Africa, it is used as foggage (frosted herbage) (Zacharias et al. 1991). Ito et al. (1985) have related this overwintering ability to the accumulation of non-structural carbohydrates (NSC). During late summer and autumn, canopy net photosynthesis exceeds the demand for assimilate for structural tissues and high NSC concentrations accumulate in consequence. These NSCs are used as an energy reserve and to lower the freezing point to survive the winter temperatures but are depleted by the time spring arrives. As temperature and daylength increase in spring, cell division and expansion speed up, more photosynthate is produced and the grass re-establishes its growing pattern.

Soil water and irrigation

Yield responses to irrigation and nutrients by plants depend on the initial soil water status before application (Russell 1973). Mears (1970) concluded that kikuyu withstands dry periods especially if nutrients are not limiting and that without N fertilisation, increasing the frequency of irrigation hardly increased DM yields. Kemp (1975) confirmed these results and concluded that during short dry periods the additional yield from N applications was greater than the extra yield from irrigation alone. The high fertiliser rates used in Forde et al.'s (1976b) or the moderate N applications in Lowe and Bowdler's (1984) studies support these conclusions, as they could not find a significant yield difference between irrigated and unirrigated plots. However, the experimental design of these studies, which were carried out to solve site-specific problems, only allowed for short periods, if any, of mild water stress, and therefore did not compromise greatly the DM yield potential of the pastures. The initial soil water status of the unirrigated plots together with the low rainfall might have been enough to permit normal growth during most of the experimental periods. Murtagh (1975) suggested that an abrupt change from dry to wet conditions would enhance the response to N fertiliser relative to response of continuously wet plots, because of the high mineralisation rates of organic N when dry soils are watered (Russell 1973). Therefore, another explanation for the lack of response to additional irrigation could be that in the dryland plots of Forde et al. (1976b) and the low-frequency irrigation plots in Lowe and Bowdler's (1984) experiments, the sporadic rainfall caused high rates of soil N mineralisation, thus masking the effects of mild water stress.
Murtagh (1988b) demonstrated a relationship between soil water potential, evaporative demand and growth rate. At a peak growth rate of 234 kg DM/ha/d and an evapotranspiration rate of 2 mm/d, water stress began to reduce growth rate when the soil water potential was less than - 134 kPa, while at an evaporative demand of 5 mm/d, growth rate was only 39% of the peak growth rate and growth ceased at a soil water potential of -101 kPa.
Critical studies designed to understand the physiological response and the limits of kikuyu grass growth to water stress have not been performed, and are needed to plan optimal irrigation schemes. These experiments should attempt to find the lowest soil water deficit required before water stress limits growth rate, rather than to test the effect of additional water on DM yield.

Effect of defoliation intervals

Cutting trials (Whitney 1974b; Kemp 1976; Soto et al. 1980; Rethman and de Witt 1988) have demonstrated that increasing the cutting interval increases annual DM production and decreases N content of kikuyu grass over a wide range of N fertiliser application rates. The physiological explanation for the increased production is an increased LAI and therefore a higher light interception from the sward as it matures.
Interactions between N fertilisation level and cutting interval are usually found in these trials. An improved N utilisation at older stages of growth has been suggested (Soto et al. 1980), especially at higher N rates. These results are consistent with the positive relationship between chlorophyll and N levels found by Whitney (1974b) and Ludlow et al. (1988). The improved N utilisation might also be masked by supracritical N concentrations (Murtagh 1990a) which would buffer the effect of a N deficiency. The lower N concentration in kikuyu as it matures is partly explained by the translocation of surplus N from the aerial parts of the plant to the roots to counteract the low soil N concentrations at longer periods of growth.
From a grazing management point of view, the balance between growth, especially of leaves, senescence, and sward composition is very important as their relative proportions will markedly influence the sward's nutritional value and utilisation (Stobbs 1975). Unfortunately, none of the trials mentioned above reports the sward composition at different stages of growth. As LAI increases, less light reaches the lower horizons of the sward, the lower leaves die (Murtagh 1987) and consequently more senescent material accumulates. The final effect is an increase of stolon relative to leaf yields, a reduction in nutritive value; and either an underutilisation of the pasture, or a reduction in animal performance.
Physiological concepts should be given more consideration in cutting trials or grazing studies to find the appropriate management strategies that produce the required sward composition, cutting intervals or paddock rest periods.

Associations of kikuyu with other grasses and legumes

Subtropical regions. In the subtropics, where marked seasonal variations in temperature and moisture define the growth habits of plant species, it is a common practice to use kikuyu in association with temperate grasses and legumes to balance the feed availability throughout the year for ruminants (Betteridge 1979; Goold 1979; Davies and Hunt 1983; Piggot and Morgan 1984; Hill 1985; Betteridge and Haynes 1986; Piggot 1988; Harris and Bartholomew 1991) or to devote an area of the farm for the production of temperate species (Murtagh and Moore 1987). The seasonality of production between kikuyu and temperate species in subtropical Australia is demonstrated in Figure 3, adapted from Murtagh and Moore (1987).
(Figure 3 about here)
Kikuyu has peak growth rates in summer and autumn (Murtagh and Moore 1987), where temperatures are favourable for its growth. It is usually associated in winter-spring with temperate grasses such as: Lolium spp. (Lambert et al. 1979; Betteridge 1979; Murtagh and Moore 1987; Harris and Bartholomew 1991), which are the predominant temperate species; Oryza sativa (Minson et al. 1993); Poa spp. (Goold 1979; Piggot 1991); Festuca arundinacea (Hill et al. 1985); Bromus spp. (Hill 1985; Hill et al. 1985; Betteridge and Haynes 1986); and/or with legumes, the most common being clovers (Trifolium spp.) (Davison 1985; Fulkerson et al. 1993) and Lupinus spp. (Royal and Hughes 1976). Fulkerson and Slack (1993) suggested that the legume, Lotus penduculatus, was a promising association. However, more recently, Fulkerson and Slack (1994) found that kikuyu associations with white clover produced higher DM yields (21% higher), their seasonal pattern of growth complemented kikuyu better, and the swards had a higher nutritive value than those with Lotus spp. They also found that a severe defoliation down to 5 cm, coupled with a flexible defoliation interval depending on the season (longer regrowth periods in winter and shorter ones in late spring), maintained a balanced association between clover and kikuyu. However, whether associations with Lotus spp. are more suitable in other climatic regimes where kikuyu grows is still not known. Tropical legumes such as Desmodium spp. (Jeffery 1971b; Brands and Cook 1976; Woomer et al. 1990) and Glycine spp. (Colman et al. 1966; Jeffery 1971b; Quinlan et al. 1975) could be used in association with kikuyu for summer feeding.
The high growth rates of kikuyu during summer and autumn and its aggressive nature (Russell and Kleinschmidt 1984; Rumball 1991), often result in the need to renovate the temperate species in mixed swards in order to obtain enough food for the winter-spring period. The preferred methods of re-introduction are oversowing (Betteridge 1985; Harris and Bartholomew 1991; Fulkerson et al. 1993) or direct drilling (Hill 1985; Minson et al. 1993), which are usually carried out in autumn (Fulkerson et al. 1993).
Pre-sowing management usually consists of slashing (Minson et al. 1993), hard grazing or mowing the kikuyu pastures to reduce their competitiveness for light and to allow more temperate species seed to reach the ground (Piggot and Morgan 1984; Betteridge and Haynes 1986). Glyphosate ('Roundup')(Hill 1985; Piggot 1991; Fulkerson and Slack 1993) and paraquat (Hill 1985; Betteridge and Haynes 1986) are also used to control over-dominance of kikuyu prior to temperate species renovation. Since these activities are carried out at a time of the year when the NSC reserves in the stem bases and stolons are high, they do not compromise the ability of kikuyu to regrow when conditions are favourable.
Seeding rates for Lolium spp. range from 15-50 kg/ha (Davies and Hunt 1983; Betteridge and Haynes 1986; Harris and Bartholomew 1991; Fulkerson et al. 1993), whereas Trifolium spp. are sown at rates 3-20 kg seed/ha, depending on the timing of the forage required and whether they are to be grown in associations with kikuyu pastures or alone as a winter-spring feed (see Fulkerson et al. 1993). Mixed pastures have a higher nutritional value in the winter and spring, but a delicate balance exists to match feed availability throughout the year (Lambert et al. 1979; Rumball and Boyd 1980). Bloat control is necessary when legumes represent a high proportion of the pastures on offer (Battese and Fulkerson 1994) in winter-spring.
It is widely recommended that kikuyu associations with Trifolium spp. be fertilised with superphosphate at 100-250 kg/ha plus 100 kg/ha muriate of potash (Colman et al. 1966; Fulkerson et al. 1993). Lime should be applied if the pH is below 5.5 (Battese and Fulkerson 1994).
Philpotts (1981) noted that poor nodulation of glycine (Neonotonia wightii) and forage lupins may occur in krasnozem soils where kikuyu grows, and suggested that rhizobia were suppressed by extracts of kikuyu roots. The subject needs further research as soil type is not the only possible cause for reduced nodulation. Chou et al. (1987; 1989) studied allelopathic interactions in a pasture-forest system in Taiwan, and found that an aqueous leachate of kikuyu promoted the growth of 2 weed species. On the other hand, Woomer et al. (1990) suggested that the mat formed in kikuyu pastures was responsible for the reduced abundance of rhizobia associated with Trifolium repens but not of rhizobia associated with Desmodium intortum.
Tropical highland regions. In tropical highland regions, seasonal temperature cycles are not as marked as in the subtropics, and moisture availability largely determines the growth pattern of kikuyu grass. Under these conditions, associations, mainly with Trifolium spp., are sometimes found (Mears 1970) but are very difficult to manage due to the mild temperatures, which promote high growth rates of N-fertilised kikuyu grass under the favourable moisture conditions. Stocking rates are seldomly adjusted to counteract these high growth rates, with the result that a dense mat of stolons is usually formed which inhibits light penetration and prevents the legume from establishing properly (Mears 1970; Fulkerson et al. 1993). Therefore, the legume component is usually lost or makes only a small contribution to the sward's yield in long established kikuyu pastures in the tropics (van der Grinten et al. 1992).
Very little information has been published regarding management of kikuyu-legume associations in tropical regions. More research is required in these areas to reduce N fertiliser and supplementation costs. Legume species with different morphological characteristics which enable them to tolerate better the aggressive growth habits of kikuyu grass should be tested (Woomer et al. 1990).

Nutritional value of kikuyu grass

Crude protein (CP)

Crude protein (N X 6.25) concentrations ranging from 74 to 282 g/kg DM have been reported in fresh kikuyu grass (Table 1) and from 120 to 164 g/kg DM in kikuyu silage (de Figuereido et al. 1989; de Figuereido 1991).
CP concentration increases with increasing N fertilisation (Minson 1973; Whitney 1974b; Castillo et al. 1983; Pearson et al. 1985; Rumball 1991), and decreases at very high day-night temperature regimes (32/24 oC) (Mannetje 1975). In the subtropics, CP initially increases during the summer (Pearson et al. 1985) probably because of the fertilisation regime, but as the growing season advances, there is a diluting effect on CP concentration caused in part by the high growth rates (Murtagh 1975; Minson 1990), and partly by changes in the structural composition of the sward. Laredo and Minson (1973) found that leaves had a higher CP than stolons (126 vs 110 g/kg DM). Similar results have been obtained by Mannetje (1975) and Marais (1990a; 1990b). Forde et al. (1976a) also reported differences between leaf lamina and sheath, with the former having a higher CP concentration (214 vs 175 g/kg DM). Dead material has a low CP concentration (< 80 g/kg DM). Thus the higher the proportion of leaves relative to stolons and dead material, the higher the crude protein concentration of kikuyu grass.
(Table 1 about here)
A particular feature of kikuyu when compared with other tropical grasses is that it can maintain a relatively high CP level (> 100 g/kg DM) when mature (Mears 1970). This is at least partly due to the essentially vegetative status of the sward. Reid et al. (1979), in Uganda, found a CP concentration of 113 g/kg DM in kikuyu after 12 weeks regrowth. Soto et al. (1980) confirmed this finding in Colombia with a crude protein concentration of 131 g/kg DM after 11 weeks regrowth. Even after a 6-month dry season in Tanzania, Rukantabula and Kusekwa (1990) report CP levels of 74.4 g/kg DM, which is close to the level at which dietary protein starts to limit intake and therefore animal performance (Milford and Minson 1965; Minson 1981). Lower concentrations are sometimes observed in the subtropics during the winter (Jeffery 1971a; Zacharias et al. 1991) and this is often related to lower leaf:stem ratios due to lower leaf appearance and extension rates (Buxton and Fales 1994). As stated by Mears (1970), it is unlikely that animals will experience protein deficiency in kikuyu grazing systems. The CP concentrations of kikuyu grass can be predicted by near-infrared reflectance spectroscopy with accuracy (r2=0.94, s.e.=11.4 g/kg DM)(Herrero et al. 1996).
Recent Australian studies (Reeves et al. 1994), have concentrated on finding defoliation intervals which produce the best compromise between forage availability and quality using morphological indicators. Preliminary results suggest that during early summer, the N concentration in leaves declines after 26 days of regrowth. This would be explained by the increasing temperature regimes in this season and an increased utilisation of leaf N to produce more structural tissue. More studies of this type are needed to find optimal defoliation intervals for particular production purposes and climatic regimes.
Current feeding systems (i.e. ARC 1984; SCA 1990) assess the protein needs of ruminants in terms of requirements for microbial protein synthesis, rumen undegradable (UDP) and rumen degradable protein (RDP). The synthesis of microbial protein in the rumen is largely dependent on the amount of energy and RDP available to the animal (SCA 1990).
Hart and Leibholz (1990) compared kikuyu grass samples containing CP and acid detergent fibre (ADF) concentrations of 130 and 294 g/kg DM, respectively, with others containing 62 g/kg DM CP and 381 g/kg DM ADF. They discovered that the true ruminal protein degradability of the samples was similar at 0.80 and 0.79, respectively. Mean rumen retention times (MRT) were 39.8 and 36.7 h for the two samples. Microbial N flow to the omasum was 25 and 14 g/kg DOM. In another set of studies, Punia et al. (1984) and Punia and Leibholz (1994) found that microbial N flow to the omasum increased with increasing level of intake. Punia and Leibholz (1994) found microbial N yields of 34, 24.1, and 20.7 g/d for kikuyu hay intake levels of 60.7, 50, and 39 g/kg metabolic weight (BW0.75). These microbial N yields represent 13-16 g/kg DOM which are low when compared with published values (ARC 1984). Low rumen ammonia concentrations or low rates of carbohydrate fermentation might have affected microbial production in their studies. Protozoal N accounted for 26.1-29% of the microbial N and protozoal numbers in rumen liquor and omasum were 1.09X10-5/ml and 0.51X10-5/ml, respectively.
Pheloung and Brady (1979) found that the solubility of CP from kikuyu grass leaves containing 181 g/kg DM CP was 26%. Similar results (24%) were obtained by Ali and Stobbs (1980). They also suggested that CP solubility was unaffected by stage of growth and that stem protein was considerably more soluble (66%) than leaf protein. The findings of Marais et al. (1987) and Marais (1990a; 1990b) support this conclusion. In their studies the stem fraction contained a high level of non-protein organic N (NPON) which is highly soluble and very low protein N, while the leaves contained very high protein N and low NPON. Fernando and Jayaratne (1980) reported mean values ranging from 24.5 to 30.2% for composite plant samples, probably reflecting a high proportion of leaf material.
Pheloung and Brady (1979) found that 13% of the soluble protein in kikuyu grass leaves was in the form of fraction 1 protein, which is the major soluble protein of the chloroplast and which has been implicated in the occurrence of bloat. However, Pienaar et al. (1993a; 1993b) attributed observations of high in vitro foaming capacity of kikuyu samples to saponins.
Accepting the importance of microbial protein in regulating protein requirements of ruminants, substantially more work is needed on the kinetic properties of kikuyu grass protein.

Fibre fractions

Mears (1970) did not include the fibre composition in his review, but this aspect has been the subject of considerable research in recent years. The importance of the fibre fractions lies in their association with digestibility, rumen fill and intake (Minson 1982, 1990; van Soest 1982).
Crude fibre (CF). When expressed as CF, levels range from 167-314 g/kg DM (Table 1), but this measurement has largely been superseded by neutral detergent fibre (NDF), acid detergent fibre (ADF) and lignin analyses which give a better description of the fibre composition (see van Soest 1982).
Neutral detergent fibre (NDF). Reported levels of NDF (cell wall) in composite samples range from 474-827 g/kg DM (Table 1), but differences exist between plant parts due in part to anatomical reasons (see Wilson 1991; 1994), leaves having a lower NDF concentration than stems (Laredo and Minson 1973; Moir et al. 1979; Marais et al. 1992). Laredo and Minson (1973) found mean NDF concentrations of 680 and 706 g/kg DM in leaves and stems, while Moir et al. (1979) and Marais et al. (1992) found higher differences (532 vs 687 and 476 vs 578 g/kg DM, respectively). These higher differences were possibily caused by a more mature stolon fraction (Minson 1990). Similarly, Bailey and Hunt (1973) reported lower hemicellulose levels in leaf blades than in sheaths.
In terms of its rumen degradability, Köster et al. (1992) working on fistulated sheep with samples of oesophageal extrusa containing 628 g NDF/kg DM and 251 g CP/kg DM found that 73.9% of the NDF disappeared from the rumen after 48 h at a rate of 2.97% per h. In another experiment, Singh et al. (1992) found that the degradation of kikuyu NDF in the rumen of steers exhibited an initial lag phase of 4.34 h. Their samples (135.1 g CP and 713 g NDF/kg DM) had a degradation rate of 4.19% per hr. Herrero et al. (1995), using gas production measurements as a proxy for NDF loss, found seasonal differences in the rate of degradation of NDF but not on its extent. Their samples, collected under tropical highland conditions, had a significantly lower NDF degradation rate during the mild dry season than during the wet season (3.68% vs 4.68%, respectively). The lower degradation rate found by Köster et al. (1992) in comparison with the previous studies was possibly caused by the short incubation times used (maximum 48 h), therefore failing to define the asymptote of the degradation curve and the lag phase.
The NDF fraction of kikuyu grass has been predicted by NIRS with accuracy (r2=0.88, s.e.=15.9 g/kg DM)(Herrero et al. 1996).
Acid detergent fibre (ADF). ADF levels in composite samples are about half of the NDF concentrations and range from 246-402 g/kg DM (Table 1). As was the case with NDF, Laredo and Minson (1973) found that leaves also have a lower ADF concentration than stems (323 vs 360 g/kg DM, respectively).
Singh et al. (1992) found that ADF degradation in the rumen also exhibited a lag phase of 7.09 h and had a rate of disappearance of 3.81% per h for a sample of kikuyu containing 402 g ADF and 135 g CP/kg DM.
Lignin. Lignin levels in composite samples of kikuyu grass range from 24-88 g/kg DM (Table 1), with levels in leaves lower than in stolons (Laredo and Minson 1973; Marais et al. 1992).
Several factors affect the concentration of these fibre components in kikuyu. Firstly, the proportion of cell wall in kikuyu leaves has been found to increase with temperature (Wilson et al. 1976; Moir et al. 1977). This is partly a reflection of the higher growth rates at increased temperatures. Secondly, Laredo and Mendoza (1982) found higher concentrations of all fibre components during the dry season in Colombia, suggesting that low moisture also plays an important role. It has long been recognised that grasses grown in dry conditions exhibit altered anatomical features such as an increase in sclerenchymatous tissues (e.g. Grace and Russell 1977). Thirdly, Reid et al. (1979) found that the fibre components increase as kikuyu matures, but as with protein, the rate of change is slow, thus maintaining low fibre concentrations. In their study, 4- and 12-week kikuyu regrowths contained 584 and 688 g NDF/kg DM while lignin increased from 35 to 53 g/kg DM. However, the stolon fraction matures faster that the leaf fraction (Minson 1990; Wilson 1994).

Non- structural carbohydrates

Non-structural (soluble) carbohydrates (NSCs) are an important energy source for rumen microbes to ensure a proper utilisation of protein (Preston and Leng 1987).
Several authors have published NSC levels of composite kikuyu samples ranging from 21.7-156 g/kg DM (Table 1). The large variation between samples is typical of the measurement of NSC. As NSCs are the product of photosynthesis, which is temperature and irradiance dependent (see section on responses to light), marked fluctuations in their levels occur throughout the day (Marais and Figenschou 1990) and between seasons (Ito et al. 1985). Marais and Figenschou (1990) observed variations of almost 38% in NSC concentrations during a single day, levels typically peaking during the afternoon when the grass had been exposed to sunlight for several hours.
The stems of graminaceous plants act as storage organs (Milthorpe and Davidson 1966). Marais and Figenschou (1990) confirmed this finding in kikuyu and noted that stems accumulated higher concentrations of NSC's than leaves. However, in the study of Taylor et al. (1976a), leaves contained slightly higher NSC levels than stems (92 vs 82 g/kg DM). NSC concentrations are reduced by high temperatures (Marais and Figenschou 1990) and this is a possible cause for the differences between authors. Perhaps in Taylor et al.'s (1976a) experiment, the temperature regimes were higher, therefore the NSCs stored in the stems would have been used as an energy source for increased respiration at high temperatures (see section on temperature). In Marais and Figenschou (1990) study, leaves and stems of plants grown at day/night temperatures of 19/9 oC and 31/16 oC contained 79 and 83, and 65 and 71 g NSC/kg DM. Sheaths typically contain less NSCs than leaves (Forde et al. 1976a). In kikuyu silages, de Figuereido and Marais (1994) found NSC concentrations of 23 g/kg DM after 120 days of fermentation.
The NSC of kikuyu grass consist mostly of sucrose although differences between leaves and stems are associated with increases in the fructose concentrations of the latter (Marais and Figenschou 1990). Starch contents are very low (< 10 g/kg DM) (Taylor et al. 1976a).
The NSC levels reported by most authors are low when compared with values for other species and therefore confirm Betteridge's (1979) observations that kikuyu pastures have a low soluble:structural carbohydrate ratio. However, it has not been possible to separate the effects of species and climate. These observations require further study as low NSC levels, coupled with low rates of degradation of structural carbohydrate, could be the cause of the low N retention and animal performance (see section on animal performance). Rumen microbes would not have had the required energy source to utilise the high N levels in kikuyu grass, much of which would have been lost as ammonia. Therefore, the often reported high N concentration would be irrelevant if it could not be properly utilised. This also suggests that the balance between the chemical fractions coupled with a knowledge of their kinetic properties should be more important in nutritional studies than looking for high or low concentrations of specific nutritional components.

Digestibility

Digestibility of the dry matter (DMD) and the chemical fractions is one of the subjects most studied in kikuyu grass due to their direct relation to the metabolisable energy concentration and forage intake (Minson 1982). Results obtained by several authors are presented in Table 1. When estimated in vitro, digestible DM values range from 500-834 g/kg DM. Ishizaki et al. (1976) and Hacker and Minson (1981) found that in vitro measurements overestimated DMD in kikuyu grass when compared with in vivo trials. In vivo estimations in sheep ranged from 473 to 686 g/kg DM. In vitro digestible DM of kikuyu silages has been found to be between 350-450 g/kgDM (de Figuereido et al. 1990; de Figuereido 1991).
In terms of plant parts, digestibility of stolons is similar to or slightly higher than that of the leaf fractions (Laredo and Minson 1973, 1975; Mannetje 1975). Hacker and Minson (1981) summarised the results of 21 in vivo digestibility trials with sheep and found that DMD of leaves and stems was 510 and 520 g/kg DM, respectively. Contrasting results were found by Taylor et al. (1976a) and Reid and Stevenson (1983), who found that leaves were between 7.5-8% more digestible than the stolons. Differences between these studies are possibly due to the age of the regrowth. Minson (1990) suggested that, at an immature stage, there are no differences in DMD between leaves and stems, but as the plant matures, the stem becomes less digestible. Wilson (1994) concluded that this was caused by an increased proportion of thick-walled cells and a more rapid lignification of the stem of tropical pastures when compared to leaves. Differences have also been found between leaves and sheaths (Forde et al. 1976a) with the latter being more digestible (661 vs 705 g/kg DM). Hacker and Minson (1981) suggested that leaves of grasses at the top of the canopy were more digestible than leaves at lower strata. In kikuyu this has been confirmed by Reid and Stevenson (1983) who found that the upper, younger leaves were slightly more digestible in vitro than side leaves (739 vs 713 g/kg DM).
High temperature decreases the DMD of kikuyu grass (Mannetje 1975; Wilson et al. 1976; Wilson 1994) and these effects are mediated through changes in the fibre composition (Wilson et al. 1976; Moir et al. 1977; Wilson 1994). Stage of growth similarly affects digestibility due to changes in the proportions of leaves, sheaths, stems and senescent material and their respective changes in digestibility. Minson (1990) estimated that DMD of kikuyu decreased at a rate of 0.2%/d with a range of 0.18-0.22%/d. N fertiliser rate appears to have no effect on DMD of kikuyu (Minson 1973).
Differences also exist between the digestibility of the different chemical fractions (Minson 1982). The digestibility of CP has been found to vary between 48.6-65.8% (Jeffery 1971a; Soto et al. 1980). NDF digestibility ranges from 44.9-69.0% (Moir et al. 1977; Soto et al. 1980) whereas ADF digestibility has been found to be 28.8-48.7% (Soto et al. 1980). The ranges in the fibre fractions depend mostly on the degree of lignification as the plant matures (Minson 1982; Wilson 1994).
Reported gross energy in kikuyu ranges from 16.9 to 20.45 MJ/kg DM (Campbell et al. 1969; Betteridge 1979; Soto et al. 1980; Bredon et al. 1987; Marais et al. 1990) but, for practical purposes, the well established value of 18.4 MJ/kg DM can be used. The DMD and the digestibility of gross energy in grasses are highly related (Minson 1981) and Minson and Milford (1966) and Jeffery (1971b) found it to be 2-4% lower than that of the DM. Metabolisable energy (ME) concentrations can be estimated as 0.81 X digestible energy (Minson 1981).

In vitro gas production

In vitro gas production measurements are gaining popularity as methods to characterise the nutritive value of forages due to their relationship with ME concentrations (Menke and Steingass 1988), and intake and digestibility (Khazaal et al. 1993, Kibon and Ørskov 1993). They can also provide kinetic parameters for digestion modelling studies (Pell and Schofield 1993, Jessop and Herrero 1996), and if gas production from fermentation of soluble carbohydrates is taken into consideration, the corrected gas volumes can be used to predict NDF disappearance (Herrero and Jessop 1996).
Herrero et al. (1995) estimated seasonal ME concentrations in kikuyu grass from gas production and obtained values ranging from 6.52 MJ/kg DM during the dry season to 7.52 MJ/kg DM during the wet season. The gas production dynamics of a range of kikuyu grass samples are presented in Table 2 (Herrero et al. 1996).
(Table 2 about here)

Mineral composition

Macro-minerals. Results from several authors are summarised in Table 3. A comparison of the mean macro-mineral concentrations of kikuyu grass against recently published mineral concentrations required in feeding stuffs for ruminants and horses suggests that:
(Table 3 about here)
* If the total Ca concentrations in kikuyu are compared with required dietary Ca concentrations, it seems that these are adequate for dairy and beef cattle, lambs and ewes (Minson 1990; SCA 1990). The only exceptions are very young animals with high growth rates (i.e. 200 kg beef cattle gaining > 1.0 kg/d; lambs gaining > 0.3 kg/d) and ewes producing 2-3 kg/milk/d. The total Ca concentrations in kikuyu grass seem to sustain maximum milk production levels for Jersey and Friesian cows of 14 and 20 kg/d, respectively. However, it is unlikely that cows eating kikuyu grass as their sole diet will produce such milk yields (see below). Growing horses or pregnant and lactating mares need to be supplemented with a Ca source as their Ca requirements would not be met (McDowell 1992) from kikuyu grass pastures as their sole diet. In subtropical regions, kikuyu pastures can be Ca deficient during winter and spring (Fulkerson et al. 1993).
Two factors make difficult a more detailed analysis of the sufficiency of the Ca concentrations in kikuyu grass. First, the true Ca requirements of ruminants and the adequacy of dietary Ca concentrations are difficult to quantify, since ruminants can use bone Ca to reduce the effects of a deficiency and dietary Ca availability is partly dependent on the level of production (Minson 1990). Second, the availability of dietary Ca can be reduced if it is bound as insoluble Ca oxalate crystals (Minson 1990), and a high proportion of Ca in kikuyu can be present in this form and may not be available to livestock (Blaney et al. 1981; Marais 1990a). Therefore, the real contribution of the Ca concentrations in kikuyu grass towards the Ca requirements of grazing livestock is considered marginal and still requires more research. It is considered that supplementation of this mineral should be given to livestock on kikuyu pastures to obtain the desired levels of production (Kayser 1975; Fulkerson et al. 1993). For example, Kaiser (1975) found increased growth rates of beef cattle grazing kikuyu when they were supplemented with a Ca source.
* The P concentrations of kikuyu grass are usually adequate to maintain high levels of animal production from ruminants and horses (SCA 1990). However, low P concentrations in kikuyu can be found in periods of rapid growth or in soils with low available P (Minson 1990).
* The concentrations of Mg, S and Cl are sufficient to maintain high levels of animal production from beef and dairy cattle, sheep and horses (Minson 1990; McDowell 1992).
* Kikuyu grass is low in Na (Sherrell 1978; Pastrana et al. 1990; Miles 1991), and beef and dairy cattle, sheep and horses grazing kikuyu need to be supplemented with a Na source. It can contain very high concentrations of K (Miles 1991; Pinkerton and Randall 1993), but under practical situations, this should only present problems for animal production if levels exceed 3.0% (McDowell 1992). As is the case with temperate pastures, excessive K uptake decreases the Ca, Mg, and Na concentrations in kikuyu and may be responsible for grass tetany (Awad et al. 1979; Miles 1991). Marais et al. (1987) suggested that the high K concentrations coupled with high rates of N fertilisation also promote high nitrate accumulation in kikuyu grass.
N fertilisation affects the mineral composition of kikuyu grass (Awad et al. 1976; Pearson et al. 1985). Pearson et al. (1985) found that increasing the rate of applied N decreased P concentrations and increased the Ca and Mg concentrations. Awad et al. (1976) found that K levels were also increased at high N applications but Mg and Mn concentrations decreased. The variation in results between studies may be related to soil pH differences as Awad et al. (1976) were working with acid soils. Awad and Edwards (1977) found that when lime was applied to suppress the acidifying effects of N fertilisation, herbage Ca, Mo and P increased while Mn decreased.
Micro-minerals. Micro-mineral concentrations in kikuyu grass (Table 4) indicate adequate concentrations of Mn, Cu, Co and Fe to meet ruminant and horse requirements (SCA 1990; McDowell 1992). Zn concentrations are within the range proposed by Minson (1990) and SCA (1990) for cattle and sheep but are slightly low for lactating dairy cattle and horses (McDowell 1992).
(Table 4 about here)
The micronutrient concentrations in kikuyu grass can vary over the growing season due to differences in uptake by roots as a consequence of soil and weather conditions. In Colombia, Pastrana et al. (1990) found that the concentrations of Cu, Co, Fe, Mn, Mo, Se, and Zn in kikuyu during the wet season were sufficient to meet the requirements of sheep but levels of Cu and Mo were deficient during the dry season.

Pasture intake

Kikuyu grass intake has been estimated by several authors. For comparative purposes, results are presented as g/kg BW0.75. Organic matter intakes range from 35-57 g/kgBW0.75 for sheep (Betteridge 1979; Meissner and Paulsmeier 1988; Köster et al. 1992; Pienaar et al. 1993b). When results have been expressed as dry matter intake (DMI), figures for sheep range from 39.8-60.5 g/kgBW0.75 (Jeffery 1971b; Minson 1972, 1973; Rees and Little 1980; Soto et al. 1980) implying that 6-12% of DM is in inorganic form. DMI for growing cattle ranges from 65.3-97.7 g/kgBW0.75 (Rees and Little 1980; Pattinson et al. 1981; Ramirez et al. 1983; Dugmore and du Toit 1988; Schiere et al. 1990). Estimates of intake of lactating cattle ranged from 87.5-128 g/kgBW0.75 (Colman and Holder 1968; Hamilton et al. 1992; Henning 1993).
Several factors related to the chemical composition of kikuyu grass have been associated with DMI and are responsible for the ranges observed in the literature. Soto et al. (1980) suggested that high DMD and low fibre constituents resulted in increased DMI, although Dugmore and du Toit (1988) could not find any relation between digestibility and intake. Laredo and Minson (1973) found that cattle consumed substantially more leaf than stem with the leaf having a slightly lower digestibility. They found a positive association between leaf DMD and intake but not between stem DMD and intake and attributed these results not to the DMD per se but to the shorter rumen retention time of the leaf fraction (20.5 vs 38.5 h) which was caused by the larger particulate surface area available for rumen degradation. They also found that the lower grinding energy of the leaf fraction contributed to higher intakes and this can be correlated to the physical breakdown of large to small particles during rumination. The higher particle size breakdown rate of the leaf fraction probably caused a higher rate of passage of leaf small particles from the rumen, thus explaining the lower mean retention time and the higher intakes. These results can also be related to the work of Moir et al. (1977), who found that the higher resistance to digestion and a higher rumen retention time of the stem fraction resulted in reduced in vitro gas production. Recent evidence suggests that in vitro fermentation characteristics are highly related to the degradation characteristics of forages and that both are good predictors of DMI (Ørskov et al. 1988; Khazaal et al. 1993; Kibon and Ørskov 1993). The relationships with degradability would therefore explain to a greater extent the earlier observations of Laredo and Minson (1973) and Moir et al. (1977).
Very little information exists on the DM degradability of kikuyu grass, but Singh et al. (1992) (see section on nutritive value) estimated the 3 parameters for the Ørskov and McDonald (1979) degradation (d) equation (d = a+b(1-e-ct), where: the rapidly soluble fraction (a) was 1.63%, the insoluble but degradable fraction (b) was 61.12%, the degradation rate constant (c) was 4.1%/h and t was time.
The low DM content of kikuyu grass in some seasons (< 100 g/kg fresh grass) has also been implicated on several occasions in reductions in voluntary feed intake (Cross 1979b; Kenney et al. 1984; Piggot 1991; Köster et al. 1992) and this is probably associated to a high water holding capacity of the NDF fraction of kikuyu grass. This subject needs to be investigated.

Diet selection and grazing behaviour.

In situations where herbage allowance is high, ruminants will preferentially select the leaf component of the sward (Stobbs 1973a, 1974, 1975; Chacon and Stobbs 1976a, 1976b; Murtagh et al. 1980b; Minson 1981). This has justified the observation of several researchers (Stobbs 1975; Chacon and Stobbs 1976a, 1977; Murtagh et al. 1980b; Cowan et al. 1986, 1993; Hughes et al. 1988) that leaf yield is a better estimator of potential animal performance than total DM. This is particularly important in tropical grasses such as kikuyu as the leaf component does not always comprise the highest proportion of the total DM in the sward (Stobbs 1973b). Where stolons comprise a large part of the herbage present, animals prefer to decrease their intake rate rather than to eat the stoloniferous material (Minson 1981; Dugmore et al. 1991; Fulkerson and Slack 1993). The data of Pattinson et al. (1981) suggest a sharp decrease in intake after the first day of grazing. Although they did not measure the quantity of leaf available, their results suggest that animals tended to decrease pasture consumption as the proportion of leaf DM diminished. Kikuyu grass can form a mat of stolons when lightly grazed and this causes a reduction in the quality of the sward and an increase in sward height. Pre-grazing sward heights range from 10-35 cm (Bransby 1980, 1981; dos Santos Abrahao 1983; Henning 1993; J.B. Hacker, personal communication), for rest periods of 3-5 weeks. Mears (1970) reported that, under some conditions, kikuyu grass may form a loose sward up to 46 cm high. It is often stated that kikuyu pastures should be kept short to maintain dense, leafy pastures (Quinlan et al. 1975; Bransby 1981), or when associated with legumes to permit proper legume establishment (Mears 1970). Reeves et al. (1993; 1994) in New South Wales, Australia; found that hard grazing or mulching to 5 cm with regrowth periods of 4-5 weeks kept the sward in a leafy state with a high nutritional value. Whether the same management regimes apply to other climatic conditions is still not known and should be investigated.
From the grazing behaviour viewpoint, pasture intake can be described as a function of grazing time, biting rate and bite size (Allden and Whittaker 1970; Stobbs 1973a, 1973b; Hodgson 1985). There is very little information on the grazing behaviour of ruminants on kikuyu pastures, but the observations of Stobbs (1973a; 1974) and Chacon and Stobbs (1976a) suggest that the maximum biting rate of cows grazing tropical pastures is around 36,000 bites per day. As the proportion of leaf in the sward decreases, grazing animals tend to decrease their bite size and increase their biting rate (Chacon and Stobbs 1976a), which would suggest a higher selection for leaf as the sward is progressively defoliated.
Stobbs (1973a) and Chacon and Stobbs (1976a) found that bite size greatly determines intake and so is one of the most important components of grazing behaviour. Chacon and Stobbs (1977) found that, at a high leaf DM allowance (2.2 t/ha), bite size of Jersey cows grazing 5-week old kikuyu regrowth was 237 g OM/bite. Using data from Stobbs (1973a), a 400 kg Jersey cow grazing for 10 h and eating 80g OM/kg BW0.75 would have had a biting rate of 28,320 bites/d which is consistent with the maximum biting rate previously mentioned. Bulk density of the pasture is positively correlated with bite size (Chacon and Stobbs 1976a; Black and Kenney 1984) and kikuyu swards usually have a high bulk density (Quinlan et al. 1975; Chacon and Stobbs 1977; Hacker and Evans 1992) if kept short.
Another way in which ruminants modify their grazing behaviour to maintain intake as sward structure changes is altering grazing time. The maximum grazing time for cattle in tropical swards is between 9-12 h/day (Stobbs 1974; Chacon and Stobbs 1976a). Animals increase the time spent grazing when leaf DM allowance decreases to counteract the smaller bite size and to try to maintain intake (Chacon and Stobbs 1976a). Henning (1993) found that lactating dairy cows grazing swards of 2, 4 or 8 weeks of regrowth grazed for 8.1, 6.3 and 5.7 h/day, respectively. Intake was similar for all treatments (128g DM/kg BW0.75). This suggests that the lower leaf availability of the shortest regrowth period (2 weeks) forced the animals to eat for longer periods and possibly to take more bites per day, while at the higher herbage allowance (8 weeks), swards were probably denser and bite size greater, therefore decreasing biting rate and grazing time. A lower nutritional value may have also played a role in limiting intake of the latter group.
It has been well recognised that grazing animals will select plant parts of higher nutritional value, and that, in the case of kikuyu grass, these are leaves (see section on nutritive value). However, recent evidence (Dugmore et al. 1991; Milne 1991) suggests that there are exceptions to this rule and that material of a lower quality may be preferred on some occasions. Dugmore et al. (1991) found that steers selected kikuyu plant material of a lower CP concentration when the swards had CP concentrations above 150 g/kg DM. They suggested that the animals 'sought' an optimum CP concentration in their diet, which in their studies was around 140 g/kg DM. Their observations are possibly related to the fact that, when CP concentrations are high (>180 g/kg DM), kikuyu accumulates high levels of NPON which are soluble (Dugmore and du Toit 1988; Marais 1990a, 1990b). Under these conditions, the low non-structural carbohydrate concentrations or the low degradation rates of the structural fraction of kikuyu grass prevent the rumen microflora from utilising the highly soluble N levels and an ammonia overflow occurs. This results in enhanced N excretion and a decrease in the N content of the grazed diet to counteract and regulate the excessive ammonia levels by the steers. Other evidence suggests that ruminants can manipulate the quality of their diet to obtain a balance between dietary components and meet their requirements for specific nutrients. Kyriazakis and Oldham (1993), in an interesting choice-feeding experiment, found that pen-fed sheep could select a diet that met their CP requirements when given access to 2 feeds with CP concentrations below and above their requirements. They found that sheep regulated the consumption of the different feeds to maintain a CP intake close to their requirements and that, to a certain extent, they could avoid an excess of protein intake. They also suggested that sheep might discriminate against feeds with excess urea which would agree with the findings of Dugmore et al. (1991) with steers grazing kikuyu with high concentrations of soluble N.
Diet selection and grazing behaviour are 2 fundamental areas that need further research in order to understand better the interactions between the physical and chemical composition of kikuyu grass swards, intake and animal performance.

Animal performance on kikuyu pastures

Dairy cattle

Daily milk production from unsupplemented dairy cows grazing kikuyu grass (Table 5) ranged from 7.7-16 kg for Hostein-Friesian cattle (Cross 1979b; Olney et al. 1982; Laredo et al. 1983; Olney and Albertsen 1984; Hamilton et al. 1992; Henning 1993; Reeves et al. 1993; Fulkerson, personal communication) and 7.2-11.2 kg for Jersey and Guernsey cows (Stobbs 1972; 1973b; Colman and Kaiser 1974; Murtagh et al. 1980a). Dos Santos Abrahao (1983) found average milk production levels of 12.4 kg/d for Flamenga cows in Brazil, and Hughes et al. (1988) found maximum productions of 10.8 kg/cow/d from a mixed group of Holstein and Guernsey cows. Cross (1979b) and Hamilton et al. (1992) report milk production levels of 15.7 and 14.7 kg/d, respectively, for Friesian cows in early lactation. Similar milk production levels can be achieved from Friesian cows in mid lactation grazing well managed kikuyu swards (Reeves et al. 1994). These results agree with the observations of Stobbs (1971), that maximum milk production levels that can be obtained from kikuyu grass are around 15 kg/d (±4500 kg/lactation). Butterfat (Stobbs 1972; Colman and Kaiser 1974; Olney and Albertsen 1984; Hughes et al. 1988; Hamilton et al. 1992) and protein (Stobbs 1972; Hughes et al. 1988; Hamilton et al. 1992) levels in milk produced on kikuyu have ranged from 3.62-4.70% and 2.79-3.29%, respectively.
(Table 5 about here)
As with other tropical grasses, the main factors controlling milk production when kikuyu grass is the only component of the diet are stocking rate (Colman and Kaiser 1974; Olney et al. 1982; Olney and Albertsen 1984), pasture availability and nutritive value (Murtagh et al. 1980a; Hughes et al. 1988; Henning 1993) and their interactions (see Humphreys 1991 for a detailed review). Colman and Kaiser (1974), working on swards fertilised with 336 kg N/ha, found a reduction in milk yields from 7.9 to 7.2 kg/cow/d when the stocking rate of Jersey and Guernsey cows was increased from 2.47 to 4.94 animals/ha. In a study with Friesian cows on swards fertilised with 200 kg N/ha, increasing stocking rate from 5 to 7 cows/ha decreased milk production/cow from 9.1 to 6 kg/day (Olney and Albertsen 1984). Obviously these experiments cannot be compared directly, but the lower stocking rates, the higher fertiliser rate and the smaller breeds used in Colman and Kaiser's (1974) study, suggest that pasture availability per animal was high even at the high stocking rate and therefore milk production per cow was only slightly reduced while milk and butterfat production/ha increased dramatically. In Olney and Albertsen's (1984) experiment the chosen stocking rates (5 and 7 cows/ha) were very high and as a consequence milk production per cow at the high stocking rate decreased to a level where milk production per ha also diminished and supplementation seemed beneficial. These results were possibly exacerbated by the use of continuous grazing which created differences in pasture availability (3.6 vs. 2.5 t/ha, respectively) between the groups kept at 5 or 7 cows per ha (Olney and Albertsen 1984). Further evidence of the effect of kikuyu grass availability on milk production is found in the trial of Hughes et al. (1988). They found that cows grazing swards with a higher dry leaf availability (1200 vs 800 kg/ha) produced more milk (9.60 vs 8.01 kg/cow/d) and lost less liveweight (-0.2 vs. -0.75 kg/cow/d) than cows on the lower leaf availability. Henning (1993) also found that cows grazing 4-week old regrowth produced more milk (10.1 vs. 8.7 kg/cow/d) than cows grazing 2-week old regrowth, which was presumably caused by the higher leaf availability at the older stage of growth. However, he found reduced milk production when the age of the regrowth was 8 weeks, possibly through reduction in the nutritive value of the grass as it matured.
Supplementation of dairy cattle grazing kikuyu grass is common practice. Milk production from cows grazing kikuyu grass and supplemented with concentrates or grain supplements at levels up to 8.3 kg/d ranges from 8.9 to 22.5 kg/cow/d (Colman and Kaiser 1974; Moir et al. 1977; Olney and Albertsen 1984; Campabadal and Sanchez 1986; Davison et al. 1991; Hamilton et al. 1992; van der Grinten et al. 1992; Reeves et al. 1993, 1994). The existing information suggests that, when concentrates or grain supplements are offered, the increase in milk production ranges from 0.4 to 1.3 kg milk/kg supplement offered (Colman and Kaiser 1974; Olney and Albertsen 1984; Hamilton et al. 1992; Reeves et al. 1994). The degree of response is dependent on pasture availability; at high availability the effect of supplements is lower and a higher substitution rate is observed, while their effect is maximised as pasture allowance decreases (see Humphreys 1991 for a review). Olney and Albertsen (1984) fed 4 kg/d barley to cows grazing swards at stocking rates of 5 or 7 cows/ha and increased milk production by 0.6 and 1.1 kg/kg supplement offered, respectively. They cautioned that supplementation was economically justifiable only at the higher stocking rate where pasture availability was low. Colman and Kaiser (1974) found responses of 0.6 kg milk/kg supplement when feeding 2.7 kg crushed oats to cows stocked at 4.94 animals/ha.
Reeves et al. (1994) also found that the degree of response was dependent on the type of supplement fed. In their studies, feeding barley as a carbohydrate source (4.8 kg/cow) and canola meal as rumen protected protein (1.2 kg/cow) produced milk yields of 21.5 kg/cow/d with an efficiency of response to supplementation of 0.9 kg milk/kg concentrate. The responses obtained were lower when feeding the energy supplement alone and decreased as supplement intake increased (0.8, 0.45 and 0.38 kg milk/kg barley at feeding levels of 3, 6 and 9 kg barley/cow). This shows the importance of a good balance of nutrients for high production. Unfortunately, it does not reflect the interactions between forage and supplement intake and their utilisation for productive purposes. When forage supplements have been offered, the increase in milk production has been low (< 0.35 kg milk/kg supplement) (Royal and Hughes 1976; Hughes et al. 1988).

Beef cattle

Since Mears' (1970) review, a considerable amount of information has been published regarding liveweight gains of cattle grazing kikuyu pastures (Table 6).
(Table 6 about here)
Authors differ in the definition of pasture available to grazing cattle. However, the evidence suggests that, at high herbage allowances, liveweight gains of cattle can range from 0.6-1.0 kg/animal/d (Cowan et al. 1976; Bransby 1981, 1990; Tainton et al. 1982; Campbell et al. 1987). Cowan et al. (1976) and Tainton et al. (1982) suggest that these liveweight gains can be obtained in periods of active pasture growth (i.e. summer and autumn in the subtropics) where more grass is available to the animals. Bransby (1981; 1990) and Karnezos et al. (1988) found similar gains when using sward height, as measured by a disc meter, as an indicator of herbage availability. At sward heights of 12-14 cm, their steers gained weight at rates above 0.6 kg/animal/d while liveweight gain decreased linearly with height as the sward was defoliated. However, in trials where results have been expressed without considering seasonal growth differences, mean liveweight gains throughout the year vary between 0.35-0.55 kg/animal/d (Evans and Hacker 1973, 1992a, 1992b; Kaiser 1975; Cowan et al. 1976; Sanchez et al. 1983). In New Zealand, Piggot (1991) reported average liveweight gains of 0.5 kg/animal/d for summer grazing pastures including 30-55% kikuyu grass.
Studies investigating the effect of stocking rate and its interaction with herbage availability and liveweight gain of cattle grazing kikuyu pastures, have shown a negative linear relationship between stocking rate and liveweight gain per animal, with the magnitude of the slope of the regression being dependent on pasture availability (Mears and Humphreys 1974b; Tainton et al. 1982; Bransby 1984; Evans and Hacker 1992a, 1992b). Mears and Humphreys (1974b) found that stocking rates giving maximum liveweight gain/ha caused reductions of almost 25% in liveweight gain/animal, which agrees with the decrease of 21% reported by Evans and Hacker (1992b). Evans and Hacker (1992b) also suggested that the productivity of stoloniferous grasses such as kikuyu was less affected than tussock-forming grasses (i.e. setarias) by high stocking rates, as the decrease in liveweight gain/animal per unit of increased stocking rate was less.
In terms of carcass composition, Mears and Humphreys (1974b) found that increasing the stocking rate decreased dressing percentage, eye muscle area and the depth of subcutaneous fat, irrespective of N application rate. However, in their study, only the animals kept at low stocking pressures (2.2, 3.3, 4.9 and 7.4 steers/ha for the paddocks receiving 0, 134, 336 and 672 kgN/ha, respectively) achieved an acceptable degree of finish for the market. In the study of Tainton et al. (1982), the lower N fertiliser rate (150 kg N/ha) and lower stocking rate (7.12 steers/ha) produced a higher proportion of high quality carcasses as a reflection of the higher liveweight gain/animal of that group (0.8 vs. 0.64 kg/d). Kaiser (1975) found a 16% increase in carcass weights when calves grazing kikuyu grass were supplemented with a mineral mixture. He postulated that the response to the mineral supplement may have been due to the alleviation of Ca deficiency which could have been caused by high proportions of this mineral bound to oxalates, and therefore, unavailable to the animals.

Sheep

Information on sheep production from kikuyu pastures is available only from subtropical regions. Early observations made by Joyce (1974) in a stall-feeding experiment suggested that performance of sheep from kikuyu pastures was disappointing due to inadequate voluntary intake of digestible energy. The wethers lost weight, and although wool growth was not affected, it was significantly lower than that for animals on a grass/barley meal diet. Similar observations have been made in South Africa by Meissner and Paulsmeier (1988) and Barnes and Dempsey (1993), who found liveweight losses when sheep grazed kikuyu during the winter. These responses are usually associated with a decrease in the seasonal quantity and quality of kikuyu grass. However, Meissner and Paulsmeier (1988) also reported average liveweight gains of 132 g/d from stall-fed 45kg Dohne Merino wethers, while Hennesy and Williamson (1976) reported gains of 117 g/d when pelleted kikuyu leaf was given to penned 25 kg Dorset Horn X Border Leicester X Merino ewes. Similar liveweight gains (112 g/d and 150g/d, respectively) from Cheviot/Romney animals at grazing have been observed by Rumball and Boyd (1980) and Betteridge (1979) in mixed kikuyu-ryegrass-clover pastures after weaning and at 18 months of age, respectively. Betteridge (1979) suggested that as the proportion of kikuyu increased in mixed pastures, the performance of the animals tended to decrease due to a lower intake of readily fermentable carbohydrate. This is consistent with the results of van Ryssen et al. (1976), who found beneficial effects of molasses or maize meal supplementation on the growth and carcass composition of lambs grazing kikuyu grass.
Rumball (1985) concluded that a high stocking rate and controlled grazing were essential for efficient management of sheep production systems based on mixed pastures including kikuyu grass. He also postulated that grazing pressure should be doubled during the autumn to suppress the competitiveness of kikuyu grass compared with the temperate species. Rumball and Boyd (1980) and Piggot (1991) also suggested that later lambing or split spring-autumn lambing improved production in this type of system. The benefits of kikuyu grass in these systems are mediated via the increased stocking rates that can be maintained (Piggot 1991).

Toxic factors

Research on toxic factors in kikuyu grass has concentrated on 3 main areas. Since the early observations of Cordes et al. (1969), Busch et al. (1969), Martinovich and Smith (1973) and Smith and Martinovich (1973) in New Zealand on the death of cattle grazing kikuyu grass due to acute ruminal indigestion and alkalosis, similar observations have been reported in Australia (Gabbedy et al. 1974; Wong et al. 1987) and South Africa (Bryson and Newsholme 1978; van Heerden et al. 1978; Bryson 1982; Newsholme et al. 1983). The condition has also affected sheep (Martinovich and Smith 1972; Peet et al. 1990) and goats (Peet et al. 1990) but to a lesser extent. Even though the clinical signs in most studies have been similar, the cause of what is called 'kikuyu poisoning' is still inconclusive. Several factors appear to predispose livestock to the disease. For example, kikuyu poisoning has been observed after rains following periods of hot, dry weather in animals grazing lush kikuyu grass. Recent infestation of kikuyu paddocks with army worm (Pseudaletia separata in New Zealand; Spodoptera exempta in South Africa; and Mythimna convecta in Australia) is commonly associated with the disease. This raises the possibility that the condition is not caused by toxic factors in kikuyu grass per se but by a product secreted by the army worms. Myrothecium sp. fungi (Martinovich et al. 1972) have also been implicated, but the results are equivocal.
Kikuyu grass is known to accumulate high levels of soluble oxalates (Blaney et al. 1981; Elphinstone 1981; Williams 1987; Marais 1990a; Williams et al. 1991). High levels (>0.3%) of soluble oxalates have caused Osteodystrophia fibrosa (bighead disease) in horses consuming kikuyu grass (Blaney et al. 1981; Elphinstone 1981). Marais (1990a) found that kikuyu samples contained equal proportions of soluble and insoluble oxalates. He also found that kikuyu leaves contained higher levels (0.79% and 0.54%, respectively) of soluble and insoluble oxalates than stems (0.02% and 0.37%, respectively) and postulated that, in some instances, the formation of calcium oxalate might reduce the bio-availability of Ca for ruminants grazing kikuyu (see section on Mineral composition). Ruminants seem to be more tolerant of oxalate poisoning than non-ruminants, as rumen bacteria can adapt to high concentrations of soluble oxalates and convert them into carbon dioxide.
Kikuyu grass may also accumulate toxic levels of nitrates under some circumstances (Marais 1980, Marais 1990a, 1990b; Williams 1987; Marais et al. 1987, 1988, 1990; Williams et al. 1991) and these are usually associated with CP concentrations higher than 180 g/kg DM (Marais 1990a, 1990b). High N fertilisation (Williams et al. 1991) and high K uptake rates (Marais et al. 1987) seem to increase the nitrate content of kikuyu grass. Stems accumulate higher concentrations of nitrates than leaves (Marais et al. 1987; Marais 1990a, 1990b) largely because they are substrate storage organs. This is an important observation, since it suggests that a reduction in the nitrate levels of kikuyu grass could be obtained by changes in management practices: i.e. by removal of the aftermath of stolons. The effect of high nitrate levels on ruminant digestion is not a direct one. Marais et al. (1988) found that accumulation of high nitrite levels during the conversion of nitrate to ammonia caused a reduction in the digestibility of kikuyu grass and a reduction in rumen microbial populations. The high proportion of soluble N in high nitrate kikuyu grass may also be responsible for the low N retention sometimes observed in grazing ruminants (Marais et al. 1990). This factor is known to cause energy-protein imbalances as shown by elevated rumen ammonia levels (Marais et al. 1990). Under these circumstances, the energy required for the extensive recycling and subsequent excretion of N in urine may play a part in the often reported low animal performance from this grass species. The subject needs further investigation.
The potential of kikuyu pastures to produce cerebrocortical necrosis in ruminants has also been studied (Meyer 1989), but their role was discounted, as very low thiaminase activities and thiamine concentrations were found in the samples analysed. Tannin contents of kikuyu grass have been found to be low (Reid et al. 1979).

Future areas of research

From this review, it can be concluded that research on kikuyu grass should be directed towards the following areas:
* Grazing management strategies for kikuyu should be based on a more ecophysiological approach and need to include morphological indicators of the optimum time to graze (Minson et al. 1993). Of particular importance are the factors determining the activity, rates of appearance, senescence and growth of leaves and roots, as substrate uptakes are dependent on them. Suitable legumes for associations should also be studied on this basis, as competition for substrates (i.e. light, nutrients) will in part determine their growth and success in a mixed pasture. This sort of information should help the development of management strategies to keep swards at their highest nutritional value without compromising DM availability and sward structure.
* In subtropical regions, the development of management strategies which use high levels of legumes to extend the seasonal production of kikuyu-based pastures in winter-spring is required (Minson et al. 1993). In tropical regions, the introduction of legumes with suitable morphological characteristics into kikuyu pastures should be tested, and adequate management strategies developed. This could increase the sustainability of kikuyu grass grazing systems.
* Kikuyu grass quality could be improved by plant breeders (Minson et al. 1993). Cherney et al. (1992) stated that introducing brown-midrib, low-lignin mutants could increase the digestibility of forage by up to 10%. However, the assessment of the nutritive value of kikuyu grass should also include a more dynamic description of the nutritional fractions. The knowledge of the concentration of a particular nutritional component, particularly N and the carbohydrate fractions, should be coupled with their kinetic characterisitics. The relationship between the different nutritional fractions should also be the subject of further studies as the balance of nutrients might be more important than the concentrations of individual nutrients. This may also help in understanding the factors controlling intake and performance of livestock grazing kikuyu grasslands. The low non-structural carbohydrate levels in this grass species should also receive more attention.
* Studies on pasture utilisation should include measurements of the different sward components (leaves, stolons and dead material) as their proportions largely affect intake. Experiments on diet selection and grazing behaviour should also be encouraged, particularly those studying the interaction between the swards' physical characteristics, herbage allowance, preference and intake rates. This type of study should provide information on the desired sward structures for particular purposes and levels of animal production.
It is believed that targeting more research on these areas will provide the understanding necessary to produce suitable management guidelines to obtain the desired levels of animal performance on kikuyu grass grazing systems.

Acknowledgements

The discussions and exchange of information with Drs. G.J. Murtagh, W. Fulkerson, J.B. Hacker, J.P. Marais and Professor L. 't Mannetje are greatly acknowledged. Financial support from the British Overseas Development Administration (ODA) is also appreciated.

References

ALI, T. and STOBBS, the late T.H. (1980) Solubility of the protein of tropical pasture species and the rate of its digestion in the rumen. Animal Feed Science and Technology, 5, 183-192.

ALLDEN, W.G. and WHITTAKER, I.A. (1970) The determinants of herbage intake by grazing sheep: the interrelationship of factors influencing herbage intake and availability. Australian Journal of Agricultural Research, 21, 755-766.

ANNESSENS, M. (1989) Essai de fertilisation du Pennisetum clandestinum au Cameroun: aspects qualitatifs et quantitatifs. Tropicultura, 7, 54-59.

ANON. (1983a) Pennisetum clandestinum Hochst. ex. Chiov. cv. Crofts. Journal of the Australian Institute of Agricultural Science, 49, 236-238.

ANON. (1983b) Pennisetum clandestinum Hochst. ex. Chiov. cv. Noonan. Journal of the Australian Institute of Agricultural Science, 49, 238-239.

ANON. (1987) Pasture production manual. Bulletin 414. Department of Agriculture and Water Supply, Natal Region, Republic of South Africa. pp. 82 and 85.

ARC (1984) The Nutrient Requirements of Ruminant Livestock. (Commonwealth Agricultural Bureau: Slough, England).

AWAD, A.S., EDWARDS, D.G. and MILHAM, P.J. (1976) Effect of pH and phosphate on soluble soil aluminium and on growth and composition of kikuyu grass. Plant and Soil, 45, 531-542.

AWAD, A.S. and EDWARDS, D.G. (1977) Reversal of adverse effects of heavy ammonium sulphate application on growth and nutrient status of a kikuyu pasture. Plant and Soil, 48, 169-183.

AWAD, A.S., EDWARDS, D.G. and HUETT, D.O. (1979) Seasonal changes in chemical composition of heavily fertilised kikuyu pasture and their potential effects on the mineral nutrition of grazing cattle. Australian Journal of Experimental Agriculture and Animal Husbandry, 19, 183-191.

BAILEY, R.W. and HUNT, W.F. (1973) Structural carbohydrate levels in kikuyu grass and ryegrass grown under identical conditions. New Zealand Journal of Agricultural Research, 16, 203-205.

BARNES, D.K. and DEMPSEY, C.P. (1993) Grazing trials with sheep on kikuyu (Pennisetum clandestinum Chiov.) foggage in eastern Transvaal highveld. African Journal of Range and Forage Science, 10, 66-71.

BATTESE, D. and FULKERSON, W.J. (1994) The kikuyu/clover system. Research to Farm Newsletter. Dairy Research and Development Corporation, NSW Agriculture. 2 pp.

BETTERIDGE, K. (1979) An evaluation of mixed kikuyu and ryegrass pastures grazed by sheep. New Zealand Journal of Experimental Agriculture, 7, 53-58.

BETTERIDGE, K. (1985) Introducing temperate grasses into resident swards. In: Kikuyu Grass Farming for High Production. Proceedings of a 'Kikuyu grass forum' held at Whangarei, New Zealand. New Zealand Grassland Association. pp 36-39.

BETTERIDGE, K. and HAYNES, D.A. (1986) Altering the growth pattern of kikuyu pastures with temperate species. Proceedings of the New Zealand Grassland Association, 47, 149-156.

BLACK, J.L. and KENNEY, P.A. (1984) Factors affecting diet selection by sheep. II. Height and density of pasture. Australian Journal of Agricultural Research, 35, 565-578.

BLANEY, B.J., GARTNER, R.J.W. and McKENZIE, R.A. (1981) The effects of oxalate in some tropical grasses on the availability to horses of calcium, phosphorus and magnesium. Journal of Agricultural Science, Cambridge, 97, 507-514.

BOLTON, J.K. and BROWN, R.H. (1980) Photosynthesis of grass species differing in carbon dioxide fixation pathways. V. Response of Panicum maximum, Panicum milioides and tall fescue (Festuca arundinacea) to nitrogen nutrition. Plant Physiology, 66, 97-100.

BRANDS, L.E. and COOK, B.G. (1976) Whittet kikuyu in the Gympie district. Queensland Agricultural Journal, 102, 429-432.

BRANSBY, D.I. (1980) Herbage height as a determinant of the preference of steers grazing three pasture species. Proceedings of the Grassland Society of Southern Africa, 15, 153-155.

BRANSBY, D.I. (1981) The effect of herbage availability on animal performance of kikuyu and Coastcross II. Proceedings of the Grassland Society of Southern Africa, 16, 49-50.

BRANSBY, D.I. (1984) A model for predicting livemass gain from stocking rate and annual rainfall. Journal of the Grassland Society of Southern Africa, 1, 22-26.

BRANSBY, D.I. (1990) Nitrogen fertilisation, stocking rate and rotational grazing effects on steers grazing Pennisetum clandestinum. Journal of the Grassland Society of Southern Africa, 7, 261-264.

BREDON, R.M., STEWART, P.G. and DUGMORE, T.J. (1987) A manual on the nutritive value and chemical composition of commonly used South African farm feeds. Department of Agriculture and Water Supply, Natal Region, Republic of South Africa. p. 7.

BROCKETT, G.M. and GRAY, N.N. (1984) The performance of tropical pastures in the drier phase of the highland sourveld of Natal. Journal of the Grassland Society of Southern Africa, 1, 4-8.

BRYSON, R.W. (1982) Kikuyu poisoning and the army worm. Journal of the South African Veterinary Association, 53, 161-165.

BRYSON, R.W. and NEWSHOLME, S.J. (1978) Kikuyu grass poisoning of cattle in Natal. Journal of the South African Veterinary Association, 49, 19-21.

BUCKMAN, H.O. and BRADY, N.C. (1969) The Nature and Properties of Soils. 7th Edn. (Collier-Macmillan: London).

BUSCH, J., HARRIS, G.H., COUP, M.R. and CORDES, D.O. (1969) Acute ruminal indigestion, alkalosis and death of cattle grazing kikuyu grass. New Zealand Veterinary Journal, 17, 182-183.

BUXTON, D.R. and FALES, S.L. (1994) Plant environment and quality. In: Fahey, G.C., Collins, M., Mertens, D.R. and Moser, L.E. (eds.) Forage quality, evaluation and utilization. (American Society of Agronomy: Madison). pp. 155-199.

CAMPABADAL, C. and SANCHEZ, J. (1986) Utilización de la soya integral en la alimentación de vacas en pastoreo. Agronomía Costarricense, 10, 1-2, 147-152.

CAMPBELL, C.M., SHERROD, L.B. and ISHIZAKI, S.M. (1969) Effect of supplemental protein and energy levels on the utilisation of kikuyu grass (Pennisetum clandestinum). Journal of Animal Science, 29, 634-637.

CAMPBELL, C.M., NOLAN, J.C., HO-A, E.B., KIMURA, C. and BERGIN, W.C. (1987) Effects of supplementation on the performance of feeder cattle on Hawaiian ranges. Research Extension Series 080. Hawaiian Institute of Tropical Agriculture and Human Resources. 6 pp.

CASTILLO, E., COWARD, J., SANCHEZ, J.M., JIMENEZ, C. and LOPEZ, C. (1983) Efecto de la fertilización nitrogenada sobre productividad, composición química y digestibilidad in vitro del pasto kikuyo bajo pastoreo en el cantón de Coronado. Agronomía Costarricense, 7, 1-2, 9-15.

CHACON, E. and STOBBS, T.H. (1976a) Influence of progressive defoliation of a grass sward on the eating behaviour of cattle. Australian Journal of Agricultural Research, 27, 709-727.

CHACON, E. and STOBBS, T.H. (1976b) Diurnal variation in bite size. Annual Report 1975-76, Division of Tropical Crops and Pastures, CSIRO, Australia. p. 77.

CHACON, E. and STOBBS, T.H. (1977) The effects of fasting prior to sampling and diurnal variation on certain aspects of grazing behaviour in cattle. Applied Animal Ethology, 3, 163-171.

CHERNEY, J.H., CHERNEY, D.J.R., AKIN, D.E. and AXTELL, J.D. (1992) Potential of brown-midrib, low-lignin mutants for improving forage quality. Advances in Agronomy, 46, 157-198.

CHOU, C.H., CHANG, S.J., CHENG, C.M., WANG, Y.C., HSU, F.H. and DEN, W.H. (1989) The selective allelopathy interaction of a pasture-forest intercropping in Taiwan. II. Interaction between kikuyu grass and three hardwood plants. Plant and Soil, 116, 207-215.

CHOU, C.H., HWANG, S.Y. and PENG, C.I. (1987) The selective allelopathic interaction of a pasture-forest intercropping in Taiwan. Plant and Soil, 98, 31-41

COLMAN, R.L. (1966) Growth curve studies on kikuyu grass. Annual Report 1965-66, Wollongbar Agricultural Research Station.

COLMAN, R.L., HOLDER, J.H. and SWAIN, F.G. (1966) Production from dairy cattle on improved pasture in a subtropical environment. Proceedings of the X International Grassland Congress, Helsinki, 1966. pp. 499-503.

COLMAN, R.L. and HOLDER, J.M. (1968) Effect of stocking rate on butterfat production of dairy cows grazing kikuyu grass fertilised with nitrogen. Proceedings of the Australian Society of Animal Production, 7, 129.

COLMAN, R.L. and KAISER, A.G. (1974) The effect of stocking rate on milk production from kikuyu grass pastures fertilised with nitrogen. Australian Journal of Experimental Agriculture and Animal Husbandry, 14, 155-160.

COLMAN, R.L. and O'NEILL, G.H. (1978) Seasonal variation in the potential herbage production and response to nitrogen by kikuyu grass (Pennisetum clandestinum). Journal of Agricultural Science, Cambridge, 91, 81-90.

CONSENTINO, J.R., BUFARAH, G. and PAULINO, V.T. (1985) Adaptaçao de plantas forrageiras no estado de Sao Paulo: Sao Luis do Paraitinga. Zootecnia, Nova Odessa, 1, 23-37.

COOK, B.G. and MULDER, J.C. (1984a) Responses of nine tropical grasses to nitrogen fertilizer under rain-grown conditions in south-eastern Queensland. 1. Seasonal dry matter productivity. Australian Journal of Experimental Agriculture and Animal Husbandry, 24, 410-414.

COOK, B.G. and MULDER, J.C. (1984b) Responses of nine tropical grasses to nitrogen fertilizer under rain-grown conditions in south-eastern Queensland. 2. Concentrations and uptakes of nitrogen, phosphorus and potassium in plant tops. Australian Journal of Experimental Agriculture and Animal Husbandry, 24, 410-414.

CORDES, D.O., COUP, M.R., HARRIS, G.H., DAVENPORT, P.G. and BUSCH, J. (1969) Acute ruminal indigestion, alkalosis and death of cattle grazing kikuyu grass. New Zealand Veterinary Journal, 17, 77-81.

COWAN, R.T., CHOPPING, G.D., BYFORD, I.J., THURBON, P.N., HUTCHINGS, A., DEANS, H. and DAVISON, T. (1976) Factors affecting growth rate of dairy heifers in Queensland. Proceedings of the Australian Society of Animal Production, 12, 26.

COWAN, R.T., DAVISON, T.M. and SHEPHARD, R.K. (1986) Observations on the diet selected by Friesian cows grazing tropical grass and grass-legume pastures. Tropical Grasslands, 20, 183-191.

COWAN, R.T., MOSS, R.J. and KERR, D.V. (1993) Northern dairy feedbase 2001. 2. Summer feeding systems. Tropical Grasslands, 27, 150-161.

CROSS, G.W. (1979a) The production potential of Pennisetum clandestinum in the Midlands and on the south coast of Natal. Proceedings of the Grassland Society of Southern Africa, 14, 53-59.

CROSS, G.W. (1979b) Maintaining the nutritive value of Pennisetum clandestinum for milk production. Proceedings of the Grassland Society of Southern Africa, 14, 61-63.

DAVIES, L.J. and HUNT, B.J. (1983) Growth of tropical grass introductions in mixed swards with ryegrass and clover under mowing. New Zealand Journal of Agricultural Research, 26, 415-422.

DAVISON, T.M. (1985) Perennial clovers as an alternative to annual temperate pastures for winter-spring milk production in Queensland. In: The challenge: efficient dairy production. Proceedings of a conference organised by the Australian and New Zealand Societies of Animal Production. March 25-28, 1985. Albury-Wodonga, Australia.

DAVISON, T.M., VEERVORT, F.P. and DUNCALFE, F. (1991) Responses to a long chain fatty acid supplement fed to dairy cows at two stages of lactation. Australian Journal of Experimental Agriculture, 31, 467-470.

de FIGUEREIDO, M.C.V. (1991) Effect of enzyme additives on kikuyu grass silage quality. Journal of the Grassland Society of Southern Africa, 8, 110-113.

de FIGUEREIDO, MARIA; MARAIS, J.P. and TAINTON, N.M. (1989) Effect of wilting on biochemical and microbiological changes in Kikuyu (Pennisetum clandestinum Hochst) silage. Journal of the Grassland Society of Southern Africa, 6, 99-103.

de FIGUEREIDO, M., MARAIS, J.P. and TAINTON, N.M. (1990) The effect of nitrogen fertilization on the quality of kikuyu grass silage. Journal of the Grassland Society of Southern Africa, 7, 41-45.

de FIGUEREIDO, M. and MARAIS, J.P. (1994) The effect of bacterial inoculants on kikuyu silage quality. Journal of Agricultural Science, Cambridge, 122, 53-60.

dos SANTOS ABRAHAO, J.J. (1983) Produçao de leite em pastagem de quicuio, sob pastejo em faixas. Comunicado Tecnico No. 59. Empresa Catarinense de Pesquisa Agropecuaria. Florianopolis, Santa Catarina, Brazil. 6 pp.

DUGMORE, T.J. and du TOIT, J.H. (1988) The chemical composition and nutritive value of kikuyu pasture. South African Journal of Animal Science, 18, 72-75.

DUGMORE, T.J., du TOIT, P.C.V. and MORNING, S.J. (1991) Dietary selection by steers grazing kikuyu. South African Journal of Animal Science, 21, 194-197.

DUGMORE, T.J., van RYSSEN, J.B.J. and STIELAU, W.J. (1986) Effect of fibre and nitrogen content on the digestibility of kikuyu (Pennisetum clandestinum). South African Journal of Animal Science, 16, 197-201.

ELPHINSTONE, G.D. (1981) Pastures and fodder crops for horses in southern coastal Queensland. Queensland Agricultural Journal, 107, 122-126.

ERICKSEN, F.I. and WHITNEY, A.S. (1981) Effects of light intensity on growth of some tropical forage species. 1. Interaction of light intensity and nitrogen fertilization on six forage grasses. Agronomy Journal, 73, 427-433.

EVANS, T.R. and HACKER, J.B. (1973) Comparison of animal production for six tropical grasses. Annual Report 1972-73, Division of Tropical Agronomy, CSIRO, Australia. pp. 11-12.

EVANS, T.R. and HACKER, J.B. (1992a) An evaluation of the production potential of six tropical grasses under grazing. 2. Assessment of quality using variable stocking rates. Australian Journal of Experimental Agriculture, 32, 29-37.

EVANS, T.R. and HACKER, J.B. (1992b) An evaluation of the production potential of six tropical grasses under grazing. 3. Responses to set stocking rates under continuous grazing. Australian Journal of Experimental Agriculture, 32, 693-699.

FERNANDO, G.W.E. and JAYARATNE, J.A. (1980) Use of nitrogenous fertilizer on kikuyu grass under zero grazing management in the wet montane zone of Sri Lanka. II. Chemical composition of kikuyu grass. Tropical Agriculturalist, 136, 1-9.

FORDE, B.J., SLACK, C.R., ROUGHAN, P.G., HASLEMORE, R.M. and McLEOD, M.N. (1976a) Growth of tropical and temperate grasses at Palmerston North. II. Total nitrogen, soluble sugar, starch, and in vitro digestibility determinations. New Zealand Journal of Agricultural Research, 19, 489-498.

FORDE, B.J., SLACK, C.R., ROUGHAN, P.G. and WHITEHEAD, H.M.C. (1976b) Growth of tropical and temperate grasses at Palmerston North. 1. Warm-season yields with and without irrigation. New Zealand Journal of Agricultural Research, 19, 135-142.

FULKERSON, W.J., LOWE, K.J., AYRES, J.F. and LAUNDERS, T. (1993) Northern dairy feedbase 2001. 3. Winter pastures and crops. Tropical Grasslands, 27, 162-179.

FULKERSON, W.J. and SLACK, K. (1993) Productivity of Lotus penduculatus and Trifolium repens in a base pasture of kikuyu grass. Proceedings of the Eight Annual Conference of the Grassland Society of New South Wales. pp. 111-112.

FULKERSON, W.J. and SLACK, K. (1994) The kikuyu/white clover system. Dairy Research Report 1993-1994. Wollongbar Agricultural Institute. pp. 19-25.

GABBEDY, B.J., GWYNN, R., HOPKINSON, W.I. and KAY, B.E. (1974) Kikuyu poisoning of cattle in Western Australia. Australian Veterinary Journal, 50, 369-370.

GARDENER, C.J., McIVOR, J.G. and JANSEN, A. (1993a) Passage of legume and grass seeds through the digestive tract of cattle and their survival in faeces. Journal of Applied Ecology, 30, 63-74

GARDENER, C.J., McIVOR, J.G. and JANSEN, A. (1993b) Survival of seeds of tropical grassland species subjected to bovine digestion. Journal of Applied Ecology, 30, 75-85.

GARTNER, J.A. (1966) The effects of different rates of fertiliser nitrogen on the growth, nitrogen uptake and botanical composition of tropical grass swards. Proceedings of the X International Grassland Congress, Helsinki, 1966. pp. 223-227.

GOMIDE, J.A., NOLLER, C.H., MOTT, G.O., CONRAD, J.H. and HILL D.L. (1969a) Effect of plant age and nitrogen fertilization on the chemical composition and in vitro cellulose digestibility of tropical grasses. Agronomy Journal, 61, 116-120.

GOMIDE, J.A., NOLLER, C.H., MOTT, G.O., CONRAD, J.H. and HILL D.L. (1969b) Mineral composition of six tropical grasses as influenced by plant age and nitrogen fertilisation. Agronomy Journal, 61, 120-123.

GOOLD, G.J. (1979) Effect of nitrogen and cutting interval on production of grass species swards in Northland, New Zealand. New Zealand Journal of Experimental Agriculture, 7, 353-359.

GRACE, J. and RUSSELL, G. (1977) The effect of wind on grasses. III. Influence of continuous drought and wind on anatomy and water relations in Festuca arundinacea Schrel. Journal of Experimental Botany, 103, 268-278.

HACKER, J.B. and MINSON, D.J. (1981) The digestibility of plant parts. Herbage Abstracts, 51, 459-482.

HACKER, J.B. and EVANS, T.R. (1992) An evaluation of the production potential of six tropical grasses under grazing. 1. Yield and yield components, growth rates and phenology. Australian Journal of Experimental Agriculture, 32, 19-27.

HAMILTON, B.A., ASHES, J.R. and CARMICHAEL, A.W. (1992) Effect of formaldehyde-treated sunflower meal on the milk production of grazing dairy cows. Australian Journal of Agricultural Research, 43, 379-387.

HARRIS, D.I. and BARTHOLOMEW, P.E. (1991) The production of four ryegrass cultivars oversown at various seeding rates into irrigated kikuyu. Journal of the Grassland Society of Southern Africa, 8, 82-85.

HART, F.J. and LEIBHOLZ, J. (1990) The effect of species of grass, stage of maturity and level of intake on the degradation of protein and organic matter in the rumen of steers. Australian Journal of Agricultural Research, 41, 791-798.

HEFER, G.D. and TAINTON, N.M. (1990) Effect of nitrogen concentration of urea ammonium nitrate, and application level on the dry matter and chemical composition of Pennisetum clandestinum and Cynodon hybrid dryland pastures. Journal of the Grassland Society of Southern Africa, 7, 36-40.

HEFER, G.D., NAIKEN, V.G. and TAINTON, N.M. (1992) Leaf scorch on Pennisetum clandestinum dryland pastures induced by liquid urea ammonium nitrate fertilizer. Journal of the Grassland Society of Southern Africa, 9, 105-110.

HENNESY, D.W. and WILLIAMSON, P.J. (1976) The nutritive value of kikuyu grass (Pennisetum clandestinum) leaf and the use of pelleted leaf in rations high or low in energy. Australian Journal of Experimental Agriculture and Animal Husbandry, 16, 729-734.

HENNING, W.P. (1993) Kikuyu pasture: its use for milk production. Döhne Bulletin, 2, 4-6.

HENZELL, E.F. and OXENHAM, D.J. (1973) The relation between growth and leaf nitrogen concentration in Nandi Setaria and some other tropical grasses. Communications in Soil Science and Plant Analysis, 4, 147-154.

HERRERO, M. (1995) Grassland modelling: a decision support tool. In: Animal Production in Central America: conservation and sustainable development. Proceedings of a Workshop held in San José, Costa Rica, 1995. Ciencias Veterinarias, 17: 72-79.

HERRERO, M., FAWCETT, R.H. and DENT, J.B. (1995) Seasonal differences in the metabolisable energy concentration of kikuyu grass as predicted from in vitro gas production measurements. Annales de Zootecnie, 44: 46.

HERRERO, M. and JESSOP, N.S. (1996) The relationship between in vitro gas production and neutral detergent fibre dissapearance in three tropical grasses. Animal Science, in press.

HERRERO, M., MURRAY, I., FAWCETT, R.H. and DENT, J.B. (1996) Prediction of the in vitro gas production and chemical composition of kikuyu grass by near-infrared reflectance spectroscopy. Animal Feed Science and Technology, 60: 51-67.

HILL, M.J. (1985) Direct drilling tall fescue (Festuca arundinacea Schreb.), prarie grass (Bromus catharticus Vahl) and Italian ryegrass (Lolium multiflorum Lam.) into kikuyu and paspalum pastures. Australian Journal of Experimental Agriculture, 25, 806-817.

HILL, M.J., PEARSON, C.J. and CAMPBELL, L.C. (1985) Growth of seedlings of prairie grass and tall fescue in small swards of kikuyu at different temperatures. Australian Journal of Agricultural Research, 36, 213-220.

HODGSON, J. (1985) The control of herbage intake in the grazing ruminant. Proceedings of the Nutrition Society, 44, 339-346.

HUGHES, R.M., ROYAL, A.J.E. and AYRES, J.F. (1988) Milk production of dairy cows grazing long or short kikuyu grass (Pennisetum clandestinum), and with access to forage lupins (Lupinus luteus). Australian Journal of Experimental Agriculture, 28, 307-313.

HUMPHREYS, L.R. (1991) Tropical Pasture Utilisation. (Cambridge University Press: Cambridge, UK).

ISHIZAKI, S.M., CAMPBELL, C.M. and TOMA, W.Y. (1976) Microdigestion techniques and chemical solubility methods as estimators of the digestibility of tropical grasses. Journal of Animal Science, 42, 1503-1508.

ITO, K., NUMAGUCHI, H. and INOSAKA, M. (1985) Relation of non-structural carbohydrate concentration in autumn to wintering ability of several tropical grasses in Southern Kyushu. Proceedings of the XV International Grassland Congress, Kyoto, 1985. pp. 363-365.

IVORY, D.A. and WHITEMAN, P.C. (1978a) Effect of temperature on growth of five subtropical grasses. 1. Effect of day and night temperature on growth and morphological development. Journal of Agricultural Science, Cambridge, 5, 131-148.

IVORY, D.A. and WHITEMAN, P.C. (1978b) Effect of temperature on growth of five subtropical grasses. 2. Effect of low night temperature. Journal of Agricultural Science, Cambridge, 5, 149-157.

JEFFERY, H. (1971a) Nutritive value of Pennisetum clandestinum based pastures in a sub-tropical environment. Australian Journal of Experimental Agriculture and Animal Husbandry, 11, 173-177.

JEFFERY, H. (1971b) The relation between various energy parameters, chemical composition and digestibility of some pasture swards in a subtropical environment. Australian Journal of Experimental Agriculture and Animal Husbandry, 11, 397-402.

JESSOP, N.S. and HERRERO, M. (1996) Interpretation of data from the in vitro gas production technique. Journal of Animal Science, submitted.

JOHNSON, I.R. and THORNLEY, J.H.M. (1984) A model of instantaneous and daily canopy photosynthesis. Journal of Theoretical Biology, 107, 531-545.

JOYCE, J.P. (1974) Nutritive value of kikuyu grass. New Zealand Journal of Agricultural Research, 17, 197-202.

KAISER, A.G. (1975) Response by grazing calves grazing kikuyu grass pastures to grain and mineral supplements. Tropical Grasslands, 9, 191-198.

KAISER, A.G., HAVILAH, E.J., CHOPPING, G.D. and WALKER, R.G. (1993) Northern dairy feedbase 2001. 4. Feeding systems during winter and spring. Tropical Grasslands, 27, 180-211.

KARNEZOS, T.P., TAINTON, N.M. and BRANSBY, D.I. (1988) A mathematical model used to describe animal performance on kikuyu and Coastcross II pastures. Journal of the Grassland Society of Southern Africa, 5, 38-41.

KEMP, D.R. (1975) The growth of three tropical pasture grasses on the mid-north coast of New South Wales. Australian Journal of Experimental Agriculture and Animal Husbandry, 15, 637-644.

KEMP, D.R. (1976) The influence of cutting frequency and nitrogen fertilizer on the growth of kikuyu grass. Australian Journal of Experimental Agriculture and Animal Husbandry, 16, 382-386.

KENNEY, P.A., BLACK, J.L. and COLEBROOK, W.F. (1984) Factors affecting diet selection by sheep. III. Dry matter content and particle length of forage. Australian Journal of Agricultural Research, 22, 35-41.

KHAZAAL, K., DENTINHO, M., RIBEIRO, J.M. and ØRSKOV, E.R. (1993) A comparison of gas production during incubation with rumen contents in vitro and nylon bag degradability as predictors of the apparent digestibility in vivo and the voluntary intake of hays. Animal Production, 57, 105-112.

KIBON, A. and ØRSKOV, E.R. (1993) The use of degradation characteristics of browse plants to predict intake and digestibility by goats. Animal Production, 57, 247-252.

KÖSTER, H.H., MEISSNER, H.H., COERTZE, R.J. and RETHMAN, N.F.G. (1992) Voluntary intake and quality of diet selected by cattle grazing bana grass, kikuyu and forage sorghum. South African Journal of Animal Science, 22, 35-41.

KYRIAZAKIS, I. and OLDHAM, J.D. (1993) Diet selection in sheep: the ability of growing lambs to select a diet that meets their crude protein (nitrogen X 6.25) requirements. British Journal of Nutrition, 69, 617-629.

LAMBERT, J.P., RUMBALL, P.J. and CHRISTIE, A.J.R. (1977) Comparison of ryegrass and kikuyu grass pastures under mowing. New Zealand Journal of Experimental Agriculture, 5, 71-77.

LAMBERT, J.P., RUMBALL, P.J. and BOYD, A.F. (1979) Comparison of ryegrass-white clover pastures with and without paspalum and kikuyu grass. New Zealand Journal of Experimental Agriculture, 7, 295-302.

LAREDO, M.A. and MINSON, D.J. (1973) The voluntary intake, digestibility, and retention time by sheep of leaf and stem fractions of five grasses. Australian Journal of Agricultural Research, 24, 875-888.

LAREDO, M.A. and MINSON, D.J. (1975) The pepsin soluble dry matter of leaf and stem fractions of grasses in relation to voluntary iintake by sheep. Australian Journal of Experimental Agriculture and Animal Husbandry, 15, 203-206.

LAREDO, M.A. and MENDOZA, P.E. (1982) Valor nutritivo de pastos de zonas frías. 1. Pasto kikuyo (Pennisetum clandestinum Hochst). Anual y estacional. Revista ICA, 17, 4, 157-167.

LAREDO, M.A., ANZOLA, H. and ARDILA, A. (1983) Fluctuaciones minerales en pastos de clima frío colombiano. 2. Kikuyo (Pennisetum clandestinum Hochst), anual y estacional. Revista ICA. 18, 4, 269-278.

LAREDO, M.A., KLEEMANN, G. and CUESTA, A. (1990) Tabla nutricional para ganado de leche. Convenio Colombo/Aleman 'ICA-GTZ'.

LESTER, D.C., CARTER, O.G., KELLEHER, F.M. and LAING, D.R. (1972) The effect of gibberelic acid on apparent photosynthesis and dark respiration of simulated swards of Pennisetum clandestinum Hochst. Australian Journal of Agricultural Research, 23, 205-213.

LIPSETT, J. (1975) A comparison of the responses by six grasses, rape and subterranean clover to application of molybdenum. Australian Journal of Experimental Agriculture and Animal Husbandry, 15, 227-230.

LOWE, K.F. and BOWDLER, T.M (1984) Performance of temperate and tropical grasses and legumes under two irrigation frequencies in south-eastern Queensland. Tropical Grasslands, 18, 46-55.

LUDLOW, M.M. and WILSON, G.L. (1971) Photosynthesis of tropical pasture plants. 1. Illuminance, carbon dioxide concentration, leaf temperature and leaf-air vapour pressure difference. Australian Journal of Biological Sciences, 24, 449-470.

LUDLOW, M.M., SAMARAKOON, S.P. and WILSON, J.R. (1988) Influence of light regime and leaf nitrogen concentration on 77K fluorescence in leaves of four tropical grasses: no evidence for photoinhibition. Australian Journal of Plant Physiology, 15, 669-676.

MANNETJE, L. 't (1975) Effect of daylength and temperature on introduced legumes and grasses for the tropics and subtropics of coastal Australia. 2. N-concentration, estimated digestibility and leafiness. Australian Journal of Experimental Agriculture and Animal Husbandry, 15, 256-263.

MARAIS, J.P. (1980) Effect of nitrate and non-protein organic nitrogen levels in kikuyu (Pennisetum clandestinum) pastures on digestibility in vitro. Agroanimalia, 12, 7-11.

MARAIS, J.P. (1990a) Effect of nitrogen on the oxalate and calcium content of kikuyu grass (Pennisetum clandestinum Hochst). Journal of the Grassland Society of Southern Africa, 7, 106-110.

MARAIS, J.P. (1990b) Relationship between nitrogen and other chemical components in kikuyu grass from long established pastures. South African Journal of Animal Science, 20, 147-151.

MARAIS, J.P., FIGENSCHOU, D.L. and DENNISON, C. (1987) The accumulation of nitrate in Kikuyu grass (Pennisetum clandestinum Hochst). South African Journal of Plant and Soil, 4, 82-88.

MARAIS, J.P., THERION, J.J., MACKIE, R.I., KISTNER, A. and DENNISON, C. (1988) Effect of nitrate and its reduction products on the growth and activity of the rumen microbial population. British Journal of Nutrition, 59, 301-313.

MARAIS, J.P. and FIGENSCHOU, D.L. (1990) Influence of temperature and growth stage on the diurnal fluctuation in the non-structural carbohydrate content of kikuyu (Pennisetum clandestinum Hochst). Journal of the Grassland Society of Southern Africa, 7, 257-260.

MARAIS, J.P., FIGENSCHOU, D.L. and WOODLEY, G.A.J. (1990) Energy deficiency in kikuyu grass containing high levels of nitrogen. South African Journal of Animal Science, 20, 16-20.

MARAIS, J.P., FIGENSCHOU, D.L. and de FIGUEREIDO, M. (1992) Effect of nutrient calcium on the cell wall composition and digestibility of kikuyu grass. Journal of the Grassland Society of Southern Africa, 9, 72-75.

MARTINOVICH, D., MORTIMER, P.H. and di MENNA, M. (1972) Similarities between so-called kikuyu poisoning and two experimental mycotoxicoses. New Zealand Veterinary Journal, 20, 57.

MARTINOVICH, D. and SMITH, B. (1972) Kikuyu poisoning in sheep. New Zealand Veterinary Journal, 20, 169.

MARTINOVICH, D. and SMITH, B. (1973) Kikuyu poisoning of cattle. 1. Clinical and pathological findings. New Zealand Veterinary Journal, 21, 55-63.

McDOWELL, L.R. (1992) Minerals in Animal and Human Nutrition. (Academic Press: San Diego, California).

MEARS, P.T. (1970) Kikuyu-(Pennisetum clandestinum) as a pasture grass-A review. Tropical Grasslands, 4, 139-152.

MEARS, P.T. and HUMPHREYS, L.R. (1974a) Nitrogen response and stocking rate of Pennisetum clandestinum pastures. 1. Pasture nitrogen requirement and concentration, distribution of dry matter and botanical composition. Journal of Agricultural Science, Cambridge, 83, 451-467.

MEARS, P.T. and HUMPHREYS, L.R. (1974b) Nitrogen response and stocking rate of Pennisetum clandestinum pastures. 2. Cattle growth. Journal of Agricultural Science, Cambridge, 83, 469-478.

MEISSNER, H.H. and PAULSMEIER, D.V. (1988) The utilization by sheep of winter and spring smuts finger and kikuyu pastures. Journal of the Grassland Society of Southern Africa, 5, 115-121.

MENKE, K.H. and STEINGASS, H. (1988) Estimation of the energetic feed value obtained from chemical analysis and in vitro gas production using rumen fluid. Animal Research and Development, 28: 7-55.

MEYER, P. (1989) Thiaminase activities and thiamine content of Pteridium alquinum, Equisetum ramosissimum, Malva parviflora, Pennisetum clandestinum and Medicago sativa. Onderstepoort Journal of Veterinary Research, 56, 145-146.

MILES, N. (1991) Nutrition of intensive pastures in the summer rainfall areas of Southern Africa. Journal of the Grassland Society of Southern Africa, 8, 39-48.

MILES, N., McDONALD, C.I. and CUNNINGHAM, J.M. (1985) The effects of lime on the dry matter production and chemical composition of Pennisetum clandestinum on highly weathered soils. In Proceedings of the XV International Grassland Congress, Kyoto, 1985. pp. 489-491.

MILFORD, R. and MINSON, D.J. (1965) Intake of tropical pasture species. Proceedings of the IX International Grassland Congress, Sao Paulo, 1965. pp. 815-822.

MILNE, J.A. (1991) Diet selection by grazing animals. Proceedings of the Nutrition Society, 50, 77-85.

MILTHORPE, F.L. and DAVIDSON, J.L. (1966) Physiological aspects of regrowth in grasses. In: Milthorpe, F.L. and Ivins, J.D. (Eds.) The growth of cereals and grasses. (Butterworths: London). pp. 241-254.

MINSON, D.J. (1972) The digestibility and voluntary intake by sheep of six tropical grasses. Australian Journal of Experimental Agriculture and Animal Husbandry, 12, 21-27.

MINSON, D.J. (1973) Effect of fertilizer nitrogen on digestibility and voluntary intake of Chloris gayana, Digitaria decumbens and Pennisetum clandestinum. Australian Journal of Experimental Agriculture and Animal Husbandry, 13, 153-157.

MINSON, D.J. (1981) Nutritional differences between tropical and temperate pastures. In: Morley, F.H.W. (ed.) Grazing Animals. World Animal Science, B1. (Elsevier Science Publishers). pp. 143-157.

MINSON, D.J. (1982) Effect of chemical composition on feed digestibility and metabolisable energy. Nutrition Abstracts and Reviews-Series B, 52, 591-615.

MINSON, D.J. (1990) Forage in Ruminant Nutrition. (Academic Press: San Diego, California).

MINSON, D.J. and MILFORD, R. (1966) The energy value and nutritive value indices of Digitaria decumbens, Sorghum almum and Phaseolus atropurpureus. Australian Journal of Agricultural Research, 17, 411-423.

MINSON, D.J., COWAN, T. and HAVILAH, E. (1993) Northern dairy feedbase 2001. 1. Summer pasture and crops. Tropical Grasslands, 27, 131-149.

MOIR, K.W., WILSON, J.R. and BLIGHT, G.W. (1977) The in vitro digested cell wall and fermentation characteristics of grasses as affected by temperature and humidity during their growth. Journal of Agricultural Science, Cambridge, 88, 217-222.

MOIR, K.W., DOUGHERTY, H.G., GOODWIN, P.J., HUMPHREYS, R.J. and MARTIN, P.R. (1979) An assessment of whether energy was the first factor limiting production of dairy cows grazing kikuyu grass pasture. Australian Journal of Experimental Agriculture and Animal Husbandry, 19, 530-534.

MURTAGH, G.J. (1975) Environmental effects on the short-term response of tropical grasses to nitrogen fertiliser. Australian Journal of Experimental Agriculture and Animal Husbandry, 15, 679-688.

MURTAGH, G.J. (1978) The effect of evaporative demand on the growth of well-watered kikuyu. Agricultural Meteorology, 19, 379-389.

MURTAGH, G.J. (1987) Estimating the proportion of degradable dry weight of kikuyu from rates of lamina senescence. Annals of Botany, 59, 159-165.

MURTAGH, G.J. (1988a) Factors affecting the growth of kikuyu. 1. Potential growth and nitrogen supply. Australian Journal of Agricultural Research, 39, 31-42.

MURTAGH, G.J. (1988b) Factors affecting the growth of kikuyu. 2. Water supply. Australian Journal of Agricultural Research, 39, 43-51.

MURTAGH, G.J. (1990a) Variation in the supracritical nitrogen concentration of kikuyu grass. Communications in Soil Science and Plant Analysis, 21, 207-220.

MURTAGH, G.J. (1990b) Solar radiation and clear sky global turbidity on the north coast of New South Wales. Australian Meteorological Magazine, 38, 255-259.

MURTAGH, G.J., KAISER, A.G., HUETT, D.O. and HUGHES, R.M. (1980a) Summer growing components of a pasture system in a subtropical environment. 1. Pasture growth, carrying capacity and milk production. Journal of Agricultural Science, Cambridge, 94, 645-663.

MURTAGH, G.J., KAISER, A.G. and HUETT, D.O. (1980b) Summer growing components of a pasture system in a subtropical environment. 2. Relationships between pasture and animal production. Journal of Agricultural Science, Cambridge, 94, 665-674.

MURTAGH, G.J. and MOORE, K.G. (1987) A pasture budget for dairy farms in the Kyogle district. Tropical Grasslands, 21, 145-153.

MURTAGH, G.J., HALLINGAN, E.A. and GREER, D.H. (1987) Components of growth and dark respiration of kikuyu (Pennisetum clandestinum Chiov.) at various temperatures. Annals of Botany, 59, 149-157.

NEWSHOLME, S.J., KELLERMAN, T.S., van der WESTHUIZEN, G.C.A. and SOLEY, J.T. (1983) Intoxication of cattle on kikuyu grass followng army worm (Spodoptera exempta) invasion. Onderstepoort Journal of Veterinary Research, 50, 157-167.

OLNEY, G.R., McGHIE, D.A. and SPRIVULIS, R. (1982) Research report 1981/82. Diary Cattle Branch, Animal Production Division. Western Australia Department of Agriculture. p. 10.

OLNEY, G.R. and ALBERTSEN, T.O. (1984) Effect of supplementary concentrate on milk production from cows grazing irrigated kikuyu pastures. Proceedings of the Australian Society of Animal Production, 15: 509-512.

ØRSKOV, E.R. and McDONALD, I. (1979) The estimation of protein degradability in the rumen from incubation measurements weighted according to rate of passage. Journal of Agricultural Science, Cambridge, 92, 499-503.

ØRSKOV, E.R., REID, G.W. and KAY, M. (1988) Prediction of intake from degradation characteristics of roughages. Animal Production, 46, 29-34.

PASTRANA, R., McDOWELL, L.R., CONRAD, J.H., WILKINSON, N.S. and MARTIN, F.G. (1990) Mineral concentrations in leaves and stems of various forages of the Colombian paramo. Communications in Soil Science and Plant Analysis, 21, 2345-2360.

PATTINSON, N.B., BRANSBY, D.I. and TAINTON, N.M. (1981) Relationship between time spent grazing, quality, quantity and availability of herbage consumed by grazing animals on kikuyu and Coastcross II. Proceedings of the Grassland Society of Southern Africa, 16, 103-105.

PEARSON, C.J. and ISON, R.L. (1987) Agronomy of Grassland Systems. (Cambridge University Press: Cambridge, UK).

PEARSON, C.J., KEMP, H., KIRBY, A.C., LAUNDERS, T.E. and MIKLED, C. (1985) Responsiveness to seasonal temperature and nitrogen among genotypes of kikuyu, paspalum and bermuda grass pastures of coastal New South Wales. Australian Journal of Experimental Agriculture, 25, 109-116.

PEET, R.L., DICKSON, J. and HARE, M. (1990) Kikuyu poisoning in goats and sheep. Australian Veterinary Journal, 67, 229-230.

PELL, A. and SCHOFIELD, P. (1993) Computarised monitoring of gas production to measure forage digestion in vitro. Journal of Dairy Science, 76: 1063-1073.

PERCIVAL, N.S. (1978) Survey of kikuyu grass (Pennisetum clandestinum Hochst.) in New Zealand pastures. New Zealand Journal of Experimental Agriculture, 6, 19-21.

PHELOUNG, P. and BRADY, C.J. (1979) Soluble and fraction 1 protein in leaves of C3 and C4 grasses. Journal of the Science of Food and Agriculture, 30, 246-250.

PHILPOTTS, H. (1981) Poor nodulation of lupins and tropical legumes in northern New South Wales. Australian Journal of Experimental Agriculture and Animal Husbandry, 21, 588-594.

PIENAAR, J.P., TAINTON, N.M. and van RYSSEN, J.B.J. (1993a) Factors affecting the voluntary feed intake of sheep grazing Pennisetum clandestinum (kikuyu) pastures: observations from forage analysis. African Journal of Range and Forage Science, 10, 140-144.

PIENAAR, J.P., TAINTON, N.M., van RYSSEN, J.B.J. and SWIEGERS, J.P. (1993b) Factors affecting the voluntary feed intakes of sheep grazing Pennisetum clandestinum (kikuyu) pastures: observations in the animal. African Journal of Range and Forage Science, 10, 145-150.

PIGGOT, G.J. (1985) Kikuyu grass in Northland: a summary. In: Kikuyu grass farming for high production. Proceedings of a 'Kikuyu grass forum' held at Whangarei, New Zealand. New Zealand Grassland Association. pp. 59-60.

PIGGOT, G.J. (1988) Herbage accumulation patterns of pastures at various sites in Northland. Proceedings of the Agronomy Society of New Zealand, 18: 81-84.

PIGGOT, G.J. (1991) Kikuyu grass farming for high production. Northland Pasture. Whangarei, New Zealand. 40 pp.

PIGGOT, G.J. and MORGAN, H.M. (1984) Autumn undersowing of grazed kikuyu grass (Pennisetum clandestinum) pasture. New Zealand Journal of Experimental Agriculture, 12, 269-271.

PIGGOT, G.J. and MORGAN, H.M. (1985) Local ecotypes compared with Australian cultivars. In: Kikuyu grass farming for high production. Proceedings of a 'Kikuyu grass forum' held at Whangarei, New Zealand. New Zealand Grassland Association. pp. 45-46.

PIGGOT, G.J. and MORGAN, H.M. (1986) A genetic basis for male sterility in kikuyu grass. Tropical Grasslands, 20, 34-36.

PINKERTON, A. and RANDALL, P.J. (1993) A comparison of the potassium requirements during early growth of Lotus penduculatus, Medicago murex, M. polymorpha, M. truncatula, Ornithopus compressus, Trifolium balansae, T. resupinatum, Pennisetum clandestinum, and Phalaris aquatica. Australian Journal of Experimental Agriculture, 33, 31-39.

PRESTON, T.R. and LENG, R.A. (1987) Matching Ruminant Production Systems with Available Resources in the Tropics and Sub-tropics. (Penambul Books: Armidale, Australia).

PUNIA, B.S., LEIBHOLZ, J. and FAICHNEY, G.J. (1984) Protozoal numbers in the rumen and omasum of cattle fed kikuyu grass. Proceedings of the Australian Society of Animal Production, 15, 733.

PUNIA, B.S. and LEIBHOLZ, J. (1994) Effect of level of intake of kikuyu (Pennisetum clandestinum) grass hay on the flow of protozoal N to the omasum of cattle. Animal Feed Science and Technology, 47, 77-88.

QUINLAN, T.J., SHAW, K.A. and EDGLEY, W.H.R. (1975) Kikuyu grass. Queensland Agricultural Journal, 101, 737-749.

RAMIREZ, L.M., LAREDO, M.A., ANZOLA, H.J. and MARTINEZ, O. (1983) Estimación del consumo voluntario de kikuyu (Pennisetum clandestinum Hochst) y raigrás tetrelite (Lolium hybridum Hausskn) por bovinos en pastoreo mediante el uso de dos marcadores externos. Revista ICA, 18, 1, 35-43.

REES, M.C. and LITTLE, D.A. (1980) Differences between sheep and cattle in digestibility, voluntary intake and retention time in the rumen of three tropical grasses. Journal of Agricultural Science, Cambridge, 94, 483-485.

REEVES, M., FULKERSON, B., KELLAWAY, R. and BRYANT, D. (1993) Milk production from Pennisetum clandestinum pastures. Dairy Research Report 1992-1993. Wollongbar Agricultural Institute. pp. 8-10.

REEVES, M., FULKERSON, B., KELLAWAY, R. and BRYANT, D. (1994) Supplementation of dairy cows grazing kikuyu pastures. Dairy Research Report 1993-1994. Wollongbar Agricultural Institute. pp. 32-35.

REID, R.L., POST, A.J. and OLSEN, F.J. (1979) Chemical composition and quality of tropical forages. Bulletin 669T. Agricultural and Forestry Experimental Station, West Virginia University. 43 pp.

REID, T.C. and STEVENSON, L.C. (1983) Laboratory assessment of the nutritive value of a pasture containing kikuyu grass (Pennisetum clandestinum Hochst.) Proceedings of the Nutrition Society of New Zealand, 8, 143-144.

RETHMAN, N.F.G. (1989) The influence of ruminant digestion on the germination of Pennisetum clandestinum Hochst.Ex Chiov. Journal of the Grassland Society of Southern Africa, 6, 14.

RETHMAN, N.F.G. and de WITT, C.C. (1988) The yield potential and crude protein content of five rhizomatous and stoloniferous grass pastures in the escarpment areas of the eastern Transvaal. South African Journal of Plant and Soil, 5, 222-224.

ROYAL, A.J.E. and HUGHES, R.M. (1976) Winter-spring forage crops as supplements for dairy cows grazing kikuyu pasture. Proceedings of the Australian Society of Animal Production, 12, 513-516.

RUKANTABULA, A.A. and KUSEKWA, M.L. (1990) Nutritive value and some agronomical aspects of selected pasture species at Mpwapwa-Tanzania by the end of the dry season. Bulletin of Animal Health and Production in Africa, 38, 131-138.

RUMBALL, P.J. (1985) Sheep production. Trial results. In: Kikuyu grass farming for high production. Proceedings of a 'Kikuyu grass forum' held at Whangarei, New Zealand. New Zealand Grassland Association. pp. 15-17.

RUMBALL, P.J. (1991) Performance of several subtropical grasses in Northland hill pastures. New Zealand Journal of Agricultural Research, 34, 375-382.

RUMBALL, P.J. and BOYD, A.F. (1980) Comparison of ryegrass-white clover pastures with and without paspalum and kikuyu grass. New Zealand Journal of Experimental Agriculture, 8, 21-26.

RUMBALL, P.J. and LAMBERT, J.P. (1985) Local ecotypes compared with Kenyan accessions. In: Kikuyu grass farming for high production. Proceedings of a 'Kikuyu grass forum' held at Whangarei, New Zealand. New Zealand Grassland Association. pp. 42-44.

RUSSELL, E.W. (1973) Soil Conditions and Plant Growth. 10th Edn. (Longman: London).

RUSSELL, J.S. (1976) Comparative salt tolerance of some tropical and temperate legumes and tropical grasses. Australian Journal of Experimental Agriculture and Animal Husbandry, 16, 103-109.

RUSSELL, M.J. and KLEINSCHMIDT, F.H. (1984) Performance of pasture grass cultivars and their associates at Queensland Agricultural College. Tropical Grasslands, 18, 55-61.

SANCHEZ, L., MENDOZA, P.E. and URBINA, N. (1983) Suplementación con henos producidos en el piedemonte llanero en el levante de novillas Holstein durante la época de verano en la sabana de Bogotá. Revista ICA, 18, 1, 45-51.

SANCHEZ, J.M., CAMPABADAL, C., VARGAS, E. and FONSECA, E. (1986) Contenido proteico y mineral en los forrajes de la zona montañosa central de Costa Rica. 2. Efecto de la especie. Agronomía Costarricense, 10, 1-2, 191-197.

SCA (1990) Feeding standards for Australian Livestock. Ruminants. Standing Committee on Agriculture. (CSIRO Publications: Australia).

SCHIERE, J.B., SILVA, K.A.I.D., BABURDEEN, A.L. and IBRAHIM, M.N.M. (1990) Dry matter intake and digestibility of kikuyu grass (Pennisetum clandestinum) offered at increasing levels. Indian Journal of Animal Nutrition, 7, 311-314.

SHERRELL, C.G. (1978) A note on sodium concentrations in New Zealand pasture species. New Zealand Journal of Experimental Agriculture, 6, 189-190.

SINGH, B. and NARANG, M.P. (1991) Rumen available carbohydrate content of various forages. Indian Journal of Animal Nutrition, 8, 93-96.

SINGH, B., NARANG, M.P. and MAKKAR, H.P.S. (1992) Rate and extent of in sacco degradation of dry matter and cell wall constituents of various forages. Indian Journal of Animal Nutrition, 9, 1-12

SMITH, B. and MARTINOVICH, D. (1973) Kikuyu poisoning of cattle. 2. Epizootiological aspects. New Zealand Veterinary Journal, 21, 85-89.

SMITH, G.S., MIDDLETON, K.R. and EDMONDS, A.S. (1980) Sodium nutrition of pasture plants. 1. Translocation of sodium and potassium in relation to transpiration rates. New Phytologist, 84, 603-612.

SOTO, L., LAREDO, M.A. and ALARCON, E. (1980) Digestibilidad y consumo voluntario del pasto Kikuyo (Pennisetum clandestinum Hochst) en ovinos bajo fertilización nitrogenada. Revista ICA, 15, 2, 79-90.

STOBBS, T.H. (1971) Quality of pasture and forage crops for dairy production in the tropical regions of Australia. 1. Review of the literature. Tropical Grasslands, 5, 159-170.

STOBBS, T.H. (1972) Milk production from tropical grasses. a. Nitrogen fertilised grasses. Annual Report 1971-72, Division of Tropical Pastures, CSIRO, Australia. pp. 68-70.

STOBBS, T.H. (1973a) The effect of plant structure on the intake of tropical pastures. 1. Variation in the bite size of grazing cattle. Australian Journal of Agricultural Research, 24, 809-819.

STOBBS, T.H. (1973b) Milk production from tropical grasses. b. Nitrogen fertilised grasses. Annual Report 1972-73, Division of Tropical Agronomy, CSIRO, Australia. p. 72.

STOBBS, T.H. (1974) Components of grazing behaviour of dairy cows on some tropical and temperate pastures. Proceedings of the Australian Society of Animal Production, 10, 299-302.

STOBBS, T.H. (1975) Factors limiting the nutritional value of grazed tropical pastures for beef and milk production. Tropical Grasslands, 9, 141-150.

TAINTON, N.M., BARROW, K.H. and BRANSBY, D.I. (1982) An economic evaluation of steer livemass gain from kikuyu differentially fertilized with nitrogen. Proceedings of the Grassland Society of Southern Africa, 17, 29-32.

TAYLOR, A.O., HASLEMORE, R.M. and McLEOD, M.N. (1976a) Potential of new summer grasses in Northland. III. Laboratory assessments of forage quality. New Zealand Journal of Agricultural Research, 19, 483-488.

TAYLOR, A.O., ROWLEY, J.A. and HUNT, B.J. (1976b) Potential of new summer grasses in Northland. I. Warm season yields under dryland and irrigation. New Zealand Journal of Agricultural Research, 19, 127-133.

TEITZEL, J.K., GILBERT, M.A. and COWAN, R.T. (1991) Sustaining productive pastures in the tropics. Nitrogen fertilised grass pastures. Tropical Grasslands, 25, 111-118.

van der GRINTEN, P., BAAYEN, M.T., VILLALOBOS, L., DWINGER, R.H. and MANNETJE, L. 't (1992) Utilisation of kikuyu grass (Pennisetum clandestinum) pastures and dairy production in a high altitude region of Costa Rica. Tropical Grasslands, 26, 255-262.

van HEERDEN, J., WILLIAMS, M.C., van RENSBURGH, I.B.J. and IPLAND, F.F. (1978) An outbreak of 'kikuyu poisoning' in Western Transvaal. Journal of the South African Veterinary Association, 49, 27-30.

van RYSSEN, J.B.J., SHORT, A.M. and LISHMAN, A.W. (1976) Carbohydrate supplementation of lambs on kikuyu pasture. Agroanimalia, 8, 43-48.

van SOEST, P.J. (1982) Nutritional Ecology of the Ruminant. (O&B Books: Corvallis, Oregon).

WENG, JEN-HSIEN (1988) Leaf photosynthesis of forage grasses in different seasons and temperatures. Botanical Bulletin Academia Sinica, 29, 163-170.

WHITNEY, A.S. (1974a) Effects of gibberelic acid on the cool season yields of kikuyugrass and pangolagrass. Proceedings of the XII International Grassland Congress, Moscow, 1974. pp. 588-597.

WHITNEY, A.S. (1974b) Growth of Kikuyugrass (Pennisetum clandestinum) under clipping. 1. Effects of nitrogen fertilization, cutting interval, and season on yields and forage characteristics. Agronomy Journal, 66, 281-287.

WHITNEY, A.S. (1976) Effects of giberrellic acid on the cool season regrowth of two tropical forage grasses. Agronomy Journal, 68, 365-370.

WHITNEY, A.S. and TAMIMI, Y.N. (1974) Efficiency of broadcast application of urea to subtropical pastures. Proceedings of the XII International Grassland Congress, Moscow, 1974. pp. 580-587.

WILLIAMS, M.C. (1987) Nitrate and soluble oxalate accumulation in kikuyu grass (Pennisetum clandestinum). Proceedings of the Western Society of Weed Science, 40, 78-79.

WILLIAMS, M.C., SMITH, B.J. and LOPEZ, R. (1991) Effect of nitrogen, sodium and potassium on nitrate and oxalate concentration in kikuyu grass. Weed Technology, 5, 553-556.

WILSON, G.P.M. (1970) Method and practicability of kikuyu grass seed production. Proceedings of the XI International Grassland Congress, Surfers Paradise, Queensland, 1970. pp. 312-315.

WILSON, G.P.M. (1975) Varieties of kikuyu. Agricultural Gazette of New South Wales, 86, 8.

WILSON, G.P.M., DALE, A.B. and SPURRS, A.B. (1975) Kikuyu seed production. Agricultural Gazette of New South Wales, 86, 9.

WILSON, G.P.M. and RUMBLE, C.J. (1975) The effect of seed rate and nitrogen fertiliser on the yield of seed and by-product leaf of Whittet kikuyu grass at Grafton, New South Wales. Tropical Grasslands, 9, 53-58.

WILSON, G.P.M. and HENNESSY, D.W. (1977) The germination of excreted kikuyu grass seed in cattle dung pats. Journal of Agricultural Science, Cambridge, 88, 247-249.

WILSON, J.R. (1991) Plant structures: their digestive and physical breakdown. In: Ho, Y.W., Wong, H.K., Abdullah, N. and Tajuddin, Z.A. (Eds.) Recent Advances on the Nutrition of Herbivores. (Malaysian Society of Animal Production). pp. 207-216.

WILSON, J.R. (1994) Cell wall characteristics in relation to forage digestion by ruminants. Journal of Agricultural Science, Cambridge, 122, 173-182.

WILSON, J.R., SANDLAND, R.L. (1976) The relation between growth and nitrogen and phosphorus content of green panic and kikuyu grass. Plant and Soil, 44: 341-358.

WILSON, J.R., TAYLOR, A.O. and DOLBY, G.R. (1976) Temperature and atmospheric humidity effects on cell wall content and dry matter digestibility of some tropical and temperate grasses. New Zealand Journal of Agricultural Research, 19, 41-46.

WONG, P.T.W., ROTH, I.J. and JACKSON, A.R.B. (1987) Kikuyu poisoning of cattle in New South Wales and its relationship to pasture fungi on kikuyu grass (Pennisetum clandestinum). Australian Veterinary Journal, 64, 229-232.

WOOMER, P., ASANO, W. and PRADHAN, D. (1990) Environmental factors related to rhizobial abundance in kikuyugrass (Pennisetum clandestinum) pastures. Tropical Agriculture (Trinidad), 67, 217-220.

ZACHARIAS, P.J.K., CLAYTON, J. and TAINTON, N.M. (1991) Leucaena leucocephala as a quality supplement to Pennisetum clandestinum foggage: A preliminary study. Journal of the Grassland Society of Southern Africa, 8, 59-62.