5. PASTURE MANAGEMENT

5.1 General Observations

Whether the pasture consists of native or exotic species, good management is necessary to obtain maximum production and maintain pasture productive capacity. Both types of pasture respond to good grazing control and adequate rest periods. However, the establishment of improved pastures described in the previous chapter involves the farmer's time, money and effort and initially a different range of management procedures (Williamson and Payne, 1978). In order to capitalize on his considerable investment it is important that the farmer devotes maximum skills to the management phase. Failure to do so will result either in establishing a poor quality improved pasture or worse, a weed covered paddock where local species of lower feed value will regrow.

There are two distinct periods in management:

• during establishment following planting;
• after establishment, or in established pastures.

5.2 Management during establishment

Where seeds have been sown Cook (1980) suggests there are two phases in establishment: germination and emergence followed by seedling growth and survival. Whether seeds, cuttings or pieces have been used the effects of weather, seed and planting material characteristics and seedbed conditions are likely to outweigh any management practices in the early stages (see Miller et al., 1993). Once rapid shoot growth takes place then management practices can influence the establishing pasture.

5.2.1 Early grazing

Cattle can be used to assist pasture establishment. With the stoloniferous species, after 2–3 months of pasture growth, or later, depending on climate, soil type, pasture species and rate of establishment (for example, Watson and Whiteman, 1981a, carried out a first light grazing after 5 or 6 months while Guzman and Allo, 1975, suggest that Para grass should be given an initial grazing within 6 to 10 weeks of planting), animals can be used at a high stocking rate to lightly graze the pasture but only for a short time (e.g. 15 animals ha^-1 for one day only) to avoid any selective grazing. Evans et al. (1992) suggest that for fully sown weed free pastures, the first grazing should occur 10–12 weeks after planting or when grasses are about 50–100 cm high in the case of tufted upright growing species, e.g., guinea grass, or at 30–50 cm for stoloniferous species such as signal or Koronivia. Tufted grasses should be grazed down to a height of about 30–50 cm and creeping grasses down to a height of 15–20 cm. Pastures should again be grazed 8–10 weeks later. As a general guide the stocking rate used in the first 6 to 12 months of pasture establishment should be about 40 percent of the optimum stocking rate for a fully established pasture. The action of grazing and trampling causes the grass to spread and to send out new stolons, thus reducing competition of the grasses on the legume component. This practice can be repeated a number of times so that gradually the paddock is brought into full production. Teitzel (1992) indicates that early grazing management is critical as the pasture is competing with weeds and there is no “seed bank” in the soil to allow plants to regrow if they are damaged. Cook and Ratcliff (1992) indicated that control or management of plant competition during the first three months of establishment was a key factor in determining pasture establishment and productivity. The important factor is neither to graze too soon nor too long (Reynolds, 1978j; Teitzel and Middleton, 1978). The first grazing given to an established pasture will have a considerable influence on the performance of the sward throughout its life. Heavy grazing will seriously weaken the pasture species, particularly legumes like Centro and Puero, and encourage the development of weeds (Middleton and Teitzel, 1978); on the other hand very light grazing wastes feed, resulting in only selective grazing and may not lead to the desired spreading of new runners. The purpose of grazing management, to suppress the grass, should be to encourage legume growth. However, a pasture should not be grazed until both grasses and legumes have established strong root systems thus avoiding damage by the grazing animal (Steel et al., 1980). The purpose of good early management is to conclude the cycle with a dense, permanent sward of the desired species after a year or so, and over the first three years the objective is to obtain a high legume content because the amount of nitrogen fixed by the legume is directly proportional to the quantity of legume in the pasture.

Where pasture emergence has been poor due to poor quality seed, poor soil/seed contact, unexpected drought or excessively prolonged wet conditions, then complete establishment may take two years. Where there is a severe weed problem grazing may have to be delayed for six months or so to allow climbing legumes to smother weeds (see Figure 127), but there is also the risk of the sown grass being smothered so careful grazing management is required. Sometimes if weeds are dominating in the early stages slashers may be used, however, pastures should not be slashed any shorter than 30 cm so that legumes are not damaged. The manager has to decide whether the benefits outweigh the costs (and Evans et al., 1992 suggest that the true cost of light slashing is of the order of 1,500 VT ha^-1 (US$ 12.50) for operating costs such as fuel, oil, repair and labour, plus 500 VT ha^-1 (US$ 4.20) depreciation on the machinery). Slashing is often followed by spot spraying of the weed regrowth (Evans et al., 1992 indicate that in Vanuatu the cost of labour to hand weed and 2, 4-D to spotspray and completely eradicate Solanum torvum during the first year of establishment usually costs from 1,000–2,000 VT ha^-1). There are many examples, however, of well established pastures completely dominating various weeds within nine months of planting without any slashing (Evans et al., 1992; Reynolds 1978k).

Figure 127. - The weed Cassia tora being smothered by climbing legumes on Reunie Plantation, Malekula, Vanuatu.

Pasture longevity is important if the mixed enterprise is to be run at the lowest possible cost. Overgrazing in the establishment phase may result in the need for pasture renewal after only 5 or 6 years (Guzman and Allo, 1975).

5.2.2 Weed control

Weeds will begin to appear in the newly established pasture in the first few months. Practices must be adopted to limit their development, so that by the second year a uniform sward has developed where weeds are only a minor problem (Jones, 1975). Efforts made at this stage will limit the problem later on.

The careful manager will quickly suppress outbreaks of new weeds as soon as they occur as well as gradually suppressing the weeds already present by preventing them from seeding and spreading. It is important to realize that single plants of, for example, pico (Solanum torvum) and peanut weed or pistache (Cassia tora) if left to mature, can produce hundreds of seeds which will germinate over a period of several years (Evans et al., 1990).

It is also important to realize that weeds are often introduced to otherwise clean farms in contaminated pasture seed or planting material (Tarilongi, 1993), by grazing animals or in soil on uncleaned machinery (Evans et al., 1990). Introduced weeds tend to radiate from farm roads, holding areas and stockyards. Care should be taken in using only certified seed, quarantining new stock in specific holding paddocks and thoroughly washing machinery from contaminated areas.

There are four main weed control methods:

1. Management - Weed problems can be minimized with good seed-bed preparation, use of fertilizer ensuring early establishment of well adapted species and good grazing management (Steel et al., 1980). Some grass species are better than others at smothering weeds. For example, B. brizantha is a particularly aggressive grass species that, given time, will overcome many weeds. Thus a 60 percent weed component after 6 months may be reduced to 10 percent after 24 months, or with ideal conditions and early management B. brizantha pastures may have only a 20 percent weed component after 6 months (Reynolds, 1978k). Wu Renun and Xu Xuejun (1993) reported that the establishment of signal grass (B. decumbens) in an area infested with the noxious weed Chromolaena odorata resulted in the decline of the weed population and within three years there was no C. odorata in the pastures. This illustrates the importance of selecting the right species and giving them the right conditions.

   The most important factor in successful long-term weed management is the maintenance of vigorous grass/legume pastures under grazing. Pasture weeds are usually a reflection of the previous management of the area. Evans et al., 1990 suggest that annual weeds are mostly opportunists that germinate when the soil is at least partially bared by seasonal conditions or following overgrazing, mowing or burning. Where overgrazing results in the dominance of a single weed species this is usually easier to control than invasion by several different species. Trends towards weed dominance should be recognized as soon as possible, and steps taken to reverse the weed build up by examining and possibly modifying management practices. Remedial measures may include grazing or cutting of weeds when young, strategic use of herbicides, effective use of fertilizer and the introduction of new competitive pasture species.

   According to Mullen and Banga (1993) a vigorous pasture maintained at the correct grazing height is the best insurance against weed invasion. “The fact that, often, more money can be made from less cattle rather than more cattle grazing a particular pasture is a concept that graziers the world over have difficulty with. This is especially true in Vanuatu where weed species are quick to invade overgrazed pastures and once weed invasion has commenced there is a rapid decrease in economic returns as cattle have progressively less pasture on offer”. They stress optimal grazing heights and sustainable stocking rates. In Vanuatu on one large plantation on Malekula spraying of Cassia tora has been abandoned in favour of better pasture management - it was found that when buffalo grass pastures were slashed and destocked for some months (3 months) then the resultant dense stand of Buffalo grass (see Figure 128) served to prevent the Cassia tora seed germinating once the existing stand had died back. Even though a considerable burden of weed seeds exist in the soil they will not germinate if a dense pasture stand is maintained. The manager of Plantation Reunie planned to stock more lightly in future at around 1.0 AU ha^-1 (and no higher than 1.5 AU ha^-1) to maintain a better stand of grass at a height of 30 cm. It was expected that liveweight gains of 0.50 kg head^-1 day^-1 should be possible.

   

   Figure 128. - A dense stand of Buffalo grass after 3 months of destocking to control Cassia tora.

   In situations of complete weed dominance “biological” control through the establishment of robust, creeping legumes and grasses and the use of smothering legumes to eradicate the target weed through shading, should be tried. (Figure 127).

   Michael (1970) emphasized that the control of weeds in grasslands by chemicals or biological control agents may be temporary or of little use if pasture improvement is not carried out at the same time. Miller (1992) noted that Mimosa pigra has been controlled with a cut stump herbicide treatment together with the sowing of calopo (C. mucunoides). He also reported on an experiment where Koronivia grass (B. humidicola) was established in pots at densities ranging form 0 to 16 plants per pot together with seedlings of M. pigra. Various growth characteristics of mimosa were reduced under competition from Koronivia grass illustrating the likely field effect of an aggressive grass species on weeds (see Figure 129).

2. Mechanical - can be carried out either by hand or using machinery. Removal of weeds, especially woody species, can be performed by hand pulling (see Figure 130) with bushknife, panga or spade (see Figure 131) where labour is readily available. However, although unskilled labour can be used, and no large investment in capital equipment is required, hand weeding has many disadvantages and is only possible on pastures where labour is inexpensive as it is very time consuming and simply slashing the tops of perennial weeds may be ineffective, unless they are slashed repeatedly, because of resprouting and regrowth. Manual weeding is the usual method of smallholder weed control. Particularly on large flat stone free paddocks use of a mower to slash the weeds, a tractor drawn set of discs, a roller, roller-chopper or weight dragged behind a tractor to crush and break the weeds, (see Figure 106) will allow pasture to grow over weeds. On rocky areas a mechanical chain slasher is preferable to a blade slasher. Mowing or slashing can be used to remove unpalatable or inedible weeds after stock have selectively grazed a paddock, and to prevent them taking a competitive advantage over the more desirable species in the pasture. Where weeds have outcompeted pasture species during the pasture establishment phase mowing may reduce weed vigour and allow sufficient light to emerging pasture species for rapid growth (Evans et al., 1990). It can also be used to prevent tall growing annual and perennial weeds from flowering. Slashing may be used one to three weeks prior to herbicide spraying to encourage lush new weed growth for maximum herbicide effect.

   

   Figure 129. - The effect of Koronivia grass density on the dry matter yield of mimosa seedlings, 12 weeks after sowing the mimosa seed. (Points followed by a common letter are not significantly different at the 5% level using the LSD based on log_e transformed yields). After Miller (1992).

3. Biological - biocontrol involves the use of natural organisms that feed on the many unwanted weed species. For example, the leaf mining hispid beetle (Uroplata giradi) which feeds on lantana (L. camara) leaves will eventually cause a decline in the population of lantana (Partridge, 1977; Reynolds, 1978k). In Western Samoa two insects Scamurius sp. and Heteropsylla sp. have been introduced (in 1988) to control giant sensitive plant, Mimosa invisa, with Heteropsylla giving promising results (Anon. 1989) - see Figure 132. In Papua New Guinea Mimosa invisa was controlled very successfully by Heteropsylla spinulosa and application of nitrogen fertilizer appeared to boost its numbers (SPC, 1994). In Sri Lanka the leaf eating Arctiid insect Ammalo insulata was introduced in 1973 to control the troublesome weed Chromolaena odorata but did not persist (Liyanage, 1984). Biological agents are also being evaluated in the South Pacific region for Mikania micrantha (Evans et al., 1990). Due to its nature this method is used by Departments of Agriculture and not by individual farmers.

   

   Figure 130. - Woody weed species hand pulled in an establishing pasture, Vailele, Western Samoa.

4. Chemical - herbicides to control weeds. Skerman (1977) has listed some of the main herbicides and methods of application. The effect of herbicide on a weed species is influenced by species characteristics, the amount of herbicide applied, the time of application in relation to age and stage of growth of the weed and the environmental conditions at and after time of application. The absence of rain for at least three hours after application for foliar applied herbicides is especially important. Herbicides are classified according to whether they kill all plant material (non-selective) or only specific species (selective). Foliar applied herbicides may be selective or non-selective and can be divided into those that kill on contact (e.g. paraquat) or those that are translocated through plant tissue into the root system (e.g. glyphosate and MSMA) and kill the whole plant (Evans et al., 1990).

   

   Figure 131. - A smallholder and his wife handweeding pico (Solanum torvum) infested pastures at Epule on Efate, Vanuatu (Photo D. MacFarlane).

   The amount of light in tree plantations has been found to affect some herbicidal activity (Headford, 1970). High light intensity reduces herbicide efficiency through photo decomposition of the herbicide. Whereas some herbicides obtain greater penetration of leaf surfaces in the dark (i.e., when sprayed late in the day) others (e.g., 2,4-D) are absorbed more in strong light than darkness (Evans et al., 1990).

   Because of the risk of killing pasture legumes and the toxic nature of many of the chemicals used and because of possible effects on the main tree crop (Ackerson and Davis, 1987; Aya and Fayemi, 1982; Isherwood and Teo, 1987; Khairudin and Teoh, 1987; Komolafe, 1978; Liyanage, 1984; Romney, 1963, 1965; Wong, 1982); herbicides should be used with care and usually for spot spraying or painting (see Figure 133) only on bush weeds such as L. camara, Psidium guajava (see Figure 134), Stachytarpheta sp., Sida sp. etc.. A list of some recommendations was given by Reynolds (1988). A more comprehensive list is given in Table 78 while for full details reference should be made to Evans et al. (1990). However, it is interesting that even though recommendations are available Mullen and Banga (1993) stress that in Vanuatu very few weeds are controlled by using herbicides (grazing management, and in particular adoption of correct stocking rates and vigorous, improved pastures are more important for effective weed control). Herbicides are commonly applied with manually operated knapsacks or tractor operated boom sprays but low volume spinning disc applicators-controlled droplet application (Bakri et al., 1987) and wick applicators are also used. The use of pre-planting herbicides should also be assessed depending on their availability, the available equipment and expertise.

   Weed control in tropical pastures has been discussed by Skerman (1977). Common weeds and weed control methods for coconut areas in the Solomon Islands, Western Samoa, the South Pacific and South-East Asia have been described by Anon. (1977c); Evans, 1995; Guzman and Allo (1975); Lambert (1970); Reynolds, (1978b, 1978k) and Steel et al., (1980). Liyanage (1984) reviewed the major weeds and their control in Sri Lankan coconut areas.

   Figure 132. - Early biocontrol trials in Western Samoa to control giant sensitive plant (Mimosa invisa).

   

   Figure 133. - Cut stump control of large Pico (Solanum torvum) on Efate, Vanuatu (Photo D. MacFarlane).

   Unpalatable weed invasion is a major problem in frequently overgrazed smallholder pastures under coconuts. Weeds such as Cassia tora, Stachytarpheta, Elephantopis, Psidium, Mimosa invisa, (see Figure 135) Nephrolepis and Sida are common (MacFarlane and Shelton, 1986; Litscher and Whiteman, 1982; Shelton et al., 1986).

   Stephen Lee (personal communication) controlled Honolulu Rose (Clerodendrum fragans) in Western Samoa (see Figure 136) with Escort. At a rate of 0.5 g l^-1 water young plants <0.5 m were completely controlled and regrowth of mature plants was sprayed 3–4 weeks after slashing.

   Recently a very comprehensive technical bulletin on “Weed identification and management in Vanuatu pastures” has been published (Evans et al., 1990) which deals with the identification and management of pasture weeds both in open situations and plantations and in smallholder systems of production. Some forty two weeds (with colour plates) and control methods are listed and appendix 1 summarizes herbicides and rates and methods of application. This publication should be widely used throughout the South Pacific. An updated second edition was published in May 1993 (Mullen et al., 1993). Also refer to a paper on “Proven strategies for managing major weeds in Vanuatu” (Mullen and Banga, 1993). They stress that in most cases overgrazing is the primary problem and weeds are merely the result of this major problem! For further details see Mullen (1993).

Chee and Ahmad Faiz (1991) reviewed weed control methods in Malaysian rubber estates and showed that the complementary use of grazing sheep reduces the overall costs of weed control by between 16 and 36 percent. Sheep selectively graze grass and some palatable broadleafed species resulting in a desirable purification of the legume Calopogonium caeruleum in cover crop mixtures under rubber.

Figure 134. - Guava (Psidium guajava); a troublesome bush weed on many Pacific Islands (Photo D. MacFarlane).

Yeoh et al. (1986) compared the economics of different weeding methods under two-year old rubber (see Table 79) and Mohd Najib (1991) indicated that animal integration can result in savings on weeding cost of 15–25%. With good grazing management, savings of up to 50% may be obtained (Wan Mohamed, 1986). Tajuddin et al. (1990) estimated that the grazing of sheep in young rubber plantations resulted in a saving of approximately 30% of the costs of chemical weed control.

Table 78. - Chemical control methods for some major weeds found in pastures under coconuts (from Evans et al., 1990).

B = Boom spraying low volume;
S = Spot spraying high volume;
SV - spray volume
% - herbicide percentage in spray solution

Bishop et al. (1993) presented a number of case studies and identified the causes of weed invasion of sown pastures on the wet tropical coast of eastern Australia. The major causes suggested for pasture decline and invasion by less productive weedy species were:

1. disease which can suddenly wipe out otherwise well adapted species;

2. inappropriate sown pasture species selected for the particular soil, climatic or weed situation;

3. failure to act quickly and decisively to eradicate weed species as soon as they appear and before soil seed reserves build up;

   Figure 135. - Giant mimosa (Mimosa invisa).

   Figure 136. - Honolulu rose (Clenodendrum fragrans).

   

   Figure 137. - Navua sedge (Kyllingia polyphylla; syn. Cyperus aromaticus).

   Figure 138. - Sword fern (Nephrolepis hirsutula).

4. over-utilization (overgrazing) of sown species creating a situation where less palatable species gain the advantage to establish and spread;

5. failure to adjust stocking rate annually in line with growing season pasture dry matter growth;

6. lack of knowledge or experience by producers of the principles of pasture growth, pasture management and assessment of pasture condition whereby stocking rate is judged on animal condition alone without including an assessment of the condition of the pastures.

These are equally appropriate to pastures under tree crops with the addition in (ii) of the environmental (shade) factor.

Table 79. - Economics of different weeding methods (Yeoh et al., 1986)

5.2.3 Fertilizer application

After 6 months, a second dressing of fertilizer should be applied (independently of fertilizer being applied to the coconuts). Again, 100 kg ha^-1 of superphosphate or potassic superphosphate might be appropriate, but will depend upon soil, climatic and other conditions in the particular area. It may be preferred to apply fertilizer more frequently, splitting the annual application into 4 treatments, so that the first application can be made after the initial early grazing (see section 5.2.1).

5.2.4 Time required for establishment

When a good seed-bed is prepared and cuttings are closely spaced, good establishment, under wet season conditions, may result in moderate grazing after 3 months and full grazing after 6 months, however, as a general rule full grazing is likely to be available after about 12 months (see Figure 139). The sooner full grazing can be undertaken, the quicker pasture establishment costs begin to be repaid (Javier, 1974a). A long establishment phase means that coconut production can be affected both through nuts lost in the long grass and possibly through reduced production because of competition from the ungrazed pasture (Ferdinandez, 1970b).

Figure 139.
-   Establishment of Palisade grass (B. brizantha)

a) One to two weeks after establishment

b) One and a half months after establishment

c) One year after establishment

5.3 Management of established pastures

5.3.1 Good pasture management

The main aim is to strive to:

1. Obtain maximum herbage yield with the highest possible nutritive value throughout the year at the lowest possible cost, with no reduction in coconut yields.

2. Keep pasture productive and prevent any overall decline in quality. An optimum stocking rate is one that will maintain pasture stability and botanical composition, and produce consistent levels of animal production (Evans et al., 1992).

3. Maintain a good grass-legume balance.

4. Convert the feed to saleable products such as meat and milk.

Unless the pastures which are grown are efficiently utilized and converted to animal products then the farmer will not get good returns on his investment. The various stages of plant and animal production in grazing systems are shown in Figure 140.

Figure 140. - Stages of plant and animal production in grazing systems. (After Hodgson, 1990).

5.3.2 Grass-legume balance

Although Whiteman (1980) and Walker (1975) have suggested that there is no optimum level of legume in a pasture, Whiteman (1980) emphasized the need for a good species balance, the selection of appropriate species and good pasture management in order to achieve high yields of good quality forage. Bryan and Evans (1968) indicated that at least 30 percent of legumes should be maintained in a pasture for high production. Evans et al. (1992) suggested that good pasture management should aim to maintain at least a 20–30 percent legume component (see Figure 141). Evans (1970) demonstrated that liveweight gains of beef cattle were linearly related to pasture legume content where this ranged from about 10 to 40 percent (Whiteman, 1980). Walker (1975) suggested that in most situations, mean annual legume contents in excess of 30 percent are difficult to achieve, although exceptionally during the year 50 percent legume contents have occurred.

Figure 141. - A good Puero (P. phaseoloides) guinea grass (P. maximum) mixture.

Among factors which are likely to affect pasture legume content are initial management during the first 6 to 12 months, together with intensity and frequency of grazing. The effect of cutting interval on dry matter yield of Siratro sown with grasses is shown in Table 80.

Furthermore the application of large quantities of nitrogen fertilizer can have negative effects on legume percentage (see 5.3.9) while the mainly positive influence of phosphate in the establishment of pastures with a high legume content has been noted (see Chapter 4, section 4.11).

The grass-legume balance is likely to fluctuate and will not remain static. An example on a countrywide scale is that in recent years there has been evidence of a decline in legume content in Australian pastures. Shelton (1990) has reviewed some of the factors involved which include:

1. changing economic incentives which have resulted in poorer pasture management and reduced legume persistence;

2. an inability of many of the twinning and more erect tropical legume cultivars released in the 1960s and 1970s to persist under moderate to heavy grazing;

3. serious challenges from a number of new and quite devastating insect pests and diseases;

4. environmental aspects, particularly acidification and changes in long-term rainfall patterns.

Table 80. - The effect of cutting interval on DM yield of M. atropurpureum cv. Siratro sown with grasses (Jones, 1967 after Whiteman, 1980)

Figure 142. - Development stages of grass plant.

5.3.3 Effect of growth stage on pasture quality and implications for grazing policy

The general relationship between stage of growth and pasture quality has been demonstrated in a number of studies (Milford and Haydock, 1965; Reid et al., 1973). In vitro dry matter digestibility coefficients for various tropical grasses and legumes harvested at different stages of maturity as well as the relationships between dry matter digestibility and number of days of growth were given by Reid et al., (1973). As plants mature, digestibility and feed quality decrease but the dry matter quantity increases. The general relationship between stage of growth and feed value is shown in Figure 142. Young leafy plants are high in protein, the cell walls containing readily digestible cellulose. As the plant matures, protein levels and digestibility decline with grasses producing stems and seed heads. This increasing proportion of stem in pasture reduces the overall nutritive value of the available dry matter (Whiteman, 1980). Grasses tend to decline more rapidly in protein content with increasing maturity than legumes (see Figure 143), some grasses have different rates of decline in crude protein levels (see Table 81).

Table 81. - Effect of length of regrowth period on DM and crude protein yields of 3 grasses¹

¹ Source: Reynolds and M. Faamoe, unpublished data.

Figure 143. - Relationship between crude protein level and plant age (Milford and Haydock, 1965; Whiteman, 1980).

The crude protein percentage of Batiki (I. aristatum) drops sharply between 2 and 4 weeks while the decline is less marked with Pangola (D. decumbens) and Guinea (P. maximum). Batiki is a grass which should have a relatively short grazing cycle and be grazed hard and frequently. In Malaysia, Romziah et al. (1986) demonstrated dry matter yield increases, and crude protein and dry matter digestibility decreases overtime with cutting interval from 3 to 12 weeks for Setaria splendida.

The critical level of crude protein in tropical grasses is about 7 percent (1.2% nitrogen, Minson, 1971), below which dry matter intake is depressed (Milford and Minson, 1965). The legume in a pasture has an important role to play in maintaining adequate protein levels in the feed on offer (Whiteman, 1980). To maintain adult cattle (in temperate conditions) at a constant weight without obtaining any production, herbage should contain about 4.0–4.2 percent digestible crude protein (DCP), (Crampton and Harris, 1969). Milford and Haydock (1965) suggest that the critical values for sub-tropical pastures are 7.2 percent crude protein and about 3.1 percent digestible crude protein. Evans et al. (1992) suggest a critical crude protein level of 9.0 percent and indicate that with a low legume content as much as 40 percent of the pasture area in Vanuatu does not have this required level of protein in forage being consumed. Good nutrition is important not only in terms of liveweight gain but also in terms of reproductive efficiency, calving percentage and other factors. Thus phosphorous deficiency in the diet may be expressed in terms of low calving percentages because of inhibited oestrus.

To achieve high production from tropical pastures animals must consume high quality feed. Thus grazing management aims to maintain pastures in a young leafy stage for as long as possible. Usually a compromise has to be made between quality (high protein content of young regrowth) and quantity (high DM content of mature pastures). Overgrazing can result in reductions in growth rate and weakening of plants. Undergrazing fails to maximize the plant's potential to produce feed resulting in a lower quality being available to the animal. According to Edgley and Quinlan (1975) the faster a pasture is defoliated to the desired level and then destocked, the greater is the opportunity for regrowth. The optimum time for grazing is shown in Figure 144. (The choice of grazing system is considered in more detail in Section 5.3.5.2. In fact, Edgley and Quinlan's ideas should be compared with the research findings of Ottosen et al., (1975) from the Atherton Tableland). This will obviously vary with species and time of year, but is likely to be within 2–6 weeks of destocking, i.e. the optimum length of regrowth period after rapid and even removal of forage by grazing cattle is from 2 to 6 weeks.

5.3.4 Stocking rate

This is usually defined as the number of grazing animals per unit (ha) of land at a particular time, although on extensive systems it becomes: ha animal^-1. It is one of the most important variables influencing both the productivity per animal and per ha (Jones, 1975). Evans (1993) indicates that the greatest influence on animal performance and sustainable pasture production is the choice of stocking rate. The long-term stocking rate is referred to as the carrying capacity implying an optimum stocking rate which can be maintained without damage to the pasture. Stocking rate involves only animal numbers and land area, whereas the most important aspect is the amount of forage produced (Whiteman, 1980). Thus grazing pressure refers to the number of animals per unit of available herbage, e.g. kg of dry matter on offer per animal unit (Skerman, 1977).

The stocking rate or carrying capacity of pastures under coconuts may vary according to differences in:

1. planting density, age of palms and therefore degree of shade;

2. botanical composition of the pasture;

3. climatic factors and time of year (stocking rate may be lower in very dry periods);

4. soil fertility and amount of fertilizer used;

5. type and age of animals;

6. grazing system and pasture condition;

7. availability of supplementary feeds;

8. management level (Plucknett, 1979).

Figure 144. - Effect of time on pasture growth rate and quality (Edgley and Quinlan, 1975).

The relationship between age and density of coconut palms and cattle carrying capacity on natural pastures in the Solomon Islands is shown in Table 82. Humphreys (1991) suggests that recommended stocking rates in the Solomon Islands for differing levels of light transmission might be: 35%, 0.7 beasts ha^-1; 45%, 1.0 beasts ha^-1; 50%, 1.3 beasts ha^-1; 60%, 1.6 beasts ha^-1; and 80%, 2.5 beasts ha^-1. Work in Vanuatu by the Vanuatu Pasture Improvement Project (MacFarlane et al., 1994) suggests that somewhat lower recommended stocking rates are appropriate for cattle under coconuts at different light transmission values (see Figure 145). Stocking rates were almost linearly related to light transmission (MacFarlane, 1993). An indication of the carrying capacity of natural pastures under coconuts in different parts of the world is given in Table 83. Rates vary from 0.3 to 2.5 animals ha^-1. A summary of available data on stocking rates on different improved pastures is given in Table 84.

Figure 145. - Recommended stocking rates for cattle under coconuts in Vanuatu (MacFarlane, 1994).

Table 82. - Influence of age and density of coconut palms on carrying capacity of natural pastures in the Solomon Islands (Walton, 1972)

Table 83. - Carrying capacity of natural pastures under coconuts (modified from Plucknett, 1979)

A comparison of estimated stocking rates on various grass species with suggested stocking rates for different pastures in Western Samoa are given in Tables 85 and 86. Robinson (1981) used up to 4 steers ha^-1 in later grazing trials in Western Samoa, while in the Solomons stocking rates of 1.5, 2.5 and 3.5 animals ha^-1 were used (Watson and Whiteman, 1981a and b). For Vanuatu the effect on stocking rate of pasture type and degree of shade are shown in Figure 146. On Bali, Rika et al., (1981) grazed a complex mixture of pasture species under coconut palms at 2.7, 3.6, 4.8 and 6.3 Bali cattle ha^-1. Clearly it is difficult to compare stocking rates from area to area or even from trial to trial when the terms used are too general such as ‘steers’, ‘number of head’, ‘yearlings’, ‘weaners’ and ‘animals’ with no indication of size. It is preferable to think in terms of number of ‘animal units’ or ‘livestock units’ ha^-1 (where 1.AU = 400 kg steer, although more recent convention is that one animal unit is a beast weighing 450 kg liveweight - Evans, 1993). The weaner steers used in the trials in the Solomons had mean liveweights of 132 kg (1980–81), 140 kg (1981–82) and 145 kg (1982–83) according to Smith and Whiteman (1983b), while male yearling Bali cattle (Bos banteng) were purchased in two groups with mean liveweights of 97 kg and 108 kg per animal. It was concluded in Bali, that stocking pastures under coconuts (60 years old, 10 m × 10 m spacing, 79 percent noon light transmission) at the rate of 5 yearlings ha^-1 with an annual change of animals (800 kg biomass ha^-1 (or 2 AU) average throughout) provided good long term production of about 550 kg ha^-1 year^-1 with no marked deterioration in species composition and pasture quality. In the Solomons, to obtain maximum liveweight gains ha^-1 pastures under coconuts (with a mean light transmission of about 60 percent) should be stocked at about 3.6 animals ha^-1 or 2.0 AU ha^-1 (Smith and Whiteman, 1983b).

According to Jayawardana (1985) natural pastures under coconuts in Sri Lanka can maintain only one animal per 2–3 ha whereas improved pasture under coconuts can support one animal per 0.5–0.8 ha. Guzman and Allo (1985) concluded that improved pasture under coconuts in Philippines could easily carry at least 1 AU ha^-1 (cattle). Ivory et al., 1984 reported a stocking rate of 0.5–1 animal ha^-1 for cattle grazing on native pasture under coconuts or other plantations such as rubber and clove in Indonesia, with average daily gains of 0.25–0.30 kg equivalent to animal liveweight gains of about 45–108 kg ha^-1 year^-1. According to Rika (1991), depending on season, the native forage (grasses and legumes) under plantation crops (coconut, clove and rubber) can support 0.7–1.4 AU ha^-1 under the cut-and-carry system. Grazing of cattle on introduced/improved pasture at a stocking rate of 6.6 head ha^-1 (average initial weight 90 kg) over a period of 252 days, resulted in an average daily weight gain of 378.9 g. Cattle grazing on native pasture under coconut achieved 311.4 g day^-1. Coconut production increased 41.4% on the improved/ introduced pasture area (Winaya et al., 1983). In Vanuatu, Evans et al. reported good liveweight gains on buffalo/carpet grass and mimosa pastures under coconuts at 2.6–3.0 animals ha^-1, whereas Berges et al. (1993) suggest that the average actual stocking rate is around 1.25–1.5 animals ha^-1.

Table 84. - Stocking rates on improved pastures

Figure 146. - Recommended stocking rates for cows and calves for open or shaded pasture (MacFarlane et al., 1992).

Table 85. - Suggested stocking rates for beef steers on pastures under coconuts in Western Samoa (Reynolds, 1980)

Note Steers in 225–360 kg liveweight range gaining 0.25–0.5 kg head^-1 day^-1, coconuts spaced 9.1 m × 9.1 m.

¹ B. brizantha, B. humidicola, B. decumbens, B. miliiformis, P. maximum cv. Embu plus associated legumes C. pubescens, M. pudica and D. heterophyllum.

² P. maximum, common and tall varieties plus associated legumes as in 1.

Table 86. - Grouping of grass species on basis of yield, estimated digestible protein production, stocking rate and number of animal grazing days (Reynolds, 1978l)

Parawan (1991) and Parawan and Ovalo (1987) noted that Buffalo (Carabao) have been tried on native pastures under coconuts in the Philippines at 1–2 AU ha^-1. Moog et al. (1988) showed that carabaos grazing on native-centro pasture under coconuts stocked at 2.0 AU ha^-1 had a cumulative liveweight gain of 190 kg ha^-1 (on a 15 month grazing trial conducted in Sorsogon, a province at the southern tip of Luzon). Carabaos on native-centro pasture stocked at 1.0 AU ha^-1 gained 104.7 kg ha^-1, while the animals on native pasture at 1.0 AU ha^-1 gained only 47.0 kg ha^-1. These findings indicate the potential of introducing legumes into native pastures under coconuts (Arganosa, 1991) although the botanical condition of the pasture at the end of the trial suggests the gains were not sustainable. At the beginning of the grazing trials the botanical composition of native-centro pastures consisted of 41 percent Centrosema and 20 percent pasture weeds. At the end of the grazing trial, the Centrosema component of pasture stocked at 2.0 AU ha^-1 was reduced to 8.6 percent (weed component increased to 65 percent) while on the pasture stocked at 1.0 AU ha^-1 the Centrosema component was reduced to 16.9 percent (and the weed component increased to 66 percent).

Average carrying capacity of the same native pastures for goats and sheep is 7– 15 head ha^-1, with the maximum carrying capacity on improved pastures rising to 20– 28 head ha^-1. Satutan et al. (1986) observed that the optimum stocking rate for cattle on an unfertilized Para/Centro pasture under coconuts was 1.0 AU ha^-1. In Sri Lanka, Perera (1972) indicated that B. miliiformis pastures under coconuts could carry 4 ewes acre^-1.

There are no general recommendations for optimum carrying capacity because of the complex of factors which governs stocking rate; however, the various data given in this section will help determine the initial choice of stocking rate which can subsequently be adjusted on the basis of local experience and which also may require seasonal adjustment depending on the rate of pasture production (Teitzel, 1992).

Although optimum carrying capacity is often measured in terms of maximum sustainable liveweight gain ha^-1, for farmers selling animals to the abbatoir or local butcher who require a quick turn off it is gain head^-1 that is the most important criterion (Evans, 1993).

This subject of meat quality is one which will have to be addressed increasingly by cattle producers. MacFarlane (1993a) notes that with the more stringent specifications of the market (see Figure 147) which requires younger, heavier carcasses (see Figure 148) than in the past it may be that rather than maximizing liveweight gains ha^-1, stocking rates may have to be adjusted to emphasize production animal^-1. Maximum profitability may be more closely related to achieving a certain critical minimum animal growth year^-1 (which in Vanuatu MacFarlane (1993a) suggests should be 180 kg an^-1 yr^-1 for improved systems and 140 kg an^-1 yr^-1 for native and buffalo grass systems) in order to turn off stock earlier rather than simply maximizing production ha^-1.

5.3.5 Stocking system

5.3.5.1 The systems

There are two systems of grazing management, continuous and rotational:

Continuous - animals are grazed within one enclosed area over a long period of time, which may last for a grazing season, or an entire year.

Rotational - the available pasture is subdivided into a number of enclosures or paddocks; animals are grazed in a regular sequence before returning to graze again the first paddock after a regrowth period of about 2–6 weeks (see Figure 149).

Figure 147. - Export beef ready to leave Santo Meat Packers, Vanuatu for the Japanese market.

Figure 148. - Beef carcasses at Santo Meat Packers, Vanuatu.

Figure 149. - Hereford steers moving to a new paddock in the rotational grazing of a Guinea-Centro pasture.

In both systems the number of animals can be kept constant (set stocking) or varied (variably stocked or put-and-take grazing) according to pasture growth and number of animals available (Booysen, 1975). (The system of zero grazing is considered below).