Sustainable grazing systems for farms in temperate Australia

Abstract

Animal Production in Australia 1998 Vol. 22 CONTRACT REVIEW SUSTAINABLE GRAZING SYSTEMS FOR FARMS IN TEMPERATE AUSTRALIA J.R. DONNELLY CSIRO Plant Industry, GPO Box 1600, Canberra, ACT 2601 INTRODUCTION If grazing systems are to be sustainable they must be profitable. It is unrealistic to consider sustainability solely in terms of resource conservation without identifying the key profit drivers and adopting a sensible approach to risk management. A grazing enterprise may also be only one of several enterprises on a farm and its integration into a whole-farm management plan is an essential step in achieving long-term sustainability. Likewise, each farm is only one of the land-use components in a catchment or district. Apart from economic realities, sustainable grazing systems must be built on sound biophysical and ecological principles and they must be socially acceptable. In this contract we discuss first well-established ways to increase profitability without substantial increases in costs and illustrate how modern decision support (DS) tools can help to provide better assessment of risks and opportunities. This is followed by a discussion of some of the main ecological and biophysical processes driving farm-level grazing systems in temperate Australia where we believe more research is required. These include issues of species persistence and compositional changes in pastures, nutrient cycling and maintenance of stable water tables as they interact with grazing systems. The grazing enterprise as a sustainable business Resource management for long-term production in grazing enterprises is a difficult and sometimes risky business. Graziers may perceive that they are not in a financial position to afford management practices compatible with wise land use and resource conservation. Rural lands have been damaged by years of inappropriate management and the scars are increasingly visible. Poorly productive pastures are widespread, with legumes and perennial grasses, both sown and native, being replaced by weeds or less productive annual grasses. Declining soil fertility and soil structure and damage from soil acidification and salinisation may be irreversible or extremely expensive to remedy. More than 30 million ha of farmland in the better-watered regions of Australia are now highly acidic and options for growing some valuable crop and pasture species are limited. A further 55 million ha are at risk due to increasing soil acidification. Current loss of national productivity due to soil acidity is estimated at $300 million and to salinity at $200 million. Overall, annual losses from all sources of land degradation are estimated to exceed $1 billion (Chartres and Isbell 1995). There is also the loss of biodiversity in our unique native flora and fauna. Other costs that reduce financial sustainability are associated with inappropriate management such as sub-optimal stocking rates coupled with unecessary supplementary feeding of livestock. The magnitude of these losses, and the potential size of the impending environmental disaster if no action is taken, have stimulated several thousand Landcare and similar groups to initiate remedial action to rehabilitate degraded land and soils. Less success has been achieved in applying existing knowledge from research which can dramatically increase profits and place graziers in a sound position to invest in better land care and complement Landcare activities. For example, even with current depressed wool prices, profits on a typical 550 ha sheep farm in southern temperate Australia can be increased from a current average loss of $22,000/ year to a profit of $68,600/year (Lean 1997). It is urgent that graziers become aware of these significant possibilities for the sustainability of their businesses. One approach which will help is to provide graziers with a comprehensive assessment of the relative importance of financial, biological and environmental risks that affect production in their own enterprises. Profit drivers for sustainable grazing systems Current farming practices clearly need adjustment if Australia is to have an agricultural sector that is capable of meeting the competitive markets of the 21st century. The processes that lead to land degradation need to be understood so that effective remedial action can be taken, but landholders must be in a sound financial position with survival no longer their immediate concern. Sustainability is as much about profitability and cash flow as it is about resource conservation. The landholder therefore needs to identify the key profit 61 Animal Production in Australia 1998 Vol. 22 drivers, which for most temperate grazing enterprises are well researched. Joining at a time that achieves the best balance between the seasonal pattern of pasture supply, animal needs and market prices, and minimising supplementary feeding, cost almost nothing to implement and can result in immediate savings. However, it is through increasing stocking rates on productive pastures that the biggest financial benefits are most likely. Extensive research over several decades shows that stocking rate is the key profit driver and unless it is close to optimum, other management tactics may have little impact on profits. Stocking rate is also a key driver of ecological sustainability in a grazing system. If stocking rates are too high then exposure to environmental damage as well as financial losses becomes inevitable. The issue then is: what is the optimum sustainable stocking rate for any grazing situation? Estimating sustainable stocking rates Serious depletion of ground cover for prolonged periods is a sure indicator that stocking rates are too high. However, critical levels of ground cover will vary widely with soil, pasture type and fertiliser policy and many native perennial grasses are susceptible to intense and prolonged defoliation. Simulation provides a powerful way to explore likely responses of a grazing system to a range of management options such as stocking rate, although there are limited experimental data to test the outcomes. The CSIRO GrassGro DS tool (Moore et al. 1997) is ideal for this purpose as it simulates pasture and animal production on local soil types using relevant weather records as the driving variables. It thus links sustainability and profit for a specific grazing enterprise to the local environment. Interactions between management variables can be explored efficiently and quickly, but user training is essential for effective use. FINANCIAL STABILITY OF GRAZING SYSTEMS J.R. DONNELLYA, D. ALCOCKA, M. FREERA and S. CLARK A B B CSIRO Plant Industry, GPO Box 1600, Canberra, A CT 2601 Agriculture Victoria, Pastoral & Veterinary Institute, Private Bag 105, Hamilton Vic. 3300 Stocking rate and financial risk The need to generate profits without exposure to unacceptable levels of financial risk is fundamental to the sustainability of grazing systems. Stocking rate is a key profit driver and gross margins (GM) generally increase with stocking rate up to a point beyond which further stocking rate increases have only a marginal impact on profits. More extreme increases in stocking rate will lead to a rapid decline in GM and can result in environmental degradation. In selecting a sustainable optimum, however, variability in annual GM is just as important as the actual level of the GM itself. The examples outlined below are derived from simulations with GrassGro of three Merino wether wool production enterprises in south-eastern Australia, located on sites with low, moderate or high soil fertility. The pastures were based on annual grass and were continuously grazed. The period simulated was for 27 years from 1970. Figure 1a illustrates the relationship between financial risk and GM as stocking rates increase at three levels of soil fertility in otherwise identical conditions. GM increases with stocking rate but the variability in GM (measured here as the standard deviation (sd) of the GM) or financial risk increases more rapidly when fertility is lower. This is shown more clearly in Figure 1b at the site of low fertility, where the simulation has been extend to 100 years from 1897. The annual estimates of the simulated GM at each stocking rate (10, 15, 20 and 25 wethers/ha), presented as boxplots, illustrate the extent of annual variability. The bar within the boxes represents the median GM. The bottom of the box is at the 25th percentile and the top at the 75th percentile. The lines extending from the box include the remaining data but with outliers indicated as dots. The length of each box and associated lines and dots directly represent the financial risk associated with each stocking rate. It is clear that risk is slight at 10/ha but the median GM is also small. GM increases by 70% at 15/ha for a small increase in risk, but risk increases substantially at 20/ha for no further increase in GM. At 25/ha GM has fallen and the chance of making a negative GM is greater than one year in four (25% of the GM values are below the bottom of the rectangle ie, a GM of less than -$40). 62 Animal Production in Australia 1998 Vol. 22 The simulations show that the main reason for the down-side variation in annual GM at each stocking rate was the need to feed relatively expensive supplements (oats at $150/t) when forage was scarce, particularly during major droughts, but also during shorter dry spells which occurred more frequently. Even so, the rule for feeding supplements was aimed merely at keeping the animals above condition score one to avoid deaths. No other feeding was specified. The up-side variation is smaller and was due to year-to-year fluctuations in production per head and product quality and insufficient animals to use extra feed grown in the better years. At low stocking rates, the simulated forage available from pastures was usually well in excess of animal needs, except in severe drought years. The need for supplements was therefore minimal and the range in GM was small (Figure 1b). At very high stocking rates, production per head was generally lower, and in most years, forage supply from pasture was inadequate; substantial and costly supplementary feeding was necessary to keep animals alive. The increase in up-side variation at 25 sheep/ha was due to livestock numbers sufficient to utilise all the feed grown in very good years. Two other features in Figure 1 deserve comment in relation to sustainability. First is the importance of recognising that an over-cautious attitude to risk will lead to greatly diminished GM. The ability to invest in improvements that may lead to better sustainability is therefore reduced. The second issue is the financial response at sites with different soil fertility. It is interesting to note that current research indicates the possibility that very much higher herbage production may be practicable through heavy application of P fertiliser (Cayley and Hannah 1995). This would result in a significant shift in the curvilinear relationship towards the much higher stocking rates shown in Figure 1a. Figure 1. (a) Gross margin (GM) for wool production v financial risk or variability in annual gross margin (s.d. GM) with stocking rate (sheep per ha) for low, medium and high fer tility. (b) The incr ease in financial risk as stocking rate increases at the site with low fer tility. The curve in (b) is plotted through the median GM for each Assessing the probability of the need for supplements in drought or dryspells When a dry spell starts, farmers are faced with tactical decisions such as to feed, to sell, or to agist. Which decision they take will depend on their assessment of the probable length of the dry spell. As we have discussed, feed costs are a significant determinant of the variability in GM and hence the financial sustainability 63 Animal Production in Australia 1998 Vol. 22 of the system. GrassGro can be used as a tactical tool to assess the probability of herbage supplies exceeding specified amounts over a short period ahead of the current date (up to one year). Two case studies presented below outline how this is done. The first study was carried out at Yass in August 1997 after a very dry winter. To prevent losses associated with nutritional problems during pregnancy, Merino ewes had been fed a supplement to keep body condition score above two. On 10 August the ewes were due to start lambing but the available green pasture was only about 500 kg DM/ha. The management issue on the 10 August, was therefore a tactical one about stocking and feeding decisions that will influence the flock situation at the start of next growing season, ie winter 1998. GrassGro was used to simulate pasture and animal production, starting with the situation of the pastures and flock on 10 August 1997, forward to 31 May 1998 using the daily weather records for the period 10 August - 31 May for each year from 1970 to 1995. These simulations also provide estimates of feed costs from which a cumulative probability distribution was calculated (Figure 2). Figure 2 shows the distribution for the 25 years of simulated feed costs over the nine month period from 10 August for sites with low and high soil fertility. Any figure read from the Y-axis is the probability that the total feed cost over this period will be at least the corresponding amount on the X-axis. For example, there is a 50% probability that the feed cost will be at least $20/ha at the high fertility site or $60/ha at the low fertility site. The spread in feed costs shown in Figure 2 reflects the range in costs encountered from wetter to drier conditions during the period August to May over the 25 years. Over the whole range there is a substantial difference in the probability distributions between the 2 sites. The reduction in feed costs associated with regular fertiliser use is at least as great in drier years as in wetter years and will lead to a reduction in financial risk. This was of course, one of the main factors contributing to the risk differences shown in Figure 1a. While no analyses can predict the future for the grazier, this analysis does provide a quantitative estimate for a particular farm of the feed costs likely to be incurred for any specified level of risk. Figure 2. The probability that the feed cost will exceed specified levels The second case study is based on analysis of seasonal grazing outlook at the Pastoral and Veterinary Institute, Hamilton, Victoria, following the dry growing season in 1997 when only 67% of average rainfall fell in the March-October period. At the end of October, measured total available herbage was 2.3 t DM/ha on perennial ryegrass pastures stocked at 15/ha with Merino ewes and their lambs. The expected daily availability of herbage DM for the period November to April, based on simulations of these pastures over the past 100 years, is shown in Figure 3a. In 50% of years, peak available herbage DM exceeds 5 t/ha and exceeds 7 t/ha in 10% of years. However, in 10% of years peak available herbage does not exceed 4 t/ha. Figure 3b shows the expected availability of herbage between November and April given the very low DM (2.3 t/ha) and soil moisture present at the end of October 1997. These tactical simulations suggest that feed availability in the summer-autumn period is likely to be sparse. This means that there is an increased 64 Animal Production in Australia 1998 Vol. 22 Figure 3. Percentiles for herbage availability (t DM/ha) over six months fr om November to April from a) historical simulations and b) tactical simulations where the situation prevailing on 1 November 1997 was taken as the starting condition. In the historical simulations the period simulated was continuous from 1879 until 1997. In the tactical simulations historical weather data for the period November to March each year from 1879 until 1996 were used. likelihood that some supplements will need to be fed in the late summer or autumn. The tactical simulations suggest that at high stocking rates and given the conditions prevailing at the end of October 1997, there is about a 2 in 3 chance that at least 8 kg of supplement will be required per ewe to maintain body condition score at 2. This compares with at least 26 kg of supplement required over the same period in the worst 4% of years since 1879. The analysis here suggests that even with the dry winter and the expectation of less than normal forage over summer and autumn, maintaining a stocking rate of 15 ewes/ha is likely to be financially sustainable. These case studies are just a few examples of the power of current DS tools to couple the mass of accumulated weather records to our scientific knowledge of grazing systems, in the interests of long-term sustainability. ECOLOGICAL SUSTAINABILITY - SPECIES PERSISTENCE IN GRAZING SYSTEMS T.P. BOLGERA and D.L. GARDEN A B B CSIRO Plant Industry, GPO Box 1600, Canberra, A CT 2601 NSW Agriculture, GPO Box 1600, Canberra, ACT 2601 General concepts of stability and change in pastures One of the cornerstones of sustainable grazing systems is a relatively stable composition of productive and nutritious species in the pasture plant community. Why do some pastures remain stable and productive while others are unstable and become degraded? What steps must be taken to improve the productivity, sustainability and stability of these pastures? A framework of ecological concepts relating to stability and 65 Animal Production in Australia 1998 Vol. 22 change in pasture species composition is needed to develop our understanding of botanical change, so that we can better manage our pasture resources. Stability can be defined as the degree to which a pasture community resists change when disturbed by management actions or natural phenomena such as climate. The stability of a pasture therefore is related to its ability to return to its original state after being disturbed. Traditional models of succession in plant communities suppose a single, stable state for the pasture in the absence of disturbance, termed the climax, which it steadily moves towards after any disturbance. Contemporary ecological models of vegetation change, such as the State and Transition model (Westoby et al. 1989), suggest a more dynamic situation where change does not necessarily move in a single direction. In this model, if a disturbance exceeds the limits of the stability of the pasture, then the pasture does not return to its original state after the disturbance, but rather crosses a threshold and reaches some new state. Transitions between states are triggered by natural events or management actions or a combination of the two. Plants, animals, soil, climate and management all interact to determine the state of a pasture, and how stable that state will be. For example, perennial grasslands are generally considered to have good stability. However, in environments with strongly seasonal rainfall, perennial grasslands have been converted to annual graslands by grazing (Heady 1958). When grazing is stopped, they often do not revert to perennial grass dominance (Hutchinson 1992). Unfortunately, it is easier to move from a stable, productive pasture to a degraded pasture than in the opposite direction. Some extreme events such as drought, especially when combined with heavy grazing, can result in rapid deterioration of the composition of a pasture (Hutchinson 1992). In contrast, improving the composition of a pasture with grazing management, where possible, is generally only successful if continued for a relatively long period of time. To apply these concepts to maintaining or developing stable, sustainable pastures, we must understand how a stable community is forced across a threshold into a less stable state, and how this change can be reversed. Why pasture systems are sometimes not sustainable Pasture improvement in Australia has generally been based on replacement philosophy (Whalley and Lodge 1987) and the assumption that sown pastures, once established, are effectively permanent. However, there are several flaws in this philosophy, which are highlighted by surveys, farmer observations and limited research. For example, Garden et al. (1993) found that many pastures previously sown to introduced species were dominated by either annual grasses or native perennial grasses. The reasons for this decline in introduced species are likely to be a combination of unsuitable soils, poor establishment, insufficient fertiliser, selective grazing, drought and poor management (Hutchinson 1992). Simpson and Langford (1996) have highlighted the problem of attempting to establish introduced species on soils which are inherently unsuitable, mainly because of shallow soil, acidity and low fertility. They recommend a more discriminatory approach to property development, with careful consideration of site and livestock requirements in selecting appropriate species. Alternatively, areas which are considered inappropriate for sown species can be managed on a long-term basis using lower inputs of fertiliser and legume seed. Pastures in these areas may contain a mixture of native perennial grasses, volunteer annual grasses and remnants of previously sown grasses and legumes. Since 8-10 years are required to recover establishment costs of sown pastures, pastures cannot be economically sustainable unless they survive beyond the payback period. The wisdom of applying the same technology across the whole of a property therefore needs to be questioned. There is also a need to balance productivity and the maintenance of ground cover to prevent erosion and other forms of land degradation. In many situations, this may mean utilising existing pastures containing native species rather than sowing poorly adapted introduced species. Fertiliser is usually applied when pastures are sown, and follow-up fertiliser applications are recommended. However, the low profitability of many grazing enterprises means that some pastures must survive without regular fertiliser inputs. Such pastures may continue to grow for some years by mining nutrients from previous applications, but the point is eventually reached where many sown species are unable to compete with species which are able to survive at lower fertility levels. If sown perennial grasses decline, their place is often taken by less productive annual grasses, creating a pasture composition which is inherently less stable and difficult to return to its previous state. In some cases, previously-sown pastures may be reinvaded by well-adapted native species which form a pasture which is stable and productive under a wide range of conditions (Garden et al. 1993). An example of the interacting effects of drought and management on pasture decline, even under high 66 Animal Production in Australia 1998 Vol. 22 fertility conditions, has been described by Hutchinson (1992). Over a period of 28 years (including the droughts of 1965 and 1980-82), a phalaris/clover pasture on the northern tablelands of NSW was relatively stable at a low stocking rate, but at stocking rates above 15 sheep/ha, the proportion of phalaris relative to annual grasses declined. Both theoretical and experimental evidence suggest that the closer a pasture is pushed towards maximum productivity the closer it is to the limits of its stability, and therefore more vulnerable to environmental fluctuation pushing it across a threshold to a degraded state (Noy-Meir 1975). A somewhat lower stocking rate will ensure a more stable pasture with less critical management requirements for sustainability, but with somewhat lower production, at least in the short term. This highlights the need to reconcile ecological stability and vulnerability with long- and short-term profitability and risk. Sustainable grazing systems can only be achieved by careful consideration of animal requirements, land capability, species choice, fertiliser inputs and appropriate management, including drought strategies. For ecological sustainability, the goal of management is to seize opportunities for favourable transitions and evade hazards which threaten unfavourable transitions (Westoby et al. 1989). Only then can a land manager ensure that pasture composition will not move from a stable to an unstable state, where production may be significantly lower, and from where it may be difficult to return. WATER LOSS FROM GRAZING SYSTEMS S.G. CLARKA, W.J. BONDB, H.P. CRESSWELLB, R.J. SIMPSONC, Z. PAYDARB, A.D. MOOREC, D.J. ALCOCKC, J.R. DONNELLYC, M. FREERC, B.A. KEATINGD, N.I. HUTHD, V.O. SNOWB, and T.P. BOLGERC Agriculture Victoria, Pastoral & Veterinary Institute, Private Bag 105, Hamilton, Vic 3300 CSIRO Land and Water, GPO Box 1666, Canberra, ACT 2601 C CSIRO Plant Industry, GPO Box 1600, Canberra, A CT 2601 D QDPI/CSIRO Agricultural Production Systems Unit, PO Box 102, Toowoomba, Qld 4350 A B Water balance of land under pasture Winter rainfall is in excess of evapotranspiration (ET) over much of the temperate grazing lands of southern Australia where surplus water must be either stored in the soil or contribute to run-off. Movement of water beyond the root zone of pastures may leach nutrients and cause soil acidification and ultimately pollution of waterways. It may also cause water tables to rise, bringing salt into the upper soil profile. Under the original native trees and deep-rooted shrubs, most of the annual rainfall was used by the plants so this deep drainage occurred infrequently, usually only in very wet years. With clearing and pasture improvement, replacement of the natural vegetation with shallow-rooted annual grasses and legumes has removed a safety buffer which protected the soil and groundwater environment. Managing annual pastures for high productivity results in only a marginal increase in total water use, the main effect being increased plant transpiration of water at the expense of soil evaporation (Bolger et al. 1993). There is now growing interest in exploiting the potential that deep-rooted perennial pasture species offer to redress the serious environmental damage that is occurring. If the perennial species can use water at depth, a solution may be available to control water balances which would help to ensure profitable production and to secure more sustainable grazing systems in the longer term. However, the impact on profitability and the role of grazing management to achieve the potential benefits need to be established. The simulation study described below (Bond et al. 1997) was undertaken for the MRC Sustainable Grazing Systems Key Program (Mason and Andrew 1998) to do this and to predict likely water-balance responses to grazing management at field sites at Hamilton, Albany and Tamworth. Although the results are expected to be indicative, this study was exploratory in nature and undertaken with minimal input data and with no opportunity for validation. The simulations were made using GrassGro (Moore et al. 1997) to predict pasture responses to grazing management on specified soils based on long-term (25 years) meteorological records from each site. The water balances were calculated using SWIMv2 (Verburg et al. 1996), running in the APSIM (McCown et al. 1996) modelling shell, and with the pasture growth profiles predicted for each pasture species by GrassGro. The pasture, animal and water balance models used the same climate file to drive the simulations. Demonstrating our ability to link outputs from the two simulation systems was a requirement of MRC. Table 1 provides a brief summary of the results for the main treatments tested in the simulations. 67 Animal Production in Australia 1998 Vol. 22 Table 1. Simulated annual waterbalance (mm) for diff erent pasture types at three locations based on rainfall, evapotranspiration (ET), drainage (D) and runoff (RO) Pasture type Annual Perennial Native Woodlot Rain 713 Hamilton ET 440 528 630 D 195 109 13 RO 6 5 5 Rain 693 Albany ET 483 554 694 D 206 139 5 RO 0 0 0 Rain 708 Tamworth ET 674 659 699 D 3 25 0 RO 27 21 11 At Albany and Hamilton, predicted drainage of water below the root zone of pastures was substantial. However, the simulated perennial pastures reduced deep drainage by about 100 mm compared with drainage from annual pastures, a result similar to that from an experiment at a site with similar rainfall near Rutherglen, Victoria (Ridley et al. 1997). At both sites, trees in a woodlot were much more effective in reducing drainage to near zero. At Tamworth, the simulations suggested that deep drainage was less likely and in terms of water balance, pasture-based enterprises appeared to be sustainable. On average, the pastures at Tamworth were able to exploit all available soil water as there is a less than 50% probability of rainfall exceeding potential ET in any month. There may be less need for trees to limit deep drainage in this environment. In southern Australia, where much of the rainfall occurs in winter, perennial pasture species although better than annual species, still did not capture all the water draining from the system. The study indicated that other deep-rooted vegetation such as trees will be needed in these environments with wet winters. The major issue that needs resolution is how much and where this vegetation should be sited in the farm landscape and what impact it will have on the physical and financial sustainability of the grazing enterprise. This contrasts with areas in northern NSW with summer-dominant rainfall, where pastures based on introduced or native perennial grasses can maintain a better water balance because of the temporal distribution of rainfall in the region. Sustainable production and profit In the simulations at Hamilton, perennial pastures were more profitable than annual pastures, at least in the short-term as this was compatible with incremental gains in reducing deep drainage. However, an issue emerging from discussion of these results from Hamilton was the persistence of deep-rooted perennial species. At Albany, the perennial grasses kikuyu and tall fescue have contrasting growth patterns to the winter-active annual grasses and have potential to contribute to a more uniform supply of feed throughout the year. These species are not commonly used and management guidelines for their inclusion in pastures are required. Pasture growth at Tamworth was more variable between years than at the southern sites. Phalarisbased pastures might be slightly more profitable than pastures based on native perennial grasses, although the latter were predicted to be more responsive to sporadic rains. The simulation study showed that a simple assessment of average rainfall and potential ET coupled with an estimate of the capacity of the soil to store moisture would be a useful indicator of the hydrological sustainability of a grazing enterprise in a given environment. Pasture alone is unlikely to achieve a satisfactory water balance in wetter predominantly winter-rainfall areas. Although stocking rate was a key profit driver in all grazing systems studied, the simulations showed that it and other grazing management practices such as seasonal pasture spelling were unlikely to have an important impact on water balance, which accords with experimental results (T.P. Bolger unpublished data). These observations indicate the need to develop strategies for including other deep-rooted vegetation in our grazing lands on farms in southern Australia. 68 Animal Production in Australia 1998 Vol. 22 EFFICIENT NUTRIENT USE R.J. SIMPSON, H. DOVE and A.D. MOORE CSIRO Plant Industry, GPO Box 1600, Canberra, ACT 2601 Grazing enterprises