Changes in chemical properties of 48 intensively grazed, rain-fed dairy paddocks on sandy soils over 11 years of liming in south-western Australia.
Dairy production in south-western Australia occurs on predominantly sandy soils in areas with annual average rainfall >800 mm. Most dairy pastures are rain-fed in the Mediterranean-type climate of the region, with typically cool, wet May-October growing seasons and dry, warm to hot November-April. The pastures comprise the annual species subterranean clover (Trifolium subterraneum L.), hereafter called clover, and annual ryegrass (Lolium rigidum Gaud.) and Italian ryegrass (L. multiflorum Lam.).
Due to diminishing returns from milk production in southwestern Australia, dairy farmers have intensified their production to remain profitable. Paddock-grown pasture is the cheapest source of feed for milk production in the region, so farmers have improved grazing management to use as much paddock-grown pasture as possible for milk production (Bolland and Guthridge 2007a, 2007b, 2009). This was achieved by adopting the 3-leaf grazing strategy for ryegrass-based pastures, described by Fulkerson and Donaghy (2001). Pastures are rotationally grazed by starting grazing when ryegrass plants have 2-3 leaves per tiller, and fertiliser nitrogen (N) is usually applied after each grazing. About 80% of the pasture is grazed, but clover in the pasture can no longer provide sufficient N through symbiotic [N.sub.2] fixation for pasture production, hence the need to apply fertiliser N after each grazing (Bolland and Guthridge 2007b). Both clover and ryegrass respond to the applied N by growing bigger leaves and stems, but between grazings, ryegrass grows over the top of and shades out clover, so clover rapidly disappears and ryegrass comprises >90% of the pasture (Bolland and Guthridge 2007a, 2007b, 2009).
When newly cleared of native vegetation for agriculture, most soils used for dairy production in the region had pH values, as measured in calcium chloride, of 5.5-6.5 (McArthur 2004). Many of these soils have since acidified and pH values of 3.7-4.5 are now common. Aluminium toxicity, induced by soil acidification, is a major problem for crop production (Dolling et al. 1991; Dolling and Porter 1994) and dairy production (Bolland et al. 2002a) in south-western Australia, and is ameliorated by applying sufficient lime to raise the pH of the top 0.10m of soil to [greater than or equal to] 5.5 (Whitten et al. 2000).
When newly cleared, the soils in the region were acutely deficient in phosphorus (P) (McArthur 2004). In addition, the soils are predominantly sandy and the surface soils explored by roots of pasture plants mostly have low capacities to retain potassium (K) and to sorb sulfur (S). Both elements leach below the root-zone in wet years, inducing deficiency of the elements for pasture production, and requiring application of fertiliser to combat the deficiency (K, Bolland et al. 2002b; Bolland and Guthridge 2009; S, Bolland et al. 2003).
To ensure that P, K, and S deficiency does not reduce pasture production, most dairy farmers in the region usually apply fertiliser P, K, and S to pastures every year. The fertilisers are typically applied twice a year: in autumn, usually 3 weeks after pasture has emerged at the start of the growing season; and in late winter-early spring (August-September) (Arkell 1995). Many dairy farmers now apply fertiliser N, P, K, and S after each grazing to their intensively grazed ryegrass pastures.
The fertiliser P, K, and S requirements are known for clover but not for ryegrass (Bolland et al. 2002b, 2003; Bolland and Guthridge 2007a). Annual ryegrass has lower requirements than clover for P (Ozanne et al. 1969; Barrow 1975) and K (Bolland et al. 2002b; Bolland and Guthridge 2009). Consequently, the fertiliser P and K requirements of intensively grazed ryegrass pastures are likely to be less than those of clover in the traditional clover-ryegrass pastures not treated with fertiliser N after each grazing.
Fertiliser P has a good residual value for pasture production on sandy soils of the region, even in sandy soils with low capacities to sorb P (Barrow 1966). As a result of application of fertiliser P each year, the P status of most dairy soils is high and well above critical soil-test P of the various soils used. When fertiliser P is applied to these soils, no pasture dry matter responses to applied P are obtained (Bolland and Guthridge 2007a). On sandy soils in high-rainfall areas, continual application of fertiliser P each year, even when none is required, has increased wasteful loss of P from fertilised pastures to drainage, resulting in eutrophication of the many waterbodies in the region (Hodgkin and Hamilton 1993; McComb and Davis 1993).
The predominantly sandy soils used for dairy production in south-western Australia were initially infertile, and the physical and chemical properties of the soils have been improved by increasing organic matter contents of the topsoils (Barrow 1969). Mineralisation of sulfate-S from soil organic matter is a major source of S for high-rainfall pastures in south-western Australia (Barrow 1966). Cultivating soil increases mineralisation and depletion of the organic matter, so the pastures are usually permanent, and the soils are rarely cultivated or cropped. The amount of organic matter in the soils through time is routinely estimated by measuring organic carbon (C) content of the soils (Rayment and Higginson 1992). In addition, soil salinity occurs in patches in grazed pastures in the region, reducing pasture production (McArthur 2004). Soil salinity levels in soil are monitored by measuring electrical conductivity (EC) of the soil (Shaw 1999).
During 1999-2009, soil and tissue testing was undertaken in 48 paddocks used for intensively grazed, rain-fed dairy production on sandy soils on the Vasse Research Centre, near Busselton, south-western Australia. The aim was to measure changes in chemical properties of the soils, as determined by the soil-test procedures routinely used by commercial soil-testing laboratories in the region for pH, P, K, S, organic carbon, and EC. The results are reported in this paper. We hypothesised that soil testing for pH and P are reliable for indicating paddocks requiring lime and fertiliser P applications, and that no fertiliser P is required if soil-test P is above the critical value for that soil. Although soil testing is used in the region to provide fertiliser K and S advice, we propose that neither of these soil tests is reliable for high-rainfall pastures because both elements can be leached below the root-zone, and reliance should be placed on tissue testing instead. We suggest tissue testing is required in conjunction with soil testing to help monitor the success of decisions based on soil testing, and indicate deficiencies of elements not covered by soil testing and monitor success of fertiliser decisions made for the elements not covered by soil testing. We do not report the results of other soil properties also measured and reported by the laboratories, because the results are either unreliable for predicting deficiency for pasture production in the next growing season (total N, ammonium-N, nitrate-N) or not calibrated in the region because tissue testing is used rather than soil testing to determine fertiliser requirements (copper, zinc, manganese), or because the elements are not deficient for pasture production in the region so it has not been possible to develop local calibrations for the elements (calcium, magnesium, iron and boron).
Materials and methods
Location and climate
The 48 paddocks were at the Department of Agriculture and Food Vasse Research Centre (VRC, 33[degrees]45'S, 115[degrees]21'E, elevation 30m), ~15km south of Busselton, south-western Australia. Average annual rainfall at VRC is 876 mm, with ~85% falling in the May-October growing season and ~70% of the annual evaporation occurring during the hot, dry November-April period.
Soil classification and topography
Soils in the 48 paddocks were 1-2 m sand to sandy loam over massive clay, known locally as Abba sand (Tille and Lantzke 1990) and classified as Chromosolic Redoxic Hydrosol (Isbell 2002) and Glossic Natrudalf (Soil Survey Staff 1987). For many soils in the region, including Abba sands, the topography is flat and the soils are waterlogged from June to early September (Tille and Lantzke 1990; McArthur 2004). The capacity of Abba sands to sorb P and S, and to retain K, increases as the clay content of the soil increases (McArthur 2004).
Grazing studies conducted in the 48 paddocks
The 48 paddocks were used for two intensively grazed dairy studies during 2000-09. The first grazing study, called the Vasse Milk Farmlets (VMF), was conducted from mid 2000 to early 2005 and measured the effects of stocking rate and feeding different amounts of concentrate on pasture and milk production (Staines et al. 2007). The second grazing study, called Greener Pastures 1 (GP1), was conducted from mid 2005 to early 2009 and measured pasture, milk, and profit responses to five fertiliser N levels applied after each grazing (0.0, 0.5, 1.0, 1.5, and 2.0 kg N/ha.day). If it was 30 days since the last grazing, then 0, 15, 30, 45, and 60 kg N/ha was applied after that grazing, and a total of 0-331 kg N/ha was applied for all grazings in each growing season (Staines et al. 2008).
Collecting soil samples
Standard procedures for soil testing of high-rainfall pastures were used (Russell 2010). The top 0.10 m of soil was collected from each paddock in April 1999 and January-February 2000-09, during the dry period before fertiliser was applied in the next May-October growing season. The samples were collected by walking on the same diagonal path across each paddock each year between two permanent markers located on fences. Samples were collected using 25-mm-diameter metal tubes that were pushed into the soil by foot every 2-3 m, with 50-100 samples collected per paddock, depending on the size of the paddock. The samples from each paddock were bulked and air-dried, and subsamples of the <2 mm fraction were sent for analyses to the major commercial soil and tissue testing laboratory in the region, CSBP Laboratories at Bibra Lakes, Western Australia.
Soil properties measured and reported
Soil pH is used to determine the need for lime in south-western Australia and is measured using 1 : 5 soil : 0.01 M Ca[Cl.sub.2] (w/v), as outlined by Rayment and Higginson (1992, p. 19).
The Colwell (1963) sodium bicarbonate procedure, which is a modification of the original sodium bicarbonate procedure of Olsen et al. (1954), is the soil P test used in south-western Australia. The capacity of soil to sorb P has a large effect both on the amount of fertiliser P needed for near-maximum yields (Ozanne and Shaw 1967) and, for the Colwell procedure in particular, on the critical value required for soil-test P (Moody 2007). Consequently, both soil-test P and the capacity of soil to sorb P are routinely measured in south-western Australia to provide P fertiliser advice (Bowden et al. 1993; Bolland and Allen 2003). Three procedures have been used to rank the capacity of soil to sorb P in south-western Australia. Reactive iron (Fe) was the first procedure used from the mid 1970s until 2007 and is the amount of Fe extracted from soil by ammonium oxalate (Schwertmann 1964). Subsequently, two single-point P sorption indices were used to directly measure P sorption by soil: (i) from the mid 1980s to 2007, the P retention index (PRI; Allen and Jeffery 1990); (ii) since 2006, the P buffering index (PBI; Burkitt et al. 2002). All three methods were used and compared in this study.
Other soil-test procedures used in south-western Australia are as follows. The Colwell soil P test procedure is also used for soil K testing (Bolland et al. 2002b). The KCl-40 procedure (Blair et al. 1991) is used for soil S. Organic C content of soil is determined using procedures outlined by Walkley and Black (1934). EC is measured using procedures described by Rayment and Higginson (1992, pp. 15-16).
No major liming program had been undertaken in the 48 paddocks before April 1999. Soil testing in 1999 indicated that soil acidification was a major problem in all 48 paddocks, so a liming program was undertaken to rectify the problem.
Local ground limestone (from Redgate), with a neutralising value of 80-90%, was used to ameliorate soil acidification in the 48 paddocks. The lime was spread over the soil surface in March-May before the start of the growing season and was not incorporated into soil after application.
Price limited the amount of lime purchased and applied each year. No lime was applied in 2001, 2002, and 2005. The same level of lime was applied to all 48 paddocks in 1999 (3 t/ha), 2000 (4 t/ha), and 2003 (2 t/ha). In 2004, 17 of the 48 paddocks with the lowest soil pH were treated with 1 t/ha of lime. Lime was applied each year during 2006-09, the level applied varying with soil pH. In 2006, levels of lime applied were 5 t/ha if pH was <4.5, 2.5 t/ha if pH was 4.5-4.9, 1 t/ha if pH was 5.0-6.0, and no lime if pH was >6.0. Levels of lime applied in 2007 were 5 t/ha for pH <4.5, 3 t/ha for pH 4.5-4.9, 2 t/ha for pH 5.0-6.0, and no lime for pH >6.0. Levels of lime applied in 2008 and 2009 were 5 t/ha for pH <5.0, 3 t/ha for pH 5.0-5.4, 1 t/ha for pH 5.5-6.0, and no lime for pH >6.0.
Tissue testing indicated copper (Cu), zinc (Zn), and manganese (Mn) applied in previous years were still adequate for pasture production, so no fertiliser Cu, Zn, or Mn was applied during 1999-2009.
During 2000-03 in the VMF project, fertiliser N was applied after each grazing, when 1 kg N/ha.day was applied, and a total of ~121 kg N/ha was applied in each growing season (Staines et al. 2007), normal practice at the time. In the VMF project, depending on treatment, 3-49 kg P/ha was imported in concentrate, and to ensure P introduced in the concentrate did not affect pasture dry matter production, 51 kg P/ha was applied as fertiliser P each year (Staines et al. 2007). The district average for dairy farms at the time ranged from 20 to 40 kg P/ha, depending on the intensity of grazing. During 2000-03, using district practice, a total of 100 kg K and 75 kg S/ha was applied each year. One-third of the P, K, and S was applied after the start of the growing season in autumn and two-thirds in late winter when paddocks were dry enough to be trafficable (Staines et al. 2007).
During the GP1 project (2005-09), fertiliser N levels applied varied from 0 to 331 kg N/ha.year, depending on the N treatment (Staines et al. 2008). In each year during 2004-09, fertiliser P was only applied when soil-test P was below the critical value for the sand to sandy loam soils, with critical values being 15-40 mg P/kg soil. Therefore, the few paddocks that were treated with P fertiliser in these years received 10-20 kg P/ha applied 3 weeks after pasture emerged at the start of the growing season (Staines et al. 2008). During the GP1 project, fertiliser N, together with ~10 kg K and 5-7 kg S/ha, was applied after each grazing (Staines et al. 2008), the practice now adopted by dairy farmers for intensively grazed ryegrass pastures in the region. A mean total of 63 kg K and 34 kg S/ha was applied each year during 2005-09 (Staines et al. 2008).
In both the VMF and GP1 projects, fertiliser N, P, K, and S was applied as various mixtures of urea, ammonium sulfate, single superphosphate, triple superphosphate, potassium chloride, and ammonium phosphate fertilisers in custom N : P : K : S fertiliser blends.
Samples of ryegrass herbage were collected from selected paddocks at different times in each growing season during 1999-2009 for tissue testing. The ryegrass samples were collected just before grazing when ryegrass plants had 2 3 leaves per tiller. Using district practice, young tissue was collected at each sampling by collecting samples 5 cm above the soil surface, being the pasture typically grazed by lactating dairy cows at each grazing. The samples collected from each paddock were bulked and dried in a forced-draught oven at 50[degrees]C for 2 days. The dried samples were ground and sent to CSBP laboratories for analyses. Subsamples of the dried, ground herbage were dissolved in a nitric and perchloric acid mixture (Yuen and Pollard 1954) and concentrations of elements in the digest were measured by inductively coupled plasma-atomic emission spectrometry (ICP-AES) (Zarcinas et al. 1987).
Analysis of data
For each soil property in each year, data presented are mean and range (minimum and maximum) values for all 48 paddocks.
Liming increased soil pH from the range 4.0-5.0 in 1999 to 5.2-6.0 in 2009 (Fig. 1). Soil pH of 5.5 was achieved in individual paddocks by 2007-09, i.e. 9-11 years after the liming program started, and required a total of 12-21 t/ha lime (Table 1). For the last soil testing in 2009, 14 paddocks (~29% of paddocks) were yet to achieve pH 5.5, but pH values for these paddocks were 5.2-5.4, so close to the target value.
Critical soil-test P for the 48 paddocks, and ranking the capacity of the soils to sorb P
Based on all available published and unpublished data for various soil types of south-western Australia, most unpublished, critical Colwell soil-test P has recently been summarised for different soils with different capacities to sorb P, as ranked using reactive Fe, PRI, and PBI (Table 2).
[FIGURE 1 OMITTED]
Mean reactive Fe, PRI, and PBI values measured in the different years for each of the 48 paddocks are listed in Table 1. Based on these mean values, data listed in Table 2 were used to provide critical Colwell soil-test P for each of the paddocks, as ranked using reactive Fe, PRI, and PBI (Table 1). The general trend was for reactive Fe to determine lower critical soil-test P values than the two single-point P-sorption indices, with PBI often estimating the largest critical soil-test P values (Table 1). For each paddock, there were variations between years for all three procedures, with variations being larger for reactive Fe than PRI and PBI. Note that comparisons are limited to 3 years for PRI and 4 years for PBI, compared with 9 years for reactive Fe.
Soil-test P significantly increased from 1999 to 2004 during the VMF project when 51 kg fertiliser P/ha was applied each year (Fig. 2a, b). In the VMF project, P was also imported in concentrate fed to the cows, increasing P surpluses from 9 to 21 kg P/ha.year as the amount of P imported in concentrate increased from 3 to 49 kg P/ha.year (Staines et al. 2007).
During 20044)9 in the GP1 project, fertiliser P was only applied in the next growing season when soil test P was below the critical value for the sand to sandy loam soils. Most soil test P values were above the critical value (Fig. 2c), so few of the 48 paddocks were fertilised with P in these years. Mean soil-test P significantly declined gradually in these years, with soil-test values remaining above critical values in most paddocks up to 2009 (Fig. 2c).
Soil-test K varied greatly among paddocks (Fig. 3) and, for each paddock, varied greatly between years (Fig. 3b). Soil-test Kwas mostly lower for sand than sandy loam soils (Fig. 3e) and tended to decline after silage and hay crops and increased in paddocks where silage and hay were fed to stock.
Soil-test S values ranged from just over l0 to just over 300mg/kg, with mean values being ~40-77 mg/kg, and were always above the critical value of ~6.5mg/kg for pasture production (Fig. 4).
Organic C content of soil
Organic C content of soils ranged from ~21 to 95 g/kg, with mean values for all 48 paddocks in each year being ~38-53 g/kg (Fig. 5). There was a significant long-term (11-year) trend for organic C to increase slightly with time (Fig. 5).
Electrical conductivity ranged from 0.08 to 1.57dS/m, with mean values for all 48 paddocks in each year being 0.29-0.47 dS/m (Fig. 6).
Tissue testing indicated that P, K, S, Cu, Zn, Mn, Fe, Ca, Mg, and B were adequate to high for ryegrass production (Table 3). Tissue testing indicated that N was adequate during the VMF and GP1 projects, except that it was deficient for the nil-N treatment of the GP1 project (Table 3).
If soil testing identifies soil acidification as a problem, liming with high levels of good quality lime should be undertaken as soon as possible and continued until the desired pH is achieved. Where possible, lime should be incorporated into soil after application to increase its effectiveness. Soil acidification continues, both during liming and after liming has been successfully completed, particularly for intensively grazed pastures treated with fertiliser N after each grazing. Therefore, once the target pH has been achieved, soil pH should be measured annually and lime again applied when pH declines below 5.5. This strategy requires smaller amounts of lime (1 t/ha) to return soil pH to [greater than or equal to] 5.5 than allowing soil pH to decline to lower values.
[FIGURE 2 OMITTED]
[FIGURE 3 OMITTED]
Progress in ameliorating soil acidification was disappointing during 1999-2003 (Fig. 1). There were three possible reasons for this. First, lime was not incorporated into soil after application to the soil surface, so liming was less effective (McLaughlin et al. 1990; Scott et al. 2000). Second, continued soil acidification during 1999 2009 increased the requirement for lime to ameliorate the acidity. Third, amounts of lime applied were limited by cost; no lime was applied in some years and only low levels were applied in other years. As the liming program continued and liming eventually increased soil pH, applications of lime after 2006 were more effective at ameliorating the problem (Fig. 1).
Although lime is more effective at ameliorating soil acidification when it is incorporated into soil after application (McLaughlin et al. 1990; Scott et al. 2000), dairy farmers in south-western Australia are reluctant to cultivate their soils. This is because in most situations the topography is flat and the soils are waterlogged for most of the growing season. After cultivation, managing the wet, unconsolidated soils can be difficult (Bolland et al. 2002a). In addition, cultivating soil increases mineralisation of soil organic matter, resulting in loss of soil fertility built up by growing pasture over many years when the soils were not cultivated.
[FIGURE 4 OMITTED]
Ranking the capacity of soil to sorb P
The capacity of soil to sorb P decreases as more P is applied to soil (Barrow 1974; Bolland and Allen 2003). This makes it difficult to rank the capacity of soils to sorb P using single-point P-sorption indices. However, adding soil-test P to P sorption when calculating single-point P sorption indices estimates P sorption for newly cleared, P-deficient soil, regardless of the P fertiliser history of that soil, making it simpler to rank the capacity of soil to sorb P (Barrow 2000). Therefore, PBI is calculated by adding soil-test P to P sorption (adjusted PBI; Burkitt et al. 2002). However, soil test P is not added to P sorption when calculating PRI; thus, PRI provides an estimate of the current capacity of soil to sorb P (Allen and Jeffery 1990). Phosphorus is likely to leach in sandy soils of south-western Australia when PRI is <2 L/kg (D. G. Allen and R. C. Jeffery, unpubl, data). We suggest that laboratories report both unadjusted and adjusted PRI and PB1 so that both values are available to provide appropriate fertiliser advice.
[FIGURE 5 OMITTED]
[FIGURE 6 OMITTED]
Spatial variation has been found to be large for reactive Fe and PRI, particularly reactive Fe (Bolland and Allen 1998).
Therefore, for the same paddock, different values for reactive Fe, PRI, and PBI measured in different years in this study are attributed to spatial variation.
In this study, there was a poor relationship between PRI and reactive Fe, a better relationship between PBI and reactive Fe, with the best relationship between PBI and PRI. These findings support results of previous studies showing poor relationships between direct measures of P sorption and reactive Fe (Singh and Gilkes 1991; Brennan et al. 1994; Bolland et al. 1996; Burkitt et al. 2002; Bolland and Windsor 2007), and better relationships between PRI and PBI (Burkitt et al. 2002; Bolland and Windsor 2007).
Soil-test P increased dramatically from 1999 to 2004 because 51 kg P/ha was applied in each of these years to all 48 paddocks. However, during 2004-09, when no P was applied to most paddocks because soil-test P was above the critical value, soil-test P decreased gradually while remaining above critical values in most paddocks up to 2009 (Fig. 2c). For paddocks not treated with P in this period, tissue testing indicated P was adequate for ryegrass production (Table 3).
The GP1 project was an N study, not a P study, but a similar result to that shown in Fig. 2c for this study occurred in a P x N field experiment in the region (Fig. 7, presenting data of Bolland and Guthridge 2007a). The data in Fig. 7 were not presented as such by Bolland and Guthridge (2007a). The P x N experiment was done on an intensively grazed dairy pasture where 6 levels of fertiliser P (0, 15, 30, 45, 90, and 180kg P/ha.year, as triple superphosphate) were applied with either no or adequate fertiliser N. The pasture was dominated by ryegrass when fertiliser N was applied after each grazing. When no N was applied, the pastures comprised >90% clover and ryegrass, and 30-70% of each species, depending on the N status of soil, grazing and fertiliser management, and seasonal conditions (Rossiter 1966). The soil-test P values were not significantly different for the two N treatments. The soil-test data shown in Fig. 7 are for the nil-P treatment and significantly, gradually decreased from 73 to 48mg/kg during 2000-04 when no fertiliser P was applied, and were always above the critical Colwell soil-test P of ~35 mg/kg for the soil type on which the experiment was located. Consequently, no response of pasture dry matter to applied P was obtained. The amount of pasture dry matter consumed by dairy cows at all grazing events in each year when adequate fertiliser N was applied is shown near the top of Fig. 7 and was the same for the nil-P and the five P levels
applied each year. Corresponding pasture yields when no fertiliser N was applied were ~2 t/ha less than the yield data for the adequate N treatment. The experiment started in mid-June 2000; hence, total pasture dry matter consumed in that year (4.9t/ha v. 9.3-12.3t/ha in the other years) was smaller because pasture yields were not measured early in the 2000 growing season before the experiment started.
[FIGURE 7 OMITTED]
We conclude soil P testing can be used to determine when to apply fertiliser P to pasture in the next growing season. No fertiliser P is required when soil test P is above the critical value for that soil, particularly if it is well above the critical value. If fertiliser P is applied to such soils, no pasture dry matter responses to applied P are obtained. When soil test P is above the critical value and no fertiliser P is applied, soil test P decreases gradually.
Because of K leaching in the next growing season, soil K testing could not be used to determine fertiliser K requirements of ryegrass in the next growing season. In addition, soil-test K values were very variable. Previous studies in high-rainfall areas of south-western Australia (Bolland et al. 2002b; Bolland and Guthridge 2009) have shown that soil K testing is unreliable for indicating K deficiency for pasture production in the next growing season. In addition, while cows graze pasture they deposit urine, containing plenty of K, in patches on the pasture. The urine patches are deposited in the 5 to 8 grazings undertaken in each paddock in each growing season. Soil test K is greatly affected by how many soil samples are collected for soil testing each year within and between the urine patches deposited during the previous growing season (Bolland and Guthridge 2009).
Therefore, in this study, fertiliser K was applied each year, and tissue testing was used to assess whether the fertiliser K strategies provided adequate K for ryegrass production. Each year during 2000-03 in the VMF project, ~30kg K/ha was applied in autumn and ~70 kg K/ha in late winter-early spring, so a total of ~100kg K/ha was applied each growing season. In the GP1 project, ~10 kg K/ha was applied after each grazing, and a total of ~63 kg K/ha was applied in each growing season. Tissue testing indicated that both methods of applying fertiliser K provided adequate K for pasture production (Table 3).
A critical KC1-40 soil-test S value of-6.5 mg/kg has been determined for subterranean clover-based pasture (Blair et al. 1991), for pasture (various species) in >560 mm average annual rainfall areas of Australia (Blair et al. 1997), and for canola (Brassica napus L.) grain production on sandy soils in southwestern Australia (Brennan and Bolland 2006). In this study, soil-test S values were always >6.5 mg/kg (Fig. 4).
Because S mineralised from soil organic matter or dissolved from applied fertiliser can be leached, soil S testing cannot be used to determine fertiliser S requirements of high-rainfall pastures in the next growing season. The extent of leaching of S cannot be predicted. Therefore, the recommendation for these pastures is to apply fertiliser S after July every year and use tissue testing to assess and improve S management for the pastures. In the VMF study during 2000-03, fertiliser S was applied twice a year, with ~25 kg S/ha applied in autumn and ~50kgS/ha applied late winter-early spring, so a total of ~75kg S/ha was applied in each growing season. In the GP1 project during 2004-09, 5-7kgS/ha was applied after each grazing, and a total of-34kg S/ha each growing season. Tissue testing indicated that both fertiliser strategies provided enough S for ryegrass production (Table 3).
Organic C content of soil
The data from this project confirm that organic C contents of the top 0.10 m of dairy soils in the region are high. The organic C content of the soils tended to significantly increase in time, supporting results of an earlier study for permanent annual pastures in the region when the soil was not cultivated (Barrow 1969). Fluctuations of soil organic C contents in different years may be due to large spatial and/or temporal variations for this soil property in different years (Brown 1999).
Electrical conductivity of soil
For sandy loam soils, soil salinity is likely to reduce annual ryegrass DM production when electrical conductivity values are >0.33 dS/m (D. L. Bennett, unpubl, data). Depending on year, electrical conductivity was above this critical value in 23-70% of the 48 paddocks (Fig. 6).
In most of the 48 paddocks there were small patches that were more saline than the remainder of the paddock. Electrical conductivity was not measured in these patches, but would have been larger than for the remainder of the paddock. Emergence of pasture at the start of the growing season was delayed in these patches, particularly following drier growing seasons.
Soil properties measured by commercial soil testing laboratories
Commercial soil-testing laboratories always measure and charge for a standard number of soil properties that are potential candidates for soil testing in all agricultural industries in south-western Australia. However, for specific agricultural industries in specific regions, we encourage the laboratories to measure only those soil properties that research has shown to be reliable and for which the results can be interpreted by advisors and consultants. This would reduce costs of analyses, encouraging farmers to soil test. For the dairy areas of southwestern Australia, this would include measuring soil pH and soil-test P every year, together with PBI, organic C content of soils, and EC every 3-5 years. For soils prone to P leaching, PBI/ PRI should be measured every year.
This study confirmed that both soil and tissue testing need to be undertaken when assessing the reliability of measuring different soil properties to help determine lime and fertiliser requirements for different agricultural enterprises in different regions, and to identify nutrient elements deficiencies not covered by soil testing. This applies to all agriculture.
In this study, soil testing for pH and P each year successfully targeted paddocks requiring lime to ameliorate soil acidification and requiring applications of fertiliser P to maintain ryegrass production. Once the acidity has been ameliorated, measurement of soil pH needs to continue regularly to target paddocks requiring lime, to prevent acidification reoccurring. No fertiliser P is required when soil-test P is above the critical value for that soil. Tissue testing indicated that P was not deficient for ryegrass production when no P was applied if it was not required. These findings are relevant to all soils, climates, and environments where soil acidification and P deficiency are problems.
In soils where leaching of K and S below the root-zone occurs, soil K and S testing cannot be used to predict fertiliser K and S requirements in the next growing seasons. Because it is not possible to predict the extent of leaching of these elements below the root-zone, the solution is always to apply K and S fertiliser each year, and use tissue testing to determine the appropriate fertiliser K and S fertiliser strategies required to ensure K and S are sufficient for pasture production. These findings are relevant to all soils prone to K and S deficiency due to leaching of the elements below the root-zone.
For soil testing in each region, it is worthwhile to measure only soil properties that research has shown to be reliable for helping to determine lime and fertiliser requirements for specific agricultural systems, such as dairy production in south-western Australia. Commercial soil and tissue testing laboratories need to be flexible, and rather than routinely measuring, reporting, and charging for a standard number of soil properties, they should only measure and report those soil properties required. This will reduce cost of analyses and encourage soil testing. This applies to all regions where soil testing is used.
Technical assistance was provided by John Baker, Patrick Donnelly, Leonarda Paszkudzka-Baizert, Kathy Lawson, and Andrew Lindsay. Martin Staines, Richard Morris, and Peter Jelinek helped collect soil samples. Peter Jelinek was responsible for data management of the VMF and GPI projects, including the soil test data reported in this paper. Funds were provided by the Government of Western Australia and Dairy Australia (Vasse Milk Farmlets DAW046, and Greener Pastures 1 DAW11011). CSBP did not charge for chemical analyses during 19992004. Positive comments of anonymous referees helped us improve the paper.
Manuscript received 4 November 2009, accepted 29 April 2010
Allen DG, Jeffery RC (1990) Methods of analysis of phosphorus in Western Australian soils. Report of Investigation No. 37. Chemistry Centre (WA), East Perth.
Arkell P (1995) Productive pastures pay: a manual on pasture establishment and management for the above 700mm rainfall zone. Agriculture Western Australia, Bulletin No. 4302. Department of Agriculture and Food, South Perth.
Barrow NJ (1966) The residual value of the phosphorus and sulfur components of superphosphate on some Western Australian soils. Australian Journal of Experimental Agriculture and Animal Husbandry 6, 9-16. doi:10.1071/EA9660009
Barrow NJ (1969) The accumulation of soil organic matter under pasture and its effect on soil properties. Australian Journal of Experimental Agriculture and Animal Husbandry 9, 437-444. doi:10.1071/ EA9690437
Barrow NJ (1974) Effect of previous addition of phosphate on phosphate adsorption by soil. Soil Science 118, 82 89. doi:10.1097/00010694-197408000-00004
Barrow NJ (1975) The responses to phosphate of two annual pasture species. I. Effect of the soil's ability to adsorb phosphate on comparative phosphate requirement. Australian Journal of Agricultural Research 26, 137-143. doi:10.1071/AR9750137
Barrow NJ (2000) Towards a single-point method for measuring phosphate sorption by soils. Australian Journal of Soil Research 38, 1099-1113. doi:10.1071/SR99135
Blair G J, Anderson GC, Crestani M, Lewis DC (1997) Soil sulfur fertiliser responses and rainfall sulfur accessions at trial sites in the National Reactive Phosphate Rock Project. Australian Journal of Experimental Agriculture 37, 995-1001. doi:10.1071/EA96114
Blair GJ, Chinoim N, Lefroy RDB, Anderson GC, Crocker GJ (1991) A sulfur soil test for pastures and crops. Australian Journal of Soil Research 29, 61%626. doi:10.1071/SR9910619
Bolland MDA, Allen DG (1998) Spatial variation of soil test phosphorus and potassium, oxalate-extractable iron and aluminium, phosphateretention index, and organic carbon content of soils of Western Australia. Communications in Soil Science and Plant Analysis 29, 381-392. doi:10.1080/00103629809369952
Bolland MDA, Allen DG (2003) Phosphorus sorption by sandy soils from Western Australia: effect of previously sorbed P on P buffer capacity and single-point P sorption indices. Australian Journal of Soil Research 41, 1369-1388. doi:10.1071/SR02098
Bolland MDA, Allen DG, Rengel Z (2002a) Responses of annual pastures to application of limestone in the high rainfall areas of south-western Australia. Australian Journal of Experimental Agriculture 42, 925 937. doi:10.1071/EA01169
Bolland MDA, Cox W J, Codling BJ (2002b) Soil and tissue tests to predict pasture yield responses to applications of potassium fertiliser in high-rainfall areas of south-western Australia. Australian Journal of Experimental Agriculture 42, 149-164. doi:10.1071/EA01060
Bolland MDA, Gilkes R J, Brennan RF, Allen DG (1996) Comparison of seven phosphorus sorption indices. Australian Journal of Soil Research 34, 81 89. doi:10.1071/SR9960081
Bolland MDA, Guthridge IF (2007a) Determining the fertiliser phosphorus requirements of intensively grazed dairy pastures in south-western Australia with or without adequate nitrogen fertiliser. Australian Journal of Experimental Agriculture 47, 801-814. doi:10.1071/ EA05184
Bolland MDA, Guthridge IF (2007b) Responses of intensively grazed dairy pastures to applications of fertiliser nitrogen in south-western Australia. Australian Journal of Experimental Agriculture 47, 927 941. doi:10.1071/EA06014
Bolland MDA, Guthridge IF (2009) Quantifying pasture dry matter responses to applications of potassium fertiliser for an intensively grazed, rain-fed dairy pasture in south-western Australia with or without adequate nitrogen fertiliser. Animal Production Science 49, 121 130. doi:10.1071/EA08106
Bolland MDA, Windsor DP (2007) Converting reactive iron, reactive aluminium, and phosphorus retention index (PRI) to the phosphorus buffering index (PBI) for sandy soils of south-western Australia. Australian Journal of Soil Research 45, 262265. doi: 10.1071/SR07026
Bolland MDA, Yeates JS, Clarke MF (2003) Comparing different sources of sulphur for high-rainfall pastures in south-western Australia. Australian Journal of Experimental Agriculture 43, 1221-1229. doi:10.1071/ EA02146
Bowden JW, Shedley C, Burgess SJ (1993) Soil test and phosphorus rate. Technote No. 5/93. Department of Agriculture and Food Western Australia, South Perth.
Brennan RF, Bolland MDA (2006) Soil and tissue tests to predict the sulfur requirements of canola in south-western Australia. Australian Journal of Experimental Agriculture 46, 1061-1068. doi:10.1071/EA04206
Brennan RF, Bolland MDA, Jeffrey RC, Allen DG (1994) Phosphorus adsorption by a range of Western Australian soils related to soil properties. Communications in Soil Science and Plant Analysis 25, 27852795. doi: 10.1080/00103629409369225
Brown AJ (1999) Soil sampling and sample handling for chemical analysis. In 'Soil analysis an interpretation manual'. (Eds KI Peverill, LA Sparrow, DJ Reuter) pp. 35 53. (CSIRO Publishing: Melbourne)
Burkitt LL, Moody PW, Gourley CJP, Hannah MC (2002) A simple phosphorus buffering index for Australian soils. Australian Journal of Soil Research 40, 497-513. doi:10.1071/SR01050
Colwell JD (1963) The estimation of phosphorus fettiliser requirements of wheat in Southern New South Wales by soil analysis. Australian Journal of Experimental Agriculture and Animal Husbandry 3, 190-197. doi: 10.1071/EA9630190
Dolling PJ, Porter WM (1994) Acidification rates in the central wheatbelt of Western Australia. 1. On a deep yellow sand. Australian Journal of Experimental Agriculture 34, 1155-1164. doi:10.1071/EA9941155
Dolling PJ, Porter WM, Robson AD (1991) Effect of soil acidity on barley production in the south-west of Western Australia. 2. Cereal genotypes and their responses to lime. Australian Journal of Experimental Agriculture 31, 811-818. doi:10.1071/EA9910811
Fulkerson W J, Donaghy DJ (2001) Plant-soluble carbohydrate reserves and senescence key criteria for developing an effective grazing management system for ryegrass-based pastures: a review. Australian Journal of Experimental Agriculture 41, 261 275. doi:10.1071/ EA00062
Hodgkin EP, Hamilton BH (1993) Fertilisers and eutrophication in southwestern Australia: setting the scene. Fertilizer Research 36, 95 103. doi: 10.1007/BF00747579
Isbell RF (2002) 'The Australian Soil Classification.' 2nd edn (CSIRO Publishing: Melbourne)
McArthur WM (2004) 'Reference soils of south-western Australia.' 2004 reprint. (Australian Society of Soil Science, WA Branch Inc.: Perth)
McComb A J, Davis JA (1993) Eutrophic waters of southwestern Australia. Fertilizer Research 36, 105-114. doi: 10.1007/BF00747580
McLaughlin MJ, Baker TG, Rundle JA (1990) Distribution and forms of phosphorus and aluminium in acidic topsoils under pasture in southeastern Australia. Australian Journal of Soil Research 28, 371 385. doi: 10.1071/SR9900371
Moody PW (2007) Interpretation of a single-point P buffering index for adjusting critical levels of the Colwell soil P test. Australian Journal of Soil Research 45, 55-52. doi: 10.1071/SR06056
Olsen SR, Cole CV, Watanabe FS, Dean LA (1954) Estimation of available phosphorus in soils by extraction with sodium bicarbonate. United States Department of Agriculture Circular No. 939.
Ozanne PG, Shaw TC (1967) Phosphate sorption by soils as a measure of the phosphate requirements for pasture growth. Australian Journal of Agricultural Research 18, 601-612. doi:10.1071/AR9670601
Ozanne PG, Keay J, Bidiscombe EF (1969) The comparative applied phosphate requirements of eight annual pasture species. Australian Journal of Agricultural Research 20, 809-818. doi: 10.1071/AR9690809
Pinkerton A, Smith FW, Lewis DC (1997) Pasture species. In 'Plant analysis an interpretation manual'. (Eds DJ Reuter, JB Robinson) pp. 287 346. (CSIRO Publishing: Melbourne)
Rayment GE, Higginson FR (1992) 'Australian laboratory handbook of soil and water chemical methods.' (Inkata Press: Melbourne)
Rossiter RC (1966) Ecology of the Mediterranean annual-type pasture. Advances in Agronomy 18, 1- 56. doi:10.1016/S0065-2113(08)60647-1
Russell WK (2010) Soil testing high rainfall pastures. Farmnote 418. Department of Agriculture and Food Western Australia, South Perth.
Schwertmann U (1964) Differenzierung der eisenoxide des bodens durch photochemische extraktion mit sauer ammoniumoxalate-losung. Zeitschrift fur Pflanzenernahrung und Bodenkunde 105, 194-202. doi: 10.1002/jpln.3591050303
Scott BJ, Ridley AM, Conyers MJ (2000) Management of soil acidity in long-term pastures of south-eastern Australia: a review. Australian Journal of Experimental Agriculture 40, 1173-1198. doi:10.1071/ EA00014
Shaw RJ (1999) Soil salinity electrical conductivity and chloride. In 'Soil analysis an interpretation manual'. (Eds KI Peverill, LA Sparrow, DJ Reuter) pp. 12%145. (CSIRO Publishing: Melbourne)
Singh B, Gilkes RJ (1991) Phosphorus sorption in relation to soil properties for the major soil types of south-western Australia. Australian Journal of Soil Research 29, 603~18. doi:10.1071/SR9910603
Soil Survey Staff (1987) 'Keys to Soil Taxonomy.' Soil Management Support Services Technical Monograph No. 6. (USDA/Cornell University: Ithaca, NY)
Staines M, Morris R J, Bennett DL, White CL, Bolland MDA, Lucey J, Russell WK, Windsor DP (2007) DAW046--Vasse Milk Farmlets: productivity, profitability and sustainability of intensive dryland dairy systems in the Mediterranean climate zone of Australia. Final Report to Dairy Australia.
Staines M, Morris RJ, Bolland MDA, Bennett DL, Fillery IRP, Lucey J, Russell WK (2008) DAW11011-Greener Pastures: profitable and sustainable use of nitrogen in dairy grazing systems (phase 1). Project Report to Dairy Australia.
Tille PJ, Lantzke NC (1990) Busselton-Margaret River Augusta land capability study. Department of Agriculture Western Australia, Land Resource Series No. 5.
Walkley A, Black 1A (1934) An examination of the Degtjareff method for determining soil organic matter and a proposed modification of the chromic acid titration method. Soil Science 37, 29 38. doi:10.1097/ 00010694-193401000-00003
Whitten MG, Wong MTF, Rate AW (2000) Amelioration of subsurface acidity in south-western Australia: downward movement and mass balance of surface incorporated lime after 2-15 years. Australian Journal of Soil Research 38, 711 728. doi:10.1071/SR99054
Yuen SH, Pollard AG (1954) Determination of nitrogen in agricultural materials by the Nessler reagent. I1. Micro-determinations in plant tissue and in soil extracts. Journal of the Science of Food and Agriculture 5, 364-369. doi:10.1002/jsfa.2740050803
Zarcinas BA, Cartwright B, Spouncer R (1987) Nitric acid digestions and multi-element analysis of plant material by inductively coupled plasma spectrometry. Communications in Soil Science and Plant Analysis 18, 131-146. doi:10.1080/00103628709367806
M. D. A. Bolland (A,B,C) and W. K. Russell (A)
(A) Department of Agriculture and Food Western Australia, PO Box 1231, Bunbury, WA 6231, Australia.
(B) School of Plant Biology, The University of Western Australia, 35 Stirling Highway, Crawley, WA 6009, Australia.
(C) Corresponding author. Retired. Present address: 35 Casuarina Street, Eaton, WA 6232, Australia.
Table 1. Total lime applied during 1999-2009 for the 48 paddocks, and mean reactive Fe (mg/kg), P retention index (PRI, L/kg), and P buffering index (PBI, no units), and corresponding critical Colwell soil-test P (critical P, mg/kg) determined from the reactive Fe, PRI, and PBI ranges listed in Table 2 Paddocks are listed in order from Greener Pastures 1 (GPI) paddocks 12 48 and 50-60. ND, No data Paddock Lime Reactive Fe (1999 2007) PRI no. applied (t/ha) Mean Critical P Mean 1 18.5 350 25 3.5 2 19.5 187 15 ND 3 16.5 249 20 6.1 4 18.0 230 20 1.8 5 17.5 178 15 0.1 6 14.0 224 20 1.0 7 19.5 234 20 0.9 8 19.5 286 25 2.2 9 19.0 255 20 0.8 10 19.5 295 25 2.6 11 19.5 273 20 1.8 12 15.5 485 25 9.1 13 15.0 343 25 4.9 14 14.0 323 25 6.1 15 14.0 355 25 6.1 16 14.0 419 25 5.1 17 15.0 493 25 6.6 18 14.0 609 25 8.6 19 14.0 544 25 6.9 20 13.0 614 25 8.8 21 15.5 571 25 9.1 22 15.5 719 30 10.3 23 14.0 728 30 12.3 24 14.0 1337 35 32.0 25 14.0 1193 30 19.8 26 14.0 818 30 9.5 27 16.0 846 30 9.8 28 17.5 1199 30 17.1 29 21.0 695 30 13.5 30 19.5 909 30 14.9 31 17.5 1127 30 72.1 32 17.0 822 30 16.2 33 16.0 847 30 20.4 34 15.5 1233 30 37.8 35 14.0 1635 35 68.0 36 19.5 1592 35 73.9 37 17.5 1498 35 95.3 38 19.5 1074 30 370.6 39 19.0 1067 30 259.7 40 16.0 537 25 83.1 41 16.0 449 25 24.5 42 18.5 472 25 15.1 43 19.0 311 25 11.4 44 19.0 228 20 7.4 45 20.0 188 15 3.2 46 21.5 168 15 3.8 47 20.0 195 15 2.9 48 15.0 222 20 3.1 Paddock (2004-07) PBI (2006-09) no. Critical P Mean Critical P 1 25 32.2 25 2 ND 16.3 25 3 25 38.3 30 4 20 18.8 25 5 20 14.2 20 6 20 17.6 25 7 20 17.6 25 8 25 21.5 25 9 20 18.5 25 10 25 23.6 25 11 20 20.7 25 12 30 50.1 30 13 25 35.6 30 14 25 35.3 30 15 25 40.8 30 16 25 42.1 30 17 25 38.4 30 18 25 48.1 30 19 25 46.5 30 20 25 47.7 30 21 30 45.3 30 22 30 49.3 30 23 30 59.4 30 24 35 99.2 35 25 30 88.1 35 26 30 67.8 30 27 30 68.4 30 28 30 66.4 30 29 30 57.8 30 30 30 71.3 35 31 35 105.2 35 32 30 64.6 30 33 30 71.8 35 34 35 90.7 35 35 35 132.7 35 36 35 143.0 40 37 40 126.7 35 38 40 265.9 40 39 40 215.0 40 40 35 153.6 40 41 30 85.0 35 42 30 64.0 30 43 30 51.7 30 44 25 46.2 30 45 25 34.1 30 46 25 27.4 25 47 25 32.8 30 48 25 35.2 30 Table 2. Critical Colwell soil-test P for different soils in south-western Australia and corresponding P sorption of soil as ranked using reactive Fe, PRI, and PBI From published and unpublished data summarised early 2009 by R. N. Summers and D. M. Weaver. Critical soil test P is for 95% of the maximum pasture yield Critical Colwell Reactive Fe PRI PBI soil-test P (mg/kg) (mg/kg) (L/kg) 15 100-200 <0 5-10 20 201-280 0-2 11-15 25 281-650 3-9 16-30 30 651-1250 10-28 31-70 35 1251-2500 29-87 71-140 40 2501-4950 88-275 141-280 Table 3. Published ranges for critical tissue test values for young tissue of perennial ryegrass as summarised by Pinkerton et al. (1997, p. 345), and range of tissue test values obtained in this study as determined for young tissue of mixed annual and ltalian ryegrass pasture measured on samples collected in selected paddocks severai times each growing season during 2000-09 Element Published critical Range of tissue test values tissue test values obtained in this study (A) N (g/kg) 30-35 37-50 (B) P (g/kg) 2.0-2.8 3.1-3.8 K (g/kg) 14-19 23-36 S (g/kg) 1.8-2.2 3.6-5.7 Ca (g/kg) 1.5-2.0 3.6--7.2 Mg (g/kg) 1.5-2.0 2.5-3.9 Cu (mg/kg) 4.0-6.0 8.1-15.2 Zn (mg/kg) 10-15 30-54 Mn (mg/kg) 15-20 29-89 Fe (mgikg) 40-60 87-1033 B (mg/kg) 3.0-5.0 8.3-16.6 (A) For each element, the lowest and highest values obtained for all samplings in all years. (B) Did net include values for the nil-N treatment of the GP1 project, which were 22-29 g/kg of N so below the critical range of 30-35 g/kg.
|Printer friendly Cite/link Email Feedback|
|Author:||Bolland, M.D.A.; Russell, W.K.|
|Publication:||Australian Journal of Soil Research|
|Date:||Dec 1, 2010|
|Previous Article:||Assessing P fertiliser use in vegetable production: agronomic and environmental implications.|
|Next Article:||Characterisation of dissolved organic matter in water extracts of thermally altered plant species found in box-ironbark forests.|