Grain mineral contents in southeastern Australia

Introduction
An understanding of the nutrient concentrations of grains such as wheat is important in developing nutrient budgets for both macronutrients and micronutrients. The values can also be used diagnostically to retrospectively assess particular nutrient deficiencies or toxicities for the crop from which the grain was derived. There have been several studies undertaken on grain nutrient densities in Australia (e.g. Schultz and French 1978) and much of that information has been collated and published in “Plant Analysis – An Interpretation Manual” (Reuter et al. 1997). These published values are considered as benchmarks and were used in developing regional nutrient budgets as part of the National Land and Water Resources Audit (2001) (Table 1).

Table 1 Nutrient wheat grain concentrations (mg/kg, 0% moisture) from Reuter as used in the National Land and Water Resources Audit (2001) and critical values are taken from Reuter et al. (1997).

	P	K	S	Ca	Mg	B	Cu	Mn	Zn
Wheat	2900	4000	1600	430	1400		-	-	-
Critical Values	2700	5000	1200	-	-	<2.0	1.0-2.5	20	5-15

However, regional, cultivar and annual changes in grain nutrient concentrations give a degree of uncertainty to possible long term nutrient balances calculated as the product of grain yield and nutrient density. The research reported here summarises nutrient densities for wheat from across in south-eastern Australia so that there can be confidence in the values used in nutrient budgets. It also provides a data set which can be used to test the amount of genetic, temporal and spatial variability in grain nutrient densities.

Materials and Methods
Wheat grain samples were obtained from site managers involved in the National Variety Testing (NVT) system. A single sample per site of two varieties, Yitpi and Gladius, was collected from the 2009 season. There were eight sites from New South Wales, 21 sites from South Australia and 17 from Victoria and these were in 12 agro-ecological regions across southeastern Australia. Samples from 23 sites from South Australian in 2008 were also included. So, a total of 70 sites across two years were analysed for grain nutrient concentration.

The NVT sites are managed using commercial best practice, which includes regional fertilizer products and rates, as well as normal establishment and crop protection operations. Crop management practices, soil test values and grain yields were taken from the NVT reports published in 2008 and 2009.

Between 18 and 30 grains of each cultivar (approx. 0.8 g) from each site were randomly selected from the harvested grain sub-sample, dried, weighed and processed for nutrient analysis by ICP-OES. Grain was digested with 11 ml of nitric acid (HNO₃)/perchloric acid (HCIO₄) mixture (10:1 v/v), boiled down to approx. 1 ml of HCIO₄ and made to 25 ml final volume using de-ionised water. This final solution was then analysed for nutrients on Inductively Coupled Plasma Optical Emission Spectrometer (ICP-OES; ARL 3580 B, Appl. Res Lab. SA, Ecublens, Switzerland) and results are reported on a dry grain basis. Analytes reported from this analysis are Al, B, Ca, Cd, Co, Cr, Cu, Fe, K, Mg, Mn, Mo, Na, Ni, P, Pb, S, Se, Ti, and Zn, but this report will only present data for B, Ca, Cu, K, Mg, Mn, P, S, and Zn and all are given on a dry grain basis.

Because the data set developed was not balanced, nor were there replicated samples for each cultivar from each site in each year, the data were assessed using a one-way analysis of variance to compare nutrient densities using either state (Victoria, New South Wales, South Australia), site (70 site years) or varieties (Gladius and Yitpi). In addition, the data set from South Australia was used to compare annual nutrient densities in 2008 and 2009.

Results
Table 2 provides a summary of the nutrient densities for nine nutrients reported here, giving the mean and standard deviation for each region, aggregated data for states and then for the complete data set.

Table 2. Mean and standard deviations of macronutrients (P, K,S, Mg, Ca) and micronutrients (Mn, B, Cu, Zn) for wheat samples from the 2008 and 2009 NVT sites for Yitpi and Gladius. All values are for dry grain (0% moisture content).

State & Region	Sites/Years	P mg/kg	K mg/kg	S mg/kg	Ca mg/kg	Mg mg/kg	Mn mg/kg	B mg/kg	Cu mg/kg	Zn mg/kg
NSW South East	4	3613 ±202	5075 ±217	1963 ±70	466 ±29	1261 ±37	57.6 ±3.6	1.6 ±0.3	3.9 ±0.4	23.0 ±2.4
NSW South West	4	2678 ±202	4238 ±217	1709 ±70	408 ±29	1173 ±37	54.4 ±3.6	1.7 ±0.3	4.1 ±0.4	23.5 ±2.4
NSW Mean		3145 ±168	4656 ±161	1836 ±55	437 ±22	1217 ±30	56.0 ±3.3	1.7 ±0.3	4.0 ±0.3	23.3 ±1.8
SA Lower EP	6	3075 ±165	4533 ±177	1492 ±57	343 ±24	1223 ±30	25.3 ±3.0	2.3 ±0.3	4.4 ±0.3	18.7 ±2.0
SA Mid North	7	3900 ±153	4686 ±164	1803 ±53	460 ±22	1359 ±28	51.1 ±2.7	1.3 ±0.3	5.6 ±0.3	25.4 ±1.8
SA Murray Mallee	9	3467 ±135	4533 ±145	1789 ±47	424 ±19	1349 ±25	38.9 ±2.4	1.9 ±0.2	5.2 ±0.2	19.2 ±1.6
SA South East	5	3620 ±181	4870 ±194	1780 ±63	488 ±26	1295 ±33	26.8 ±3.2	1.5 ±0.3	3.5 ±0.3	24.5 ±2.2
SA Upper EP	12	3117 ±117	4758 ±125	1778 ±41	419 ±17	1237 ±22	49.3 ±2.1	2.4 ±0.2	4.9 ±0.2	26.0 ±1.4
SA Yorke Penn.	6	3083 ±165	4433 ±177	1650 ±57	411 ±24	1202 ±30	41.8 ±3.0	1.8 ±0.3	5.6 ±0.3	22.2 ±2.0
SA Mean		3354 ±71	4641 ±68	1729 ±23	423 ±9	1278 ±13	40.8 ±1.4	1.9 ±0.1	4.9 ±0.1	22.9 ±0.8
Vic Mallee	8	3088 ±143	4256 ±154	1662 ±50	396 ±21	1291 ±26	36.8 ±2.6	3.4 ±0.2	5.1 ±0.3	18.9 ±1.7
Vic North Central	2	2900 ±286	4250 ±307	1793 ±99	348 ±41	1293 ±53	55.8 ±5.1	1.7 ±0.5	4.7 ±0.5	25.5 ±3.4
Vic North East	2	2950 ±286	4275 ±307	1933 ±99	380 ±41	1230 ±53	53.3 ±5.1	1.4 ±0.5	5.1 ±0.5	28.8 ±3.4
Vic Wimmera	5	4110 ±181	5040 ±194	1733 ±63	470 ±26	1424 ±33	49.2 ±3.2	4.6 ±0.3	4.8 ±0.3	27.3 ±2.2
Vic Mean		3350 ±115	4488 ±111	1730 ±38	410 ±15	1323 ±20	44.6 ±2.2	3.3 ±0.2	4.9 ±0.2	23.3 ±1.3
Mean		3329 ±671	4606 ±645	1742 ±220	421 ±89	1282 ±122	43.5 ±13.8	2.2 ±1.3	4.8 ±1.2	23.0 ±7.3

The values reported show considerable regional differences, with P, K and S showing coefficients of variation (CV) of 20%, 14% and 13% respectively although within each region the CV’s for P, K and S were usually less than 5%. There were some larger CV’s though, in particular for P in southern New South Wales and P, K and S CV’s for the North East and North Central regions of Victoria. These values were high although they were derived from the smallest subsets within the data. The variation in grain S content is smaller and more consistent than those values for the other macronutrients.

Table 3 gives the P value of the F statistic from the one-way analyses of variance for each of the main factors tested. This shows that nutrient concentration for the macronutrients did not differ among the three sets of state means for P, K, S and Ca, although there were significant regional differences for these nutrients. P values differed between cultivars, but grain S contents were not different. Grain micronutrient concentration did differ between states and regions, except for state level Zn. The analysed data set does have a strong weighting to data from South Australia so the “national” mean values presented in Table 3 are not indicative of the values across each agroecological zone. For all the nutrients tested, regional values differed significantly (Table 3), suggesting that there is no universal value that can be used for a nutrient budget at regional or sub-regional (i.e. farm) level, rather that values should be based on data from that region. Cultivar and annual differences were not significant for P, S, Mg or Na so that a single regional value would seem appropriate for these nutrients. If the level of interest is in a national scale, for example to assess the national nutrient in-flows and out-flows, then the values in Table 2 would appear to be appropriate as there is no significant difference between the values at the state levels for the macronutrients.
Table 3. P values for the F test in one-way analyses of variance for states, regions or cultivars from the NVT data set analyzed.

	P	K	S	Ca	Mg	Mn	B	Cu	Zn
States	0.51	0.48	0.19	0.60	0.01	0.00	0.00	0.01	0.95
Regions	0.00	0.01	0.00	0.00	0.00	0.00	0.00	0.00	0.00
Cultivars	0.02	0.00	0.24	0.00	0.09	0.57	0.66	0.57	0.00
Year (SA only)	0.86	0.00	0.12	0.00	0.00	0.01	0.00	0.04	0.05

Figure 1 gives a distribution of grain P contents for the data set analysed. P concentration ranged from under 2000 mg/kg to over 4000 mg/kg, with a mean of 3329 mg/kg which is around 400 mg/kg, 15% more than the value reported in the National Land and Water Resources Audit (2001). There was no relationship between grain yield and P content, and soil pH or soil P test value (Colwell P) appeared to have no effect on grain P. P removal, the product of grain yield and grain P content was not significantly correlated to soil P test value or soil pH (data not shown). All the experiments received at least 10 kg P/ha from fertilizers at sowing, but despite this grain P was quite variable within this data set. Similar results have been reported by from the northern grains zone for sorghum (D Lester, pers. comm.), and no driving variable could be proposed for large differences noted.


Figure 1 Wheat grain P content cumulative distribution function from the NVT 2008 and 2009 dataset analysed.

In all regions, mean grain zinc levels (Table 2) were less than the 33 mg/kg target suggested as part of the Harvest Plus Zinc bio-fortification program (Cakmak 2010), and in the Lower Eyre Peninsula, and the Victorian and South Australia Mallee mean levels are around half the Harvest Plus target.

Conclusions:
Macro and micro nutrient densities show significant regional variation, and may also differ in some cases among cultivars. To complete reliable nutrient removals for the macro nutrients, regionally or paddock determined grain P, S and K values should be used. For aggregation of data at larger scales – such as for state level nutrient removals, nominal values could be used although they are also likely to have temporal differences may be related to grain yield.

Grain micronutrient densities should relate to soil conditions, and while the generalities of these relationships with soil pH can be seen, there are sufficient deviations to suggest that – like the macro nutrients, particular values could be appropriate for local regions. For an aggregate measure in undertaking farm nutrient audits, it would appear that 3300 mg/kg is a reasonable estimate, but if state or regional nutrient balances are required, then the values need to be more closely.

AcknowledgmentsThe author would like to thank the trial managers Rob Wheeler (SARDI), Frank McRae (Industry and Investment, New South Wales), Harpreet Gill (Agrisearch P/L) and Angela Clough (Victorian DPI) for supplying grain samples. Also NVT CEO and trials manager Alan Bedggood provided soil test and location information from the NVT database as well as support for the collection of this information. The study was funded by the International Plant Nutrition Institute.

References
Cakmak I (2010) Biofortification of cereals with zinc and iron through fertilization strategy. 19th World Congress of Soil Science, Soil Solutions for a Changing World, 1 – 6 August 2010, Brisbane, Australia. http://www.iuss.org/19th%20WCSS/Symposium/pdf/1165.pdf Accessed 27 April, 2012.
National Land and Water Resources Audit (2001).Regional Farming Systems and Soil Nutrient Status Final Report September 2001, Land and Water Australia, Canberra.
Reuter DJ, Edwards DG and Wilhelm NS (1997). Temperate and Tropical Crops. In Plant Analysis: An Interpretation Manual. Eds DJ Reuter, JB Robinson. p 83-253. CSIRO Publishing, Melbourne.
Schultz J and French RJ (1978). Mineral content of herbage and grain of Halberd wheat in South Australia. Australian Journal of Experimental Agriculture and Animal Husbandry 16, 887-892.