Fertiliser management for higher productivity of established lucerne pasture

Abstract

Lucerne (Medicago sativa L.), a deep-rooted perennial legume forage, is a significant part of pastures, and is also grown in rotation with wheat to improve the yield and protein content of grains. Adequate supply of key nutrients is one of the important factors in higher productivity and sustainability of lucerne pastures. However, farmers mostly overlook the importance of nutrient management of established lucerne pastures, resulting in poor persistence, low productivity and weed invasion. The aim of this on-farm field study was to investigate the effects of the combined use of P, K, S and Zn and different sources of Zn on herbage yield, crude protein, nutrient composition, nitrogen fixation, weed invasion and persistence of an established (2 years old) stand of ‘Aurora’ lucerne pasture. A randomised block design was used to test ten fertiliser treatments. The combined use of P, K, S and Zn significantly (P<0.05) increased herbage yield (139% increase) over 2 years, nitrogen fixation, leaf-to-stem ratio, crude protein, nutrient composition (P, K, Ca, S, Zn concentrations), persistence and groundcover, and also led to significant weed suppression (weed biomass decreased from 53 to 69%). Of the three forms of zinc tested, zinc sulphate and zinc-coated single superphosphate proved superior to zinc oxide. The results suggest that nutrient management is essential for improved persistence, reduced weed invasion and higher sustainable productivity of established lucerne pastures, which is important for livestock production systems and overall profitability of farming systems.

Keywords:

• lucerne
• herbage yield
• nitrogen fixation
• nutrient composition
• persistence
• weeds invasion

Introduction

Lucerne (Medicago sativa L.), a deep-rooted perennial, constitutes a significant part of pastures and is high-quality forage in Australia and many other countries. It is also an integral part of mixed farming systems, being mostly grown in rotation with wheat to improve the yield and protein content of grains by virtue of its high nitrogen-fixing capacity (Holford & Crocker 1997; Grewal 2001; Grewal & Williams 2003; Mullen et al. 2006; Pietsch et al. 2007). Furthermore, in organic farming, lucerne has a pivotal role in contributing to a soil's nitrogen balance (as the legume in legume–rhizobia symbiosis) (Bruulsema & Christie 1987; Pietsch et al. 2007). In addition, leys/rotations based on lucerne also have significant beneficial effects on soil structure, soil organic matter content and soil nitrogen content (Holford 1981; Gault et al. 1995; Holford & Crocker 1997; Holford et al. 1998; Peoples et al. 1998; Pietsch et al. 2007). Besides fixing atmospheric nitrogen (Bruulsema and Christie 1987; Pietsch et al. 2007), lucerne also benefits subsequent crops by reducing diseases and weed invasion, and improving water infiltration (Meek et al. 1990).

Many sustainable farming systems rely on lucerne's nitrogen-fixing ability (Peoples et al. 1995 ,1998). Lucerne-based pastures represent a more reliable means to improve soil fertility for subsequent crops than annual pastures, fixing 22–25 kg N for every tonne of lucerne's dry matter (DM) production irrespective of the environment in which it was grown (Peoples et al. 1998).

The recent escalation in the costs of nitrogenous fertilisers and a growing emphasis on eco-friendly farming have increased interest in the use of biological nitrogen fixation to sustain agricultural production systems. Whilst working with lucerne growers in northern NSW (Australia), the author observed that most lucerne growers overlooked the importance of adequate nutrition for established lucerne leys, and lucerne nutrient levels were mostly sub-optimal. This was having serious effects not only on overall productivity of lucerne-based cropping systems, but also on livestock nutrition and productivity and therefore on overall farming profitability. Plants with sub-optimal levels of nutrients are generally more vulnerable to environmental stresses and are more prone to diseases, leading to reduced lucerne persistence (ability of the plants to maintain an adequate population in terms of plant density over a period of time) and productivity. For example, Zn-deficient lucerne was observed to be more prone to Phytophthora root rot and common leaf spot disease (Grewal 2001), and water stress had a greater effect on herbage yield in lucerne plants with low-Zn status (Grewal and Williams 2000). Low-K status also had a significant impact on the disease tolerance and persistence of lucerne (Grewal & Williams 2003). Berg et al. (2007) reported a significant improvement in lucerne herbage yield with the addition of P and K, and a marked improvement in lucerne persistence with K addition.

Lucerne persistence and productivity were observed to be progressively declining after about 2 years following establishment. To test the hypothesis that adequate nutrition plays an important role in persistence, nitrogen fixation, weed suppression and overall productivity of established lucerne leys, an on-farm field experiment was conducted. The research involved the combined use of P, K, S and Zn and different sources of Zn on lucerne pasture that had been established 2 years previously.

Materials and methods

Experiment set up and treatments

The study site was a 2 year old, established stand of ‘Aurora’ lucerne situated about 36 km from Tamworth (near Duri) in northern NSW, Australia. This paddock was selected after an initial survey of ten lucerne paddocks for which growers reported comparatively poor growth, reduced persistence and consequent decreased overall productivity and sustainability of their lucerne-based cropping systems. Plant tissue and soil analysis confirmed the sub-optimal nutrient status of the lucerne crop and soil, respectively. Specifically, the plants had 3.5% N, 0.19% P, 1.9% K, 0.13% S, 0.17% Ca, and 16 mg Zn/kg. The soil was a brown clay with a pH of 7.7 (1:5 soil to CaCl₂ ratio), and was low in available P (Colwell P 7 mg/kg), S (SO₄ sulphur 3 mg/kg), K (ammonium acetate extracted K 85.8 mg/kg) and Zn (DTPA-extractable Zn 0.2 mg/kg).

A randomised block design was used, consisting of ten nutrient treatments (Table 1) replicated four times, with a plot size of 10 m × 5 m. The fertilisers were uniformly broadcasted as per treatment after lucerne cutting on 20 July 1999.

Table 1  Fertiliser treatments

T1	Control (no nutrients applied)
T2	20 kg P/ha (as triple superphosphate)
T3	60 kg K/ha (as muriate of potash, KCl)
T4	4 kg Zn/ha (as zinc oxide)
T5	4 kg Zn/ha (as zinc sulphate)
T6	20 kg P/ha + 20kgS/ha (as single superphosphate)
T7	20 kg P/ha + 20 kg S/ha + 4 kg Zn/ha (P and S applied as single superphosphate, and Zn as zinc oxide)
T8	20 kg P/ha + 20 kg S/ha + 4 kg Zn/ha (as Zn-coated single superphosphate)
T9	20 kg P/ha + 20 kg S/ha + 4 kg Zn/ha (as single superphosphate + zinc sulphate)
T10	20 kg P/ha + 20 kg S/ha + 60 kg K/ha + 4 kg Zn/ha (as single superphosphate + muriate of potash + zinc sulphate)

Observations and measurements

Visual observations to differentiate the treatments were regularly recorded over the course of the experiment. Plant characteristics that were noted included stunted growth, leaves of a pale yellow colour, necrotic spots and interveinal chlorosis.

Herbage (above ground biomass) was harvested ten times over a period of 2 years. Harvesting dates were 21 September 1999, 2 November 1999, 7 December 1999, 8 February 2000, 27 September 2000, 8 November 2000, 20 December 2000, 4 February 2001, 3 June 2001 and 22 September 2001. At each harvest, herbage was dried in a dehydrator and then DM yields were determined.

Fifty shoots were randomly cut from each plot before the 2nd, 4th and 8th harvests. Leaves of these shoots were separated from stems, dried in an oven at 65°C for 48 h and weighed to determine the leaf-to-stem ratio. The total numbers of leaflets and the number of leaflets that had dropped were also recorded from the 50 stems and used to calculated leaf drop (as a percentage of total leaves).

To determine the extent of weed invasion and the DM of weeds in the lucerne stand, weeds were collected from five randomly selected areas (each 1 m²) in each plot before the 1st, 4th, 6th, 8th and 10th harvests. The weed biomass was then determined (oven drying at 65°C for 48 h).

Each plot was scored for groundcover percentage based on groundcover of five random areas (each 1 m²) before the 1st, 4th, 6th, 8th and 10th harvests. Persistence was determined by calculating the average number of crowns in five randomly selected areas (each 1 m²) in each plot after harvests.

Lucerne plant samples taken from each plot at the 2nd harvest were also analysed by inductively coupled plasma spectrometry to determine nutrient concentration (Zarcinas et al. 1987). Nitrogen content was also determined to calculate the crude protein content.

Since lucerne fixes 22–25 kg N for every tonne of DM production irrespective of its environment (Peoples et al. 1998), nitrogen fixation in the current study was calculated by multiplying the lucerne dry weight (t/ha) of each cutting by 22.

Statistical analysis

All data, including herbage yield, leaf-to-stem ratio, leaf drop, crude protein, nitrogen fixation, ground cover, number of crowns/m² (plant population) and weed biomass, were subjected to an analysis of variance (ANOVA). Least significant difference (LSD_0.05) was used to determine significant differences between treatment means when the F values of the ANOVA indicated significance.

Results

Plant appearance and herbage yield

Visually, there were marked differences in growth between lucerne plants that had received combined P, K, S and Zn and control plots and plots receiving P or K or Zn. These differences were normally more marked after rains and persisted over 2 years. Plants receiving combined nutrients were more vigorous and succulent. In contrast, lucerne plants in the control plots had stunted growth and pale green leaves, and some of the leaflets showed necrotic spots, interveinal chlorosis and marked leaf drop. Also, lucerne was also invaded by low-quality weeds in the control plots and the plots receiving only P or K or Zn (see section on weed biomass for details).

Herbage yield was significantly higher in plots receiving P or K or Zn only for the 1st and 2nd harvests (Table 2). Total herbage yield (22.49 t/ha) of ten cuttings with the combined use of P and S was significantly greater than that of the control plots (12.67 t/ha) or the plots receiving P only (13.09 t/ha). There was 107–125% increase in herbage yield when Zn was also applied (regardless of form) along with P and S. The highest total herbage yield (30.33 t/ha) of all ten cuttings was achieved when K and Zn were also applied along with P and S. Among the three zinc forms used in conjunction with P and S, zinc sulphate and zinc-coated single superphosphate were comparatively superior and resulted in herbage yields of 28.46 and 27.43 t/ha, respectively, compared with a 26.19 t/ha herbage yield for zinc oxide application.

Table 2  Effects of different nutrient treatments on herbage yield (shoot DM) of lucerne during different cuttings.

	Herbage yield (t/ha)
Treatment	Cut 1	Cut 2	Cut 3	Cut 4	Cut 5	Cut 6	Cut 7	Cut 8	Cut 9	Cut 10	Total
T1	0.35	0.79	0.60	0.72	1.15	1.98	1.92	2.08	1.63	1.45	12.67
T2	0.55	0.94	0.68	0.73	1.14	2.00	1.90	2.06	1.62	1.47	13.09
T3	0.59	0.97	0.73	0.75	1.17	2.08	1.88	2.04	1.60	1.49	13.30
T4	0.54	0.97	0.64	0.72	1.18	2.06	1.88	2.03	1.70	1.48	13.20
T5	0.67	1.29	0.76	0.74	1.20	2.02	1.90	2.04	1.66	1.50	13.78
T6	0.66	2.09	1.53	0.97	2.13	3.98	3.65	3.10	2.37	2.01	22.49
T7	0.77	2.44	1.91	1.19	2.10	4.33	4.00	3.59	2.94	2.92	26.19
T8	0.82	2.48	2.17	1.24	2.21	4.50	4.17	3.84	3.03	2.97	27.43
T9	0.90	2.60	2.21	1.31	2.33	4.60	4.31	4.01	3.11	3.08	28.46
T10	1.17	2.69	2.40	1.41	2.44	4.77	4.42	4.29	3.49	3.25	30.33
LSD 0.05	0.11	0.14	0.19	0.08	0.21	0.27	0.23	0.20	0.15	0.12	1.20

Plant population (persistence)

There was no significant effect of nutrient treatment on lucerne plant population (number of crowns/m²) in the 1st harvest (Table 3). However, plant population declined progressively in subsequent harvests for all treatments, but the rate of decline was greater in the control plots and in plots receiving P or K or Zn compared with plots receiving combined P, K, S and Zn. Applying P or K or Zn alone resulted in a significantly larger plant population at the 4th and 6th harvests compared with applying no fertiliser (control); however, the effects of individual nutrient application were not apparent in the 8th and 10th harvests. Applying P and S together resulted in a significantly larger plant population that persisted to the final harvest compared with applying no fertiliser or P or K or Zn alone. At 4th, 6th, 8th and 10th harvests, the combined use of P, K, S and Zn resulted in a significantly greater plant population compared with the control or with individual application of P or K or Zn. Omitting P, K, S and Zn resulted in a 72.8% decline in plant population at the 10th harvest compared with that of the first harvest. However, the plant population decreased only 24.4% with the combined use of P, K, S and Zn: there were 23.3 crowns/m² at the 10th harvest in plots receiving P, K, S and Zn compared with only 8.5 crowns/m² in the control plots.

Table 3  Effects of different nutrient treatments on plant population reflecting longevity (persistence) of lucerne.

	Plant population (number of crowns/m^2)
Treatment	Harvest 1	Harvest 4	Harvest 6	Harvest 8	Harvest 10
T1	31.2	20.5	15.6	10.4	8.5
T2	30.1	23.3	18.3	12.3	9.4
T3	29.8	23.6	19.2	12.0	10.3
T4	28.5	24.0	19.0	12.4	9.8
T5	29.7	24.5	18.7	12.6	10.0
T6	28.6	26.3	24.6	20.9	15.3
T7	30.5	27.8	25.5	23.8	18.7
T8	31.0	28.2	26.3	24.5	19.4
T9	30.7	28.5	26.0	24.9	19.2
T10	30.8	29.6	27.5	26.4	23.3
LSD 0.05	NS	2.6	2.9	3.1	2.1

Groundcover

Applying P or K or Zn alone significantly (P<0.05) improved the groundcover in the 1st and 4th harvests compared with the control (Table 4). The highest groundcovers at 2nd, 4th, 6th, 8th and 10th harvests were achieved with the combined use of P, K, S and Zn. Groundcover declined markedly at the 4th, 6th, 8th and 10th harvests and the magnitude of the decrease was highest in control plots. Whilst groundcover dropped from 65% at the first harvest to 35% at the 10th harvest in the control plots, the corresponding decline in plots receiving combined P, K, S and Zn was only from 94% to 88%. All the three Zn sources were equally effective in maintaining groundcover.

Table 4  Effect of different nutrient treatments on groundcover of lucerne before different harvestings.

	Groundcover (%)
Treatment	Harvest 1	Harvest 4	Harvest 6	Harvest 8	Harvest 10
T1	65	49	44	41	35
T2	74	56	46	44	37
T3	73	55	45	45	35
T4	72	57	44	42	38
T5	77	60	49	46	40
T6	80	76	62	58	56
T7	84	87	89	79	78
T8	86	88	90	82	79
T9	87	82	86	81	81
T10	94	92	96	90	88
LSD 0.05	7	5	8	9	6

Leaf drop and leaf-to-stem ratio

There was significantly (P<0.05) greater leaf drop and markedly lower leaf-to-stem ratios in the control plots compared with the plots receiving combined P, K, S and Zn (Tables 5 and 6). Applying P or K or Zn alone had a significant beneficial effect on reducing leaf drop and improving the leaf-to-stem ratio for the 2nd and 4th harvests. Leaf drop was lowest and leaf-to-stem highest in plants receiving a combined application P, K, S and Zn. Leaf drop for the 2nd, 4th and 8th harvests was 40, 36.3 and 43.7, respectively in the control plots, while values for the plots receiving combined P, K, S and Zn were 10.3, 9.0 and 14.3%. Leaf-to-stem ratios of 1.01, 0.98 and 0.92 were recorded for the 2nd, 4th and 8th harvests, respectively, for plots receiving combined P, K, S and Zn; the respective values or the control plots were 0.71, 0.67 and 0.64. Among the zinc sources, zinc sulphate and zinc-coated single superphosphate were slightly superior to zinc oxide in reducing leaf drop and improving leaf-to-stem ratio.

Table 5  Effects of different nutrient treatments on leaf drop of lucerne at different harvestings.

	Leaf drop (%)
Treatment	Harvest 2	Harvest 4	Harvest 8
T1	37.3	40.0	43.7
T2	33.7	37.3	42.0
T3	33.2	31.7	41.6
T4	32.8	32.3	40.6
T5	31.7	34.7	40.0
T6	18.3	19.0	20.4
T7	16.2	16.7	19.3
T8	15.6	15.0	16.7
T9	15.8	14.5	16.2
T10	9.0	10.3	14.3
LSD 0.05	3.3	2.5	3.7

Table 6  Effects of different nutrient treatments on leaf-to-stem ratio of lucerne at different harvestings.

	Leaf-to-stem ratio
Treatment	Harvest 2	Harvest 4	Harvest 8
T1	0.71	0.67	0.64
T2	0.83	0.81	0.71
T3	0.84	0.80	0.70
T4	0.81	0.79	0.69
T5	0.84	0.83	0.72
T6	0.90	0.88	0.76
T7	0.94	0.93	0.84
T8	0.96	0.95	0.85
T9	0.97	0.95	0.86
T10	1.01	0.98	0.92
LSD 0.05	0.05	0.08	0.10

Crude protein (CP) and composition of other nutrients

Adding nutrients had a significant effect on the crude protein (CP) and nutrient composition of lucerne (Table 7). Compared with the control plots, there was a slight increase in CP content with the application of P or K alone and a significant increase with the application of Zn (the effect of zinc sulphate was more pronounced than zinc oxide). Applying S along with P resulted in significantly higher CP compared with application of P alone. There were non-significant differences in CP between the plots receiving a combination of P and S and those receiving P, S and Zn or P, K, S and Zn.

Table 7  Effects of different nutrient treatments on crude protein (CP) content and nutrient concentration of established lucerne at second cutting.

Treatment	CP (%)	P (mg/kg)	K (mg/kg)	S (mg/kg)	Ca (mg/kg)	Zn (mg/kg)	Cu (mg/kg)	Fe (mg/kg)	Mn (mg/kg)
T1	20.1	1890	19100	1280	1655	15.9	14.8	45.1	39.0
T2	21.9	2025	18900	1295	1701	15.7	13.5	46.3	40.1
T3	21.9	1905	19900	1305	1625	16.1	15.7	46.4	40.6
T4	22.5	1785	19100	1330	1642	21.6	14.7	44.7	39.6
T5	23.8	1870	19500	1410	1670	22.6	14.3	46.8	39.5
T6	25.0	2085	19600	1526	1732	16.8	16.3	47.3	38.8
T7	25.3	2050	19500	1630	1720	23.0	16.6	47.0	39.2
T8	25.6	2000	19352	1590	1765	22.5	16.3	47.5	39.4
T9	25.9	2110	19600	1660	1760	23.0	16.5	47.7	38.0
T10	25.9	2105	19735	1630	1745	23.2	16.0	46.9	39.9
LSD0.05	1.9	140	598	164	108	0.7	1.2	NS	NS

There was a significant (P<0.05) increase in P, K and Zn concentrations of lucerne plants when one of these nutrients was applied individually or in combination with others compared with no application. There were non-significant differences in P and K concentrations in plants receiving P and S or P, S and Zn or P, K, S and Zn. However, lucerne plants that received S with P had a significantly higher S concentration compared with control plants or plants that received P or K alone. Applying Zn as zinc sulphate or P, S and Zn or P, K, S and Zn also resulted in a significant increase in S concentration compared with the control and other treatments. The zinc concentration in lucerne plants in the control plots was only 15.9 mg/kg DM; this increased to 21.6 and 22.6 mg/kg DM when Zn alone was applied as zinc oxide and zinc sulphate, respectively. The Zn concentration was significantly greater in plants receiving Zn alone or in combination with other nutrients. Zinc sulphate proved to be superior over zinc oxide in increasing the Zn concentration of lucerne plants. The Cu concentration of lucerne plants was significantly less with application of P alone compared with the application of combined P, K, S and Zn. However, the effects of nutrient treatments were not significant on Fe and Mn concentration of the plants.

Nitrogen fixation by lucerne

Nitrogen fixation of lucerne was significantly (P<0.05) influenced by the application of nutrients (Fig. 1). Only 279 kgN/ha was fixed over 2 years in the control plots compared with 667 kg N/ha (a 140% increase) in plots receiving combined P, K, S and Zn. Applying P or K or Zn alone resulted in only a 3–9% increase in nitrogen fixation. Combined use of P and S resulted in the fixation of 495 kg N/ha, 78% more than the amount fixed on the control plots. There was 107–125% increase in nitrogen fixation when Zn was also applied in different forms in conjunction with P and S. Among the three zinc forms used with P and S, zinc sulphate and zinc-coated single superphosphate resulted in the fixation of 626 and 603 kgN/ha, respectively, which was significantly higher than 576kgN/ha associated with zinc oxide.

Fig. 1  Effects of different nutrient treatments on total nitrogen fixed (kg/ha) by lucerne over 2 years. The effects of nutrient treatments were significant and LSD_0.05 value was 24.

Weeds biomass

The lucerne plots were invaded by low-quality weeds, mainly wild turnip (Brassica rapa var. rapa), wild mustard (Sisymbrium orientale), saffron thistle (Carthamus lanatus), bathurst burr (Xanthium spinosum), capeweed (Arctotheca calendula), barley grass (Hordeum leporinum) and silver grass (Vulpia bromoides), and mainly in the control plots and plots receiving only P, K or Zn alone. The weed biomass (i.e. dry matter) was significantly greater in control plots and plots receiving only P or K or Zn compared with those receiving combined P, K, S and Zn at 4th, 6th, 8th and 10th harvests (Table 8). In the 4th to 10th harvests of lucerne, there was 53–69% less weed biomass with combined P, K, S and Zn compared to the no-fertiliser plots (controls). Applying P, S and Zn (omitting only K) resulted in a 35–56% reduction in weed dry weight during different cuttings compared with the control treatment. Similarly, applying only P and S (omitting both K and Zn) resulted in only 21–27% reduction in weed dry weight during different cuttings. The total dry weight of weeds with combined P, K, S and Zn (Zn applied as zinc sulphate) was only 0.73 t/ha, which was significantly lower that of the control plots (1.86 t/ha) and plots receiving P or K or Zn (1.69–1.74 t/ha) or a combination of P and S (1.42 t/ha). There was no difference in weed dry weight between the three forms of Zn.

Table 8  Effect of different nutrient treatments on DM of weeds in lucerne at different harvestings.

	DM of weeds (t/ha)
Treatment	Harvest 4	Harvest 6	Harvest 8	Harvest 10	Total
T1	0.34	0.45	0.52	0.55	1.86
T2	0.31	0.4	0.49	0.54	1.74
T3	0.32	0.41	0.47	0.52	1.72
T4	0.3	0.39	0.48	0.52	1.69
T5	0.31	0.38	0.47	0.55	1.71
T6	0.27	0.33	0.4	0.42	1.42
T7	0.21	0.26	0.28	0.26	1.02
T8	0.21	0.25	0.29	0.28	1.03
T9	0.18	0.27	0.27	0.24	0.96
T10	0.16	0.19	0.21	0.17	0.73
LSD 0.05	0.06	0.07	0.1	0.08	0.22

Discussion

This study revealed that, compared with no fertiliser application, the combined use of P, K, S and Zn in established lucerne pasture led not only to significantly improved herbage yield, crude protein, nutrient composition, nitrogen fixation and crop persistence but also contributed to a significant reduction in weed invasion. Applying P or K or Zn alone was unable to compensate the losses to herbage yield, persistence and nitrogen fixation. These results suggest that application of combined P, K, S and Zn plays a pivotal role in realising optimal productivity, sustainability and environmental protection of established lucerne pastures. Omitting to provide P, K, S and Zn to an established lucerne crop that has sub-optimal nutrient status will have severe penalties not only on herbage yield and on lucerne's contribution to nitrogen fixation, but will also affect the crop's ability to suppress weeds. This may also have strong implications for crops grown in rotation with lucerne.

The results also demonstrated that Zn applied as zinc sulphate or zinc-coated single superphosphate was more effective than if applied as zinc oxide. The sulphur content of zinc sulphate and the granular form of zinc-coated single superphosphate appear to contribute to their superiority over zinc oxide. Also, Zn may be less available in zinc oxide because the powder presents a greater surface area of contact with the soil and therefore a greater sequestering of Zn in the soil colloidal complex.

The major impact of using combined P, K, S and Zn in the current study was improved lucerne persistence, which was reflected in a greater plant density, even up to the 10th harvest. There was a positive linear relationship between the number of lucerne crowns/m² and herbage yield, with R ² values of 0.86 and 0.96, respectively, for the 4th (Fig. 2(a)) and 10th harvest (Fig. 2(b)), indicating the importance of persistence in sustaining lucerne productivity. Herbage yield of lucerne has been described as the product of number of plants per unit area, number of shoots per plant and mass per shoot (Volenec et al. 1987; Berg et al. 2007). There was a significantly higher number of crowns per unit area (Table 3) and shoots per unit area (data not given) with combined P, K, S and Zn in the current study. Berg et al. (2005) reported higher herbage yields for lucerne with the combined application of P and K compared to applying each element alone. Adequate K nutrition has been reported to be positively associated with improved lucerne persistence (Burmester et al. 1991; Simons et al. 1995; Grewal and Williams 2002; Berg et al. 2007).

Fig. 2  Relationship between lucerne plant population (number of lucerne crowns/m²) and herbage yield for (a) the 4th harvest and (b) the 10th harvest.

The inverse linear relationship (R ²=0.99) between lucerne herbage yield and weed biomass found in this study (Fig. 3) suggests that maintaining a high lucerne biomass is important to avoid the invasion of weeds. The inverse linear relationship between lucerne population size (persistence) and weed biomass (R ²=0.98) (Fig. 4) indicates that the greater suppression of weeds was due to the higher number of lucerne plants associated with the use of combined P, K, S and Zn. Lucerne plants were more vigorous and more successfully competed with weeds, leading to greater suppression. Increased invasion of weeds in control plots and in plots receiving P or K or Zn alone probably contributed to greater utilisation of water and nutrients by weeds, and consequently affected the growth and herbage yield of lucerne.

Fig. 3  Relationship between herbage yield of lucerne and weed biomass (dry weight) for the 10th harvest.

Fig. 4  Relationship between lucerne plant population (number of lucerne crowns/m²) and weed biomass (weeds dry weight) for the 10th harvest.

The major agronomic implications of this study are that growers having low levels of nutrients, particularly P, K, S and Zn, in their established lucerne pastures are more at risk of decreased productivity, nitrogen fixation, sustainability and overall profitability. Besides severe penalties on herbage yield and reduced forage quality, weeds are more likely to invade the pasture and compete for subsoil water and nutrients, and thereby contribute to reduced efficiency of water and nutrient use by lucerne. This will also affect livestock productivity and therefore the overall profitability of farming systems. Furthermore, plants with sub-optimal levels of nutrients are generally more vulnerable to environmental stresses and more prone to diseases, leading to decreased persistence.

Besides the reduction in weed invasion, the improved plant population and greater groundcover associated with combined P, K, S and Zn are expected to reduce runoff and protect against soil erosion. Runoff and soil erosion are related to groundcover (Snyman & VanRansburg 1986; Silburn et al. 1992; Zobisch 1993; McIvor et al. 1995), with groundcover being particularly important in reducing runoff under high rainfall events by reducing the amount of soil disturbed by raindrops and the ability of water to transport soil particles (McIvor et al. 1995).

Another major beneficial effect of combined P, K, S and Zn is on lucerne's contribution to soil nitrogen, which is important for wheat or other cereal crops grown in rotation with lucerne. Any additional nitrogen fixed by nutrient-sufficient lucerne will increase the productivity and protein content of subsequent wheat or other cereal crops. Lucerne plants that had not received P, K, S and Zn had sub-optimal levels of nutrients (mainly N, P, K, S and Zn) and were probably unable to maintain sufficient carbohydrates in their roots to continue vegetative growth. Consequently, these plants suffered markedly in their persistence and nitrogen-fixing ability. Zinc-deficient lucerne plants have significantly fewer nodules (Grewal 2001), and sulphur deficiency had been reported to decrease nitrogen-fixation by lucerne plants (DeBoer & Duke 1982).

The results demonstrate that the combined use of P, K, S and Zn also benefited lucerne by reducing leaf drop and enhancing leaf-to-stem ratio. Improved leaf retention and a high leaf-to-stem ratio are desirable traits associated with improved forage and hay quality in lucerne (Grewal 2001), perhaps due to the fact that proteins are also more concentrated in leaves than in stems (Grewal & Williams 2002 ,2003). Addition of P or K alone had little or no effect on CP content of lucerne in the current study. These results differ from those of Lissbrant et al. (2009), who reported a significant increase in CP with the addition of P, and a reduction in CP with the addition of K. In the current study, the significant increase in CP was observed only with combined P, K and Zn and/or combined P, K, S and Zn. Low levels of P, K, S and Zn in the plants may have limited the synthesis of enzymes responsible for the reduction of inorganic nitrogen into amino acids. Friedrich & Schrader (1978) demonstrated a decrease in extractable nitrate reductase activity preceding a decline in soluble protein, chlorophyll and the enzymes of NH₄ assimilation in maize seedlings grown without S but fertilised with nitrate.

The improved CP and other nutrient (P, K, S, Zn and Cu) concentrations with combined P, K, S and Zn in the present study will improve the nutrition of grazing/feeding cattle, sheep and other animals, as P, K, S and Zn are essential both for animals and plants. Animals fed on low-Zn forages or feed (Zn less than 5 mg/kg DM) may develop signs of severe Zn deficiency (Underwood 1981; White 1993). Marginal Zn deficiency (characterised by sub-optimum growth, reduced fertility and skin disorders) has been reported in cattle and sheep grazing on pastures containing less than 20 mg Zn/kg (White 1993). Pregnant and lactating cows have much greater Zn requirements. Milking cows fed on nutrient-rich pastures (particularly Zn) improved the concentration of Zn in their milk, leading to premium prices in the market. Zinc is an essential nutrient for humans and dairy products are the major source of Zn for humans (Welch 1993).

Conclusions

The results of this study revealed that omitting P, K, S and Zn application to a 2 year old established lucerne ley had significant detrimental effects on herbage yield, persistence, nitrogen fixation, crude protein content and concentrations of other nutrients, along with greater weed invasion. The combined application of P, K, S and Zn resulted in maximum beneficial effect not only on herbage yield but also on nitrogen fixation, groundcover, leaf retention, leaf-to-stem ratio, crude protein and nutrient (P, K, S, Ca, Zn and Cu) concentration and weed suppression. Among the zinc forms tested, zinc sulphate and zinc-coated single superphosphate proved to be equally effective and superior to zinc oxide for improving herbage yield and other attributes of lucerne. The results of this study suggest that the application of P, K, S and Zn to established lucerne leys with low-nutrient status is pivotal in realising the potential beneficial effects of lucerne leys. Omitting these nutrients will result not only in severe penalties on herbage yield, crude protein and nutrient composition, but will also significantly reduce lucerne's contribution to soil nitrogen and increase weed invasion, leading to deterioration in the quality of lucerne which may also have a detrimental effect on succeeding crops.

Acknowledgements

This research was partly supported by Grains Research and Development Corporation of Australia. The author thanks Stan Lee (Farmer), Bill Davidson and John Paul (Technical Officers) and Rex Williams (Lucerne Breeder) for their help with this trial.