Effect of surface applied glycine betaine on herbage production and quality of perennial ryegrass white clover pastures

Transcription

1 CSIRO PUBLISHING Australian Journal of Experimental Agriculture, 2008, 48, Effect of surface applied glycine betaine on herbage production and quality of perennial ryegrass white clover pastures J. M. Lee A,G, K. Elborough B,E, W. D. Catto C, D. J. Donaghy D and J. R. Roche A A DairyNZ, Private Bag 3221, Hamilton 3240, New Zealand. B ViaLactia BioSciences, PO Box , Auckland 1061, New Zealand. C Ballance Agri-Nutrients Limited, Private Bag 12503, Mount Maunganui 3116, New Zealand. D University of Tasmania, PO Box 3253, Burnie, Tas. 7320, Australia. E Present address: HortResearch, 120 Mt Albert Road, Auckland 1025, New Zealand. F Corresponding author. julia.lee@dairynz.co.nz Abstract. Osmoprotectants have been reported to reduce the detrimental effects of various environmental stresses in many different plant species. However, there is little research available concerning pasture species. Two experiments were undertaken with the aim of quantifying the effect of surface applications of exogenous glycine betaine (GB) on herbage production and quality of perennial ryegrass (Lolium perenne L.) white clover (Trifolium repens L.) pastures during periods of moisture stress and cold temperatures over 2 years. Pastures fertilised with GB were compared with unfertilised pastures and pastures fertilised with nitrogen (N). Rates of 0.5, 1.0 and 1.5 kg GB/ha.defoliation were applied in experiment 1 and 5 kg GB/ha.defoliation was applied in experiment 2. Surface applications of GB did not significantly affect herbage production relative to unfertilised pastures; herbage yields averaged and kg DM/ha over 11 months in experiment 1, and 7253 and 7177 kg DM/ha over 6 months during summer and autumn in experiment 2, for the unfertilised control and GB, respectively. During both experiments, herbage quality parameters were not affected by GB application, although the proportion of white clover in the sward between summer and winter during experiment 1 was greater (P < ) in plots treated with GB than in untreated plots. Application of N fertiliser increased (P < 0.001) herbage production, but did not consistently affect herbage quality. The failure of surface applications of exogenous GB to improve the herbage production or quality of perennial ryegrass white clover pastures suggests that it is not an appropriate method to enhance plant tolerance to environmental stress at the concentrations applied in these studies. Additional keywords: water deficit. Introduction Environmental variables, such as low moisture availability, can Rhodes and Hanson 1993; Kishitani et al. 1994; Gorham 1995). have a considerable impact on plant growth and development Glycine betaine is considered a particularly effective (Wilson 1981; McKenzie et al. 2000). As a result, plant species osmoprotectant in times of abiotic stress (Le Rudulier et al. exhibit a variety of responses under conditions of moisture 1984), due to its ability to stabilise the structure of proteins, stress. One adaptation in drought-tolerant plants is osmotic enzymes and membranes, and the fact that it does not inhibit adjustment, where cells raise the osmotic pressure of the enzyme activity (for a review, see Gorham 1995). cytoplasm by accumulating solutes that do not have toxic Furthermore, exogenous applications of GB to leaves or roots effects. This contributes to turgor maintenance of both roots and have been shown to increase the tolerance of several species of shoots (Datta 2002). Some of these solutes may also protect plants, including both natural accumulators and noncellular components from injury caused by dehydration accumulators, to various stresses (Harinasut et al. 1996; Mäkelä (Mickelbart et al. 1999). Osmoprotectants (compatible solutes) et al. 1997; Allard et al. 1998; Lopez et al. 2002). The leaf area of include sugars and sugar alcohols (Thomas et al. 1995), amino tobacco (Nicotiana tabacum L.) plants increased 14% following acids (Talwar and Yanagihara 1999) and tertiary sulfonium and application of 0.1 mol GB/L before water stress (Agboma et al. quaternary ammonium compounds (Rhodes and Hanson 1993). 1997b). In a glasshouse experiment, a foliar application of 0.2 Glycine betaine (N,N,N-trimethylglycine; hereafter GB) is mol GB/L increased the growth rates of both drought-stressed pea one such quaternary ammonium compound that occurs (Pisum sativum L.) plants and those recovering from drought by naturally in a wide variety of plants, animals and 45% and 13%, respectively, compared with untreated plants as microorganisms (Sakamoto and Murata 2000). A positive measured 14 days after GB application (Mäkelä et al. 1997). correlation exists between the accumulation of GB and the Grain yields from maize (Zea mays L.), sorghum (Sorghum acquisition of tolerance to a range of abiotic stresses, such as bicolour L.), soybean (Glycine max L. Merr.) and wheat (Triticum salinity, drought and cold temperatures (Rhodes et al. 1989; aestivum L.) have also been increased by between 4 and 37% CSIRO /EA /08/050687

2 688 Australian Journal of Experimental Agriculture J. M. Lee et al. following applications of 3 6 kg GB/ha (Agboma et al. 1997a; Agboma et al. 1997c; Diaz-Zorita et al. 2001). Herbage production can also be reduced by low temperatures. In late autumn winter, the temperature below which there is minimal grass growth is 8 C. This critical temperature drops to 5 C after mid winter, following vernalisation (Brereton et al. 1985), but growth is very sensitive to changes in temperature between 5 C and 10 C. Allard et al. (1998) improved the survival of wheat plants through treatment with relatively low concentrations of GB before exposure to cold temperatures (6 C/2 C day/night temperatures). While the protection of membrane integrity against freezing has been demonstrated in lucerne (Medicago sativa L.) sprayed with 0.2 mol GB/L (Zhao et al. 1992), little research exists on the response of forages treated with GB to cold stress. Although Naidu et al. (2003) discovered that a foliar application of 1 kg GB/ha during winter increased the DM production of temperate pasture (species unknown) by 24% (average from two sites), there is a lack of information on the effect of GB on perennial ryegrass white clover pastures subjected to environmental stresses. Most studies applied GB as a foliar spray to the leaves of growing plants. However, fertiliser is generally applied to grazed pastures as a prill following defoliation. The use of GB as an osmoprotectant would be minimal if requiring regular foliar applications. The aim of the current study was to therefore quantify the effect of a surface application of GB on herbage production and quality of perennial ryegrass white clover pastures. The effects of GB were compared with unfertilised pastures and those fertilised with N. Materials and methods Experimental site Two experiments were conducted at Scott Farm, Dexcel, Hamilton, New Zealand (37 47 S, E; elevation 40 m) between September 2004 and May 2006.A uniform experimental site recently sown with perennial ryegrass (cv. Bronsyn sown at 18 kg/ha) and two cultivars of white clover (cv. Aran and Kopu, both sown at 3.5 kg/ha) was selected. The pasture was grazed lightly once by dry dairy cows before the experiments began. The soil type of the experimental site was ate Kowhai silt loam (Singleton 1991) or Typic Ochraqualf (Soil Survey Staff 1990). Soil macronutrient concentration was analysed before the first experiment, revealing adequate mineral concentrations (Roberts and Morton 1999) of 27 µg/kg phosphorus (P; Olsen), 38 mg/kg sulfate sulfur (S; potassium phosphate extraction), 159 mg/kg calcium (ammonium acetate extraction), 120 mg/kg potassium (K; ammonium acetate extraction), 27 mg/kg magnesium (ammonium acetate extraction), 4 mg/kg sodium (ammonium acetate extraction). The ph was 6.1 (1:2.1 v/v water slurry). Experimental design and treatments Experiment 1 In the first experiment, treatments were assigned to plots (2 5 m) in a randomised block design, with seven treatments replicated nine times. Each plot was separated from adjacent plots by a 1-m border. Treatments consisted of a control, two rates of N fertiliser (25 and 50 kg N/ha.defoliation in the form of urea; N25 and N50, respectively) and three rates of GB (0.5, 1.0, 1.5 kg/ha.defoliation; GB0.5, GB1.0 and GB1.5, respectively) applied as a prill. This application method was chosen because it is the most practical method of fertiliser application for regularly defoliated pastures. A further treatment consisted of 1.0 kg GB/ha.defoliation (GBP) applied in anhydrous powder form. The nine replicates were randomly assigned to three harvestdate groups of three treatment replicates. Each group was harvested in 1 day, with sequential groups harvested each subsequent week. The first group was harvested on 22 October 2004 (H 1 ), 16 November 2004 (H 2 ), 17 December 2004 (H 3 ), 19 January 2005 (H 4 ), 24 February 2005 (H 5 ), 31 March 2005 (H 6 ), 16 May 2005 (H 7 ), 7 July 2005 (H 8 ) and 31 August 2005 (H 9 ). The second and third harvest-date groups were harvested 7 and 14 days later, respectively. Immediately following each defoliation, treatment plots received the appropriate treatment application. Following the final harvest for experiment 1, all plots were fertilised with 142 kg urea (65 kg N)/ha, 468 kg single superphosphate (49 kg S and 45 kg P)/ha and 800 kg muriate of potash (400 kg K)/ha, to replace nutrients removed. Experiment 2 Treatments in experiment 2 consisted of a control, one rate of N fertiliser (35 kg N/ha.defoliation in the form of urea; N) and one rate of GB (5.0 kg/ha.defoliation; GB), applied as a prill. Treatments were randomly allocated within blocks, such that each treatment occurred at least once within each block and each treatment occurred at least twice within previous treatment. Each treatment was replicated 15 times. The first harvest-date group was harvested on 21 December 2005 (H 1 ), 18 January 2006 (H 2 ), 15 February 2006 (H 3 ), 22 March 2006 (H 4 ), 19 April 2006 (H 5 ) and 27 May 2006 (H 6 ). The second and third harvest-date groups were harvested 7 and 14 days later, respectively. Plots received an initial treatment application immediately following the covariate harvest (November 2005) and again following each of the three subsequent defoliations (H 1 H 3 ; Treatment period). Plots did not receive any further treatment applications from H 4 onwards (water deficit recovery period). Climatic measurements Climatic data, including 24-h maximum and minimum air temperature ( C), 24-h rainfall (mm), soil temperature at 100-mm depth ( C) and radiation (MJ/m 2 ), were recorded daily at 0900 hours at a meteorological station less than 3 km from the experimental site. In both experiments between December and May, soil volumetric water content was measured weekly in 20 locations within the plots using a FieldScout TDR300 soil moisture meter (120-mm length probe; Spectrum Technologies Inc., Plainfield, IL). Herbage measurements Prior to each experiment, all treatment plots were harvested and herbage DM yields determined for use as a covariate (H 0 ). The decision to harvest following H 0 in both experiments was based on the emergence of three full leaves on ryegrass tillers in the control treatment, or before canopy closure in plots fertilised with N whichever occurred first. This was assumed to be the

3 Effect of glycine betaine on herbage production Australian Journal of Experimental Agriculture 689 optimal time for defoliation (Fulkerson and Donaghy 2001). In the 2 months between experiments, plots continued to be harvested based on the same decision rules. Once plots were ready to harvest, two parallel strips (each m) were cut to a residual height of 35 or 40 mm (experiments 1 or 2, respectively) from the centre of each plot using a rotary lawnmower. Sample fresh weight was recorded in the field, then representative subsamples (200 g) were ovendried in duplicate at 95 C to constant weight (~48 h) to estimate DM content. Herbage mass was calculated by multiplying the sample fresh weight by the average DM content. Total herbage accumulation for each experimental plot was calculated from the sum of herbage mass from each individual harvest. Following sample harvest, the remaining plot area was mown to the prescribed residual height for each experiment. In experiment 1, additional herbage samples were harvested from the control, N50 and GB1.0 treatment plots at the end of each 3-month period (November, February, May and August) on the day before harvest, for botanical and herbage quality measurements. During experiment 2, additional herbage samples were harvested from all treatments on the day before H 3 and H 4, for botanical and herbage quality measurements. Representative herbage samples were cut to either 35 or 40 mm (experiments 1 or 2, respectively) using hand shears from the centre of each plot, weighed, blended and subsampled. One subsample (20 g) was dissected into ryegrass leaf and pseudostem or stem, white clover, other grasses, weeds and dead material, then oven-dried at 90 C to constant weight (~24 h) to determine botanical composition. Another subsample was oven-dried at 60 C for 48 h, ground to pass through a 1-mm sieve (Christy Laboratory Mill, Suffolk, UK) and analysed for crude protein (CP), lipid, ash, acid detergent fibre (ADF), neutral detergent fibre (NDF), non-structural carbohydrates (NSC) and organic matter digestibility (OMD) content using near infrared spectroscopy (Corson et al. 1999). In both experiments, 20 whole perennial ryegrass leaves were randomly selected from the final subsample. The length and width of the leaves were measured in milllimetres, and the leaf area calculated as described by Gay (1993) and Kemp (1960). Leaves were then dried at 95 C for 24 h to constant weight. Leaf area index (LAI) was calculated using the following formula (Roche 1995): LAI = total length of leaf average width 1000 yield/ha subsample weight During experiment 2, in addition to the ryegrass samples for LAI, 20 white clover leaves were collected from each plot before the third harvest. Growing points from 20 randomly selected clover plants were identified, and the second emerged leaf from the growing point collected. Leaves were electronically scanned and the area of each leaf determined using Delta-T-Scan (Delta-T Devices Ltd, Burwell, UK). Statistical analysis Individual harvest data from experiment 1 were analysed by ANOVA as a randomised block design using the statistical procedures of GENSTAT 7 (2004), with treatment and harvest-date group as fixed effects. Contrasts of treatments were included to test the effects of rate of GB and N application, mean N v. control, mean GB v. control, and N50 v. GB1.0 v. control. Individual harvest data from experiment 2 were analysed using REML as a mixed model, using the statistical procedures of GENSTAT 7 (2004), with treatment, previous treatment and harvest-date group as fixed effects, and group and plot as random effects. Pre-experimental measurements from both experiments were used as covariates. A probability of P < 0.05 was used to determine statistical significance, unless otherwise stated. Results Experiment 1 Mean daily temperatures and radiation levels, and total monthly evapotranspiration and rainfall are presented in Table 1. A soil moisture content of 20 23% is reported to be the permanent wilting point of perennial ryegrass on the soil type of the experimental area (Singleton 1991). Soil moisture levels remained above 30% until mid January, declining thereafter and remaining below the critical level of 20% from 21 January to 8 May 2005 (Fig. 1). Table 1. Mean (± s.d.) daily maximum, minimum and soil temperature at 100-mm depth, radiation and total monthly evapotranspiration and rainfall data for experiments 1 and 2 Oct. Nov. Dec. Jan. Feb. Mar. Apr. May June July Aug. Experiment 1 Maximum air temperature ( C) 17.8 (1.8) 20.3 (2.0) 19.0 (1.7) 23.6 (2.3) 25.6 (2.1) 24.4 (1.7) 20.7 (2.5) 18.2 (2.3) 14.2 (1.9) 14.1 (1.7) 14.7 (1.1) Minimum air temperature ( C) 9.2 (3.8) 9.4 (3.1) 10.5 (2.9) 12.6 (2.6) 14.4 (3.7) 12.2 (3.1) 7.5 (4.3) 8.4 (4.0) 5.3 (3.5) 5.3 (4.1) 3.0 (2.6) Soil temperature at 100 mm ( C) 14.9 (1.8) 17.9 (1.7) 17.0 (1.4) 20.8 (2.1) 22.1 (1.6) 19.7 (1.6) 15.1 (2.7) 12.4 (2.2) 8.9 (1.9) 8.5 (2.5) 8.5 (2.0) Radiation (MJ/m 2 ) 16.3 (6.1) 21.0 (4.8) 20.8 (7.1) 23.6 (7.1) 19.7 (4.2) 19.4 (2.1) 12.6 (2.5) 8.2 (2.7) 7.1 (2.4) 7.3 (2.4) 11.3 (3.2) Total evapotranspiration (mm) A Total rainfall (mm) Experiment 2 Maximum air temperature ( C) 22.5 (1.5) 23.9 (2.2) 24.4 (1.9) 22.2 (1.9) 20.7 (1.7) 17.1 (2.0) Minimum air temperature ( C) 13.3 (3.1) 13.3 (3.9) 13.0 (3.1) 11.0 (3.7) 11.3 (2.8) 6.6 (3.8) Soil temperature at 100 mm ( C) 20.1 (1.3) 21.5 (2.0) 21.1 (1.6) 17.6 (1.5) 15.6 (1.6) 11.2 (2.6) Radiation (MJ/m 2 ) 23.1 (6.5) 26.7 (4.9) 20.8 (5.2) 16.9 (5.1) 11.5 (3.3) 8.7 (2.9) Total evapotranspiration (mm) A Total rainfall (mm) A Priestly Taylor evapotranspiration.

4 690 Australian Journal of Experimental Agriculture J. M. Lee et al. Soil water content (%v/v) Dec. Jan. Feb. Mar. Apr. May June Fig. 1. Volumetric water content of the soil (% v/v) between December 2004 and May 2005 (solid line; experiment 1) and December 2005 and May 2006 (dashed line; experiment 2). Values below 20% moisture are below the permanent wilting point of ryegrass. smaller reduction in leaf size than both the control and N fertiliser treatments (P < 0.05; reduction of 17%, compared with 20% and 34%, for the mean of GB, N and control treatments, respectively). As a result, compared with unfertilised pastures, pastures treated with GB produced larger leaves for the remainder of the experiment. Leaf size in plots fertilised with N was greater, relative to both the control and GB treatments throughout the year. During autumn, application of GB increased the proportion of white clover in the sward, while the proportion of ryegrass leaf decreased (Table 4). Application with N fertiliser had the opposite effect, with the proportions of ryegrass stem and dead material also increased following N fertilisation. Irrespective of season, there was no difference in herbage quality parameters between unfertilised pastures and those that received GB. Application of 50 kg N/ha tended to increase herbage fibre content and reduce OMD and ME content, but the effect was inconsistent and varied with season. There was no difference between GB1.0 and GBP on any of the measured parameters. From this point onwards, only results for GB1.0 will be referred to. Nor was there any effect of rate of GB application, therefore, comparisons shall be made between the mean of GB, N and control treatments. The effects of N and GB application on individual harvest and total herbage yields during the experiment are presented in Table 2. GB did not significantly affect total herbage production relative to the control (12248 and kg DM/ha for the unfertilised control and mean of the GB treatments, respectively). Application of N fertiliser increased herbage yields throughout the year, except during summer. This resulted in an increase (P < 0.001) in total herbage production of 3994 kg DM/ha (12248 and kg DM/ha, for the unfertilised control and N50, respectively). Increasing the N application rate from 25 to 50 kg N/ha only increased herbage production during spring (H 1 H 2 ). At the end of spring, there was no difference in ryegrass leaf size in pastures treated with GB, relative to the control (Table 3), but both were smaller (P < ) than those fertilised with N. During summer, pastures fertilised with GB experienced a Experiment 2 Mean daily temperatures and radiation levels, and total monthly evapotranspiration and rainfall are presented in Table 1. Soil moisture levels declined below the critical level of 20% for 10 days in mid January, and again for 38 days in late February March (Fig. 1). The effects of GB and N application on individual harvest and total herbage yields during the experiment are presented in Table 5. Glycine betaine did not significantly affect herbage production relative to the control at any point during the experiment, resulting in total herbage yields of 7253 and 7177 kg DM/ha for the unfertilised control and GB, respectively. Application of N fertiliser increased individual herbage yields during both the treatment and recovery periods, resulting in an increase (P < 0.001) in total herbage production of 1012 kg DM/ha (7253 and 8265 kg DM/ha, for the unfertilised control and N, respectively). At the end of summer (H 3 ), there was no difference in the size of white clover leaves in pastures fertilised with GB or N, relative to the control (Table 6). Nor was there any effect of GB on the size of perennial ryegrass leaves relative to the control (averaging Table 2. Individual harvest and total herbage yields (kg DM/ha) from unfertilised plots (Control) and those receiving nitrogen (25 and 50 kg N/ha.defoliation in the form of urea; N25 and N50, respectively) or glycine betaine (0.5, 1.0, 1.5 kg/ha.defoliation; GB0.5, GB1.0 and GB1.5, respectively) during experiment 1 The first harvest-date group was harvested on 22 October 2004 (H 1 ), 16 November 2004 (H 2 ), 17 December 2004 (H 3 ), 19 January 2005 (H 4 ), 24 February 2005 (H 5 ), 31 March 2005 (H 6 ), 16 May 2005 (H 7 ), 7 July 2005 (H 8 ) and 31 August 2005 (H 9 ). The second and third harvest-date groups were harvested 7 and 14 days later, respectively., P < 0.001;, not significant Control N25 N50 GB0.5 GB1.0 GB1.5 s.e.d. P-value N GB N50 v. GB1.0 H H H H H H H H H Total

5 Effect of glycine betaine on herbage production Australian Journal of Experimental Agriculture 691 Table 3. Perennial ryegrass leaf size (mm 2 ) and leaf area index (LAI) of unfertilised plots (Control) and those receiving nitrogen (25 and 50 kg N/ha.defoliation in the form of urea; N25 and N50, respectively) or glycine betaine (0.5, 1.0, 1.5 kg/ha.defoliation; GB0.5, GB1.0 and GB1.5, respectively), as measured at the end of each season during experiment 1, P < 0.05;, P < ;, P < 0.001;, not significant Control N25 N50 GB0.5 GB1.0 GB1.5 s.e.d. P-value N GB N50 v. GB1.0 Leaf size (mm 2 ) Spring Summer Autumn Winter Leaf area index Summer Autumn Winter and 382 mm 2 for H 3 and H 4, respectively), although both were smaller than those fertilised with N (Table 6). The effect of GB and N application on the botanical composition of the sward are presented in Table 6. While GB application did not significantly affect the proportions of ryegrass and white clover in the sward, N fertiliser tended to increase the proportions of ryegrass leaf and stem, while decreasing the proportion of white clover in the sward. There was no effect of GB on herbage quality parameters and N fertiliser had little effect on herbage quality, with slightly increased levels of ash (97, 97 and 100 g/kg DM for control, GB and N treatments, respectively; P < ), CP (267, 266 and 274 g/kg DM for control, GB and N treatments, respectively; P < 0.1) and NDF (318, 327 and 337 g/kg DM for control, GB and N treatments, respectively; P < 0.1) in herbage harvested at H 4. Discussion The soil type of the experimental site was a Te Kowhai silt loam, well suited to dairy farming due to the high potential value for herbage production. This soil type reaches field capacity at 50 55% moisture and its wilting point is 20 23% moisture (Singleton 1991). At wilting point, capillary forces become too great to allow roots to extract water from the soil, placing the plant Table 4. Botanical composition (proportion of above 35 mm herbage mass; DM basis) of herbage from unfertilised plots (Control) and those receiving nitrogen (25 and 50 kg N/ha.defoliation in the form of urea; N25 and N50, respectively) or glycine betaine (0.5, 1.0, 1.5 kg/ha.defoliation; GB0.5, GB1.0 and GB1.5, respectively), as measured at the end of each season during experiment 1, P < 0.05;, P < ;, P < 0.001;, not significant Control N25 N50 GB0.5 GB1.0 GB1.5 s.e.d. N P-value GB N50 v. GB1.0 Spring Summer Autumn Winter Ryegrass leaf Ryegrass stem A White clover Dead material Spring Summer Autumn Winter Spring Summer Autumn Winter Spring Summer Autumn Winter A Ryegrass stem refers to the actual stem plus the pseudostem.

6 692 Australian Journal of Experimental Agriculture J. M. Lee et al. Table 5. Treatment herbage yields (H 1 H 4 ; kg DM/ha), herbage yields during the water deficit recovery period (H 5 H 6 ) and total herbage yield for unfertilised plots (Control) and those receiving nitrogen (35 kg N/ha.defoliation in the form of urea; N) or glycine betaine (5.0 kg/ha.defoliation; GB) during experiment 2 The first harvest-date group was harvested on 21 December 2005 (H 1 ), 18 January 2006 (H 2 ), 15 February 2006 (H 3 ), 22 March 2006 (H 4 ), 19 April 2006 (H 5 ) and 27 May 2006 (H 6 ). The second and third harvest-date groups harvested 7 and 14 days later, respectively., P < ;, P < 0.001;, not significant Period Control GB N s.e.d. P-value H 1 Treatment H 2 Treatment H 3 Treatment H 4 Treatment H 5 Recovery H 6 Recovery Total under considerable stress and reducing herbage growth rates (Singleton 1991; McKenzie et al. 2000; Karsten and MacAdam 2001). Without the addition of water in such situations, plants will eventually die. In both experiments, there was adequate soil moisture for plant growth until mid January, at which point soil moisture levels declined below wilting point. Although the water deficit was more severe in experiment 1, lasting until mid May (106 days), in experiment 2 soil moisture levels did decline below the critical level for 10 days in January and again for 38 days in late February March. The lack of response of N in experiment 1 and the 40% decline in response to N in experiment 2 during summer confirm a lack of moisture availability, providing suitable conditions to test the osmoprotectant capabilities of GB. GB did not significantly affect herbage production or quality during either experiment, relative to unfertilised pastures. Previous research demonstrated an accumulation of GB in some plants, in response to a variety of environmental stresses (Rhodes et al. 1989; Rhodes and Hanson 1993; Kishitani et al. 1994; Gorham 1995). This response is normally seen in those plants which already contain some (>1 2 mol/m 3 plant water) GB in the unstressed condition (Gorham 1995). Attempts have been made to demonstrate the beneficial effects of GB on stressed plants by exogenous application and positive results have been recorded in some species (Agboma et al. 1997a; Agboma et al. 1997c; Diaz-Zorita et al. 2001). However, since soil and leaf microorganisms readily take up and metabolise GB (Rhodes and Hanson 1993; Gorham 1995; Agboma et al. 1997a), it can be difficult to interpret results from these studies. Although Naidu et al. (2003) demonstrated increased herbage growth rates during winter following foliar application of 1 kg GB/ha, it is possible that the surface application employed in the current study resulted in GB being degraded by soil microbes. However, even when GB application rates were increased to 5.0 kg/ha (experiment 2), there was still no effect of GB, relative to unfertilised pastures. This suggests that either the pasture plants were not taking up GB or, if they were, that it was not affecting the cellular structure to an extent that was translated into increased herbage production. In experiment 1, perennial ryegrass leaf size decreased in all treatment pastures during summer, coinciding with the period of hotter temperatures and largest decline in soil moisture. Research has shown that during periods of water deficit, there is a decrease in cell growth and therefore leaf elongation, resulting in smaller leaves (McKenzie et al. 2000). Interestingly, pastures treated with GB displayed the smallest decline in leaf size following Table 6. White clover leaf size (mm 2 ), perennial ryegrass leaf width and length (mm), leaf size (mm 2 ), leaf area index (LAI) and proportions (of above 40 mm herbage mass; DM basis) of perennial ryegrass leaf, perennial ryegrass stem and white clover in unfertilised plots (Control) and those receiving nitrogen (35 kg N/ha.defoliation in the form of urea; N) or glycine betaine (5.0 kg/ha.defoliation; GB) as measured at the third and fourth treatment harvests during experiment 2 The first harvest-date group was harvested on 15 February 2006 (H 3 ) and 22 March 2006 (H 4 ). The second and third harvest-date groups harvested 7 and 14 days later, respectively., P < 0.10;, P < 0.05;, P < ;, P < 0.001;, not significant. Harvest Control GB N s.e.d. P-value White clover leaf size H Perennial ryegrass leaf width H H Perennial ryegrass leaf length H H Perennial ryegrass leaf size H H Perennial ryegrass leaf area index H H Perennial ryegrass leaf H Perennial ryegrass stem A H H H White clover H H A Ryegrass stem refers to the actual stem plus the pseudostem.

7 Effect of glycine betaine on herbage production Australian Journal of Experimental Agriculture 693 commencement of moisture stress and therefore continued to produce larger leaves than unfertilised plants throughout the year. Increases in leaf area following GB application have also been recorded in drought-stressed tobacco plants (Agboma et al. 1997b). These increases are presumed to be a result of the ability of GB to prevent cellular dehydration and maintain turgor pressure and photosynthetic activity under conditions of low water potential (Larkum and Wyn Jones 1979; Borowitzka 1981). This effect, however, was not observed in experiment 2. Similar inconsistent effects of GB application on botanical composition were evident across experiments. In experiment 1, GB altered the botanical composition of the sward, increasing the proportion of white clover while decreasing that of ryegrass leaf. However, no such effect was recorded in experiment 2. There are several reasons why GB application may have affected plant growth factors in experiment 1 but not experiment 2. It is possible that the extent of the moisture deficit in experiment 2 was not severe enough to elicit a measurable response from GB, but that the moisture restriction in experiment 1 exceeded the required threshold, as it was both more severe and lasted longer. An additional reason for the difference between years may be due to differences in application rate. The response of plants to exogenous osmoprotectant application is dependant on application rate (Pospíšilová 2003). In previous experiments, application of low concentrations (1 µmol/l) of an osmoprotectant to water stressed bean (Phaseolus vulgaris L.) plants improved recovery, while higher concentrations (10 µmol/l) had a negative impact (Rulcová 2000; Rulcová and Pospíšilová 2001). It is possible that GB also has a concentration threshold, above which there is no effect, or potentially even a negative effect, and the higher concentration (5.0 kg/ha) of GB used in experiment 2 may have surpassed this threshold. Further research is required to determine if GB confers a protection to moisture-stressed perennial ryegrass and white clover plants and at what moisture restriction and application rate this protection exists. Consistent with the increased clover content in experiment 1, previous research has demonstrated that betaines confer protection against the detrimental effects of osmotic stress on N-fixing bacteria in lucerne, as well as to the host plant (Smith et al. 1988; Gorham 1995). It is plausible that this protection against osmotic stress also occurs in the N-fixing bacteria in white clover, as well as in the white clover plant itself. This would act to not only increase the plant s tolerance to water deficits, but also increase plant growth when compared with unfertilised plants, through maintaining N fixation processes and hence N availability. Components that determine the yield of clover include the number of growing points and the size of leaves, neither of which were measured in experiment 1. Further, although an effect was not evident in experiment 2, failure to exceed a moisture deficit threshold required for measurable responses cannot be ruled out. Application of N fertiliser increased (P < 0.001) total herbage production in both experiments. In experiment 1, individual harvest yields were increased throughout the year, except during summer, when the greatest decline in soil moisture levels occurred. Pastures fertilised with N during experiment 2 displayed increased herbage yields at all harvests, relative to unfertilised pastures, but the response was 40% less during the period of soil moisture deficit. Research demonstrates greater responses to N fertiliser where conditions of light, temperature and, in particular, moisture availability are favourable for herbage growth (Whitehead 1995). The fact that N fertiliser was effective in increasing herbage production during the period of low moisture availability in experiment 2 confirms that the water deficit was not as severe as that experienced during experiment 1. Research has extensively demonstrated that N fertiliser promotes greater herbage accumulation through production of larger leaves (Ryle 1964; Langer 1979; Whitehead 1995) and also increased tillering (Whitehead 1995; Wilman and Fisher 1996). Consistent with previous findings, N fertilisation in the current experiment increased (P < ) the size of perennial ryegrass leaves during both experiments by up to 54%, relative to unfertilised pastures. The increase in size was largely a result of increased leaf length (154 v. 175 mm for the unfertilised control and N, respectively; experiment 2). When combined with the increased herbage yields, this increase in size resulted in a greater LAI. Conclusion To compete with N fertiliser, any new growth enhancer needs to either result in similar increases in herbage production, or needs to be effective during periods that N fertiliser would not (i.e. during periods of moisture stress). A surface application of GB did not increase herbage production or quality relative to either unfertilised pastures, or those receiving N, thereby signifying that it does not confer reliable protection from environmental stresses to perennial ryegrass or white clover when applied in this way, and it is not suitable for use as a fertiliser. Acknowledgements The authors would like to acknowledge the statistical expertise of Barbara Dow, the help afforded by Scott Farm staff and the technical expertise of Penny Aspin, Glenise Ferguson, Angela Sheahan and Talia and Jordan van der Wel. This project was funded by Ballance AgriNutrients. References Agboma PC, Jones MGK, Peltonen-Sainio P, Rita H, Pehu E (1997a) Exogenous glycinebetaine enhances grain yield of maize, sorghum and wheat grown under two supplementary watering regimes. Journal Agronomy & Crop Science 178, Agboma PC, Peltonen-Sainio P, Hinkkanen R, Pehu E (1997b) Effect of foliar application of glycinebetaine on yield components of droughtstressed tobacco plants. Experimental Agriculture 33, doi: /s Agboma PC, Sinclair TR, Jokinen K, Peltonen-Sainio P, Pehu E (1997c) An evaluation of the effect of exogenous glycinebetaine on the growth and yield of soybean: timing of application, watering regimes and cultivars. Field Crops Research 54, doi: /s (97) Allard F, Houde M, Krol M, Ivanov A, Huner NPA, Sarhan F (1998) Betaine improves freezing tolerance in wheat. Plant & Cell Physiology 39, Borowitzka LJ (1981) Solute accumulation and regulation of cell water activity. In The physiology and biochemistry of drought resistance in plants. (Eds LG Paleg, D Aspinall) pp (Academic Press: Sydney)

8 694 Australian Journal of Experimental Agriculture J. M. Lee et al. Brereton AJ, Carton OT, O Keefe WF (1985) Tissue turnover in perennial ryegrass (Lolium perenne L.) during winter. Irish Journal of Agricultural Research 24, Corson DC, Waghorn GC, Ulyatt MJ, Lee J (1999) NIRS: forage analysis and livestock feeding. Proceedings of the New Zealand Grassland Association 61, Datta SK (2002) Recent developments in transgenics for abiotic stress tolerance in rice. International Rice Research Institute, JIRCAS Working Report, Philippines. Diaz-Zorita M, Fernandez-Canigia MV, Grosso GA (2001) Applications of foliar fertilizers containing glycinebetaine improve wheat yields. Journal Agronomy & Crop Science 186, doi: /j x x Fulkerson WJ, 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, doi: / EA00062 Gay AP (1993) Leaf area and light interception. In Sward measurement handbook. (Eds A Davies, RD Baker, SA Grant, AS Laidlaw) pp (British Grassland Society: Reading, UK) GENSTAT 7 (2004) GENSTAT for Windows. Release 7.2. (VSN International Ltd: Oxford, UK) Gorham J (1995) Betaines in higher plants biosynthesis and role in stress metabolism. In Amino acids and their derivatives in higher plants. (Ed. RM Wallsgrove) pp (Cambridge University Press: Cambridge, UK) Harinasut P, Tsutsui K, Takabe T, Nomura M, Takabe T, Kishitani S (1996) Exogenous glycinebetaine accumulation and increased salt-tolerance in rice seedlings. Bioscience, Biotechnology, and Biochemistry 60, Karsten HD, MacAdam JW (2001) Effect of drought on growth, carbohydrates, and soil water use by perennial ryegrass, tall fescue, and white clover. Crop Science 41, Kemp CD (1960) Methods of estimating the leaf area of grasses from linear measurements. Annals of Botany 24, Kishitani S, Watanabe K, Yasuda S, Arakawa K, Takabe T (1994) Accumulation of glycinebetaine during cold acclimation and freezing tolerance in leaves of winter and spring barley plants. Plant, Cell & Environment 17, doi: /j tb00269.x Langer RHM (Ed.) (1979) How grasses grow. 2nd edn. (Edward Arnold Ltd: London) Larkum AWD, Wyn Jones RG (1979) Carbon dioxide fixation by chloroplasts isolated in glycinebetaine, a putative cytoplasmic osmoticum. Planta 145, doi: /bf Le Rudulier D, Strom AR, Dandekar AM, Smith LT, Valentine RC (1984) Molecular biology of osmoregulation. Science 224, doi: /science Lopez CML, Takahashi H, Yamazaki S (2002) Plant-water relations of kidney bean plants treated with NaCl and foliarly applied glycinebetaine. Journal Agronomy & Crop Science 188, doi: /j x x Mäkelä P, Kleemola J, Jokinen K, Mantila J, Pehu E, Peltonen-Sainio P (1997) Growth response of pea and summer turnip rape to foliar application of glycinebetaine. Acta Agriculturæ Scandinavica Section B. Soil and Plant Science 47, McKenzie BA, Kemp PD, Moot DJ, Matthew C, Lucas RJ (2000) Environmental effects on plant growth and development. In New Zealand pasture crop and science. (Eds J White, J Hodgson) pp (Oxford University Press: Auckland, New Zealand) Mickelbart MV, Ejeta G, Rhodes D, Joly RJ, Goldsbrough PB (1999) Assessing the contribution of glycinebetaine to environmental stress tolerance in sorghum. In Molecular approaches for the genetic improvement of cereals for stable production in water-limited environments. An International Workshop, International Maize and Wheat Improvement Centre, El Batan, Mexico. (Eds JM Ribaut, D Poland) pp (CIMMYT: Mexico) Naidu BP, Stevenson G, Page R, Munford S (2003) Glycinebetaine foliar application increases pasture winter growth and milk yield. In Solutions for a better environment. Proceedings of the 11th Australian agronomy conference, Geelong. (Australian Society of Agronomy: Gosford, NSW) Pospíšilová J (2003) Participation of phytohormones in the stomatal regulation of gas exchange during water stress. Biologia Plantarum 46, doi: /a: Rhodes D, Hanson AD (1993) Quaternary ammonium and tertiary sulfonium compounds in higher plants. Annual Review of Plant Physiology and Plant Molecular Biology 44, doi: / annurev.pp Rhodes D, Rich PJ, Brunk DG, Ju GC, Rhodes JC, Pauly MH, Hansen LA (1989) Development of two isogenic sweet corn hybrids differing for glycinebetaine content. Plant Physiology 91, Roberts AHC, Morton JD (1999) Fertiliser use on New Zealand dairy farms. (New Zealand Fertiliser Manufacturers Association: Auckland, New Zealand) Roche JR (1995) The effect of closing date in autumn and turnout date in spring on sward characteristics, dry matter output and the milk performace of spring-calving dairy cows in early lactation. Masters Thesis. National University of Ireland, Dublin, Ireland. Rulcová J (2000) Effect of cytokinins on plant recovery after water stress. PhD thesis, Charles University, Czech Republic. Rulcová J, Pospíšilová J (2001) Effect of benzylaminopurine on rehydration of bean plants after water stress. Biologia Plantarum 44, doi: /a: Ryle GJA (1964) A comparison of leaf and tiller growth in seven perennial grasses as influenced by nitrogen and temperature. Journal of the British Grassland Society 19, Sakamoto A, Murata N (2000) Genetic engineering of glycinebetaine synthesis in plants: current status and implications for enhancement of stress tolerance. Journal of Experimental Botany 51, doi: /jexbot/ Singleton PL (1991) Soils of Ruakura a window on the Waikato. (DSIR Land Resources: Lower Hutt, New Zealand) Smith LT, Pocard J, Bernard T, Le Rudulier D (1988) Osmotic control of glycine betaine biosynthesis and degradation in Rhizobium meliloti. Journal of Bacteriology 170, Soil Survey Staff (1990) Keys to soil taxonomy. 4th edn. (United States Department of Agriculture: Blacksburg, VA) Talwar HS, Yanagihara S (1999) Physiological basis of heat tolerance during flowering and pod setting stages in Groundnut (Arachis hypogaea L.). Japan International Research Center for Agricultural Sciences, Working report no. 14, Ibaraki, Japan. Thomas JC, Sepahi M, Arendall B, Bohnert HJ (1995) Enhancement of seed germination in high salinity by engineering mannitol expression in Arabidopsis thaliana. Plant, Cell & Environment 18, doi: /j tb00584.x Whitehead DC (1995) Grassland nitrogen. (CAB International: Oxon, UK) Wilman D, Fisher A (1996) Effects of interval between harvests and application of fertilizer N in spring on the growth of perennial ryegrass in a grass/white clover sward. Grass and Forage Science 51, doi: /j tb02037.x Wilson JR (1981) Environmental and nutritional factors affecting herbage quality. In Nutritional limits to animal production from pastures. Proceedings international symposium, St Lucia, Queensland. (Ed. JB Hacker) pp (Commonwealth Agricultural Bureaux: Farnham Royal, UK) Zhao Y, Aspinall D, Paleg LG (1992) Protection of membrane integrity in Medicago sativa L. by glycinebetaine against the effects of freezing. Journal of Plant Physiology 140, Manuscript received 3 May 2007, accepted 25 November