SEARCH

SEARCH BY CITATION

Keywords:

  • grazing management;
  • pasture management;
  • sheep;
  • weeds

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

Chilean needle grass [Nassella neesiana (Trin. & Rupr.) Barkworth; CNG] is a perennial spear grass that has invaded pastures in south-eastern Australia and can lead to a substantial reduction of stockcarrying capacity during the summer months. This study examined a range of grazing, herbicide and pasture resowing options, alone or in combination, on CNG and introduced pasture grass basal cover, for several CNG-infested sites in south-eastern Australia. At each site, options were chosen on the basis that they were most likely to control the CNG infestation while maintaining a productive sheep-grazing enterprise on grass pastures. After 2 years of management, the reduction in CNG basal cover in set stock plots that were sprayed with flupropanate, versus those not sprayed, ranged from non-detectable to a reduction of 80%, depending on site location. After 5 years of management, the reduction ranged from zero to 50%. Grazing management or sowing of competitive pastures did not generally reduce CNG basal cover to low levels. None of the management options maintained reasonable levels of desirable perennial species by the end of 4–5 years. We conclude that, because of the persistence of CNG, the need for regular spraying of herbicide, and the relative ineffectiveness of other control methods, management systems may need to be developed that utilize CNG while minimizing its input to the soil seedbank.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

Chilean needle grass [Nassella neesiana (Trin. & Rupr.) Barkworth; CNG] is a highly invasive perennial spear grass (McLaren et al., 1998). Concern about the invasion of pastures in south-eastern Australia by CNG began to mount in the 1970s. Native to temperate South America, CNG can over-run pastures resulting in canopy cover exceeding 60%. Such infestations lead to a substantial reduction of stockcarrying capacity during the summer months when the weed produces large quantities of unpalatable inflorescences. By the 1990s, many livestock producers in New South Wales and Victoria found that they had expanding cover of CNG in their paddocks. Furthermore, the weed has invaded conservation areas with native grasslands, grassy woodlands and riparian vegetation. CNG is a declared noxious weed across Australia and is recognized as one of twenty ‘Weeds of National Significance’ due to its severe agricultural and environmental impacts (Thorp and Lynch, 2000).

Good grazing management can increase competition of beneficial pasture species resulting in a long-term decline in undesirable weed species (Lodge and Whalley, 1985; Dowling et al., 1996; Popay and Field, 1996; Bowman et al., 2009). In some circumstances, increased soil fertility can provide a competitive advantage to beneficial grass species over undesirable weed species (Kemp et al., 1996; Taylor and Sindel, 2000). Similarly, pasture rehabilitation is a recognized tool for renovating pastures and decreasing weed dominance (Jackson and Caldwell, 1992; Dowling et al., 2000; Taylor and Sindel, 2000).

It is known that herbicides such as glyphosate and flupropanate can be used to control CNG infestations (Lowien et al., 2001; Pritchard, 2004; Gaur et al., 2006; Grech, 2007). However, these herbicides can affect desirable pasture species (Campbell, 1997a; Taskforce, 2006). Thus, there is an issue of how to control CNG infestations, with or without the use of these herbicides, while maintaining a productive extensive grazing enterprise. Strategic grazing can preferentially advantage tall fescue over CNG because tall fescue has a faster growth rate after strategic grazing episodes (Gardener et al., 2005).

This study examined a range of grazing and chemical management options, alone or in combination, for several CNG-infested sites in south-eastern Australia. It monitored how they affect the control of CNG while maintaining desirable perennial grasses over a 5- to 6-year period. At each site, combinations of herbicide, pasture rehabilitation and grazing management were chosen based on their likely potential to control the CNG infestation while maintaining a productive sheep-grazing enterprise.

Materials and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

Experimental design

Two sites in Victoria (Toolleen and Greenvale) and two sites in New South Wales (Goulburn and Glen Innes) were chosen as representative of sheep-grazing properties with CNG-infested pasture in the high rainfall zone of south-eastern Australia (ANRA, 2011) (Fig 1, Table 1). Prior to the study, all sites were used for sheep grazing.

Table 1. Summary table of four sites. Rainfall data are obtained from the nearest meteorological site on Climate Data Online (http://www.bom.gov.au/climate/data/, accessed 1 September 2011)
MeasurementUnitGreenvaleToolleenGoulburnGlen Innes
pH (1:5 water) 5·576·305·905·8
Phosphorus
Olsen Pmg kg−110·335·00  
Colwellmg kg−1  21·00 
Braymg kg−18·003·9013·0012
Potassium
Skenemg kg−1110·00350·00  
Colwellmg kg−1  307·00 
Amm Ac.meq 100 g−1  0·600·4
Sulphur
CPC by ICPmg kg−18·007·00  
KCL40mg kg−1  18·005
Sodium
NH4OAc exchangemeq 100 g−10·800·32  
Amm Ac.meq 100 g−1  0·20<0·1
Calcium % of cations46·0049·00 65
Amm Ac.meq 100 g−1  8·18 
Average annual rainfallmm619579514849
summer rainfall (% of annual)mm24182836
Average Chilean needle grass basal cover – initial%11162613
Average desirable perennial grass basal cover – initial%72147
image

Figure 1. Map of trial site locations in Victoria and New South Wales, Australia. Capital cities are denoted by ▲.

Download figure to PowerPoint

At each site, an experiment was constructed using a three-replicate randomized block design. After consultation with district agronomists and land managers, each experiment had 8–12 grazing, herbicide spray type and pasture resown treatments that were chosen to suit district practice and seasonal conditions (Table 2).

Table 2. Sites with various combinations of grazing, herbicide spray type and pasture resown treatments
GrazingHerbicide Spray TypePasture ResownSites
  1. Go, Goulburn; Gr, Greenvale; T, Toolleen; GI, Glen Innes.

SetNoneNoGo, Gr, T, GI
SetGlyphosateNoGI
SetGlyphosateYesGI
SetSpraytopNoGI
SetFlupropanateNoGo, Gr, T, GI
StrategicNoneNoGo, Gr, T, GI
StrategicNoneYesGo, Gr
StrategicGlyphosateNoGI
StrategicGlyphosateYesGo, Gr, T, GI
StrategicSpraytopNoGI
StrategicFlupropanateNoGo, Gr, T, GI
Lock-upNoNoGo, Gr, T
Lock-upNoYesGo, Gr
Lock-upGlyphosateYesGo, Gr, T
Lock-upFlupropanateNoGo, Gr, T

All plots were 100 m2, except for lock-up and set stock treatments at Greenvale that were 25 m2. All plots were fenced off, within existing paddocks used for grazing sheep. ‘Set stock’ plots were unfenced at one end so that they received the same grazing pressure [typically eight Dry Sheep Equivalents (DSE) per ha (McLaren, 1997)] as the commercial paddock within which the experiment was located. Plot fences were made of sheep-proof mesh (Ringlock), with water troughs placed in the plots during grazing periods. Sheep used in the experiment were generally merino and merino-cross ewes and lambs, with a few Suffolk-cross ewes and merino-cross wethers at Greenvale. Sheep ranged from lambs to 7-year-old stock, and all sheep were accustomed to the district in which they were grazing. All grazing regimes were approved prior to experimentation by the relevant state Department of Primary Industries animal ethics committees.

At Greenvale and Toolleen, herbicide was sprayed using a hand-held boom sprayer (Azo-Dutch Sprayer). At Goulburn and Glen Innes, a boom sprayer mounted on a four-wheel motorbike or a hand-held sprayer was used to spray the plots. The rates of application were chosen to align with seasonal conditions and to suit the growth stage of the plant. The rate of flupropanate was chosen to limit non-target damage based on the soil type of each experimental site.

Flupropanate plots were sprayed at a rate of 1·12 kg a.i. ha−1 on 4 July 2003 at Toolleen and 17 July at Greenvale, except for lock-up flupropanate treatments, which were sprayed on 21 October 2003. Flupropanate plots at Glen Innes were sprayed at a rate of 2·24 kg a.i. ha−1 on 30 September 2003. Goulburn flupropanate plots were sprayed at a rate of 1·49 kg a.i. ha−1 on 1 November 2002. On 1 August 2007 at Greenvale and on 31 July 2007 at Toolleen, the flupropanate plots were split, and further basal cover measurements were carried out on one half of the plot.

At Glen Innes, glyphosate was sprayed at a rate of 1·13 kg a.i. ha−1 on 18 March 2004, for those plots that were also sown with perennial grass species. For plots that were not sown, glyphosate was sprayed on 18 March 2004 at a rate of 540 g a.i. ha−1. These non-sown plots were also resprayed with glyphosate on 14 April 2004 due to heavy storm rain within 2 h of the initial spray application on 18 March 2004. At Greenvale, glyphosate was sprayed on 4 May 2004 at a rate of 1·08 kg a.i. ha−1. Glyphosate (450 g a.i. L−1) was applied at Toolleen on 4 June 2004 at a rate of 2·5 L ha−1. Goulburn glyphosate treatments were sprayed on 18 February 2005 with 360 g a.i. L−1 formulation applied at 2 L ha−1.

Spraytop plots at Glen Innes were sprayed with glyphosate at 180 g a.i. ha−1 on 18 November 2003, with a follow-up application on 11 November 2005.

At all sites except Goulburn, destocking of the commercial paddock (i.e. all set-stock treatments) occurred during periods when CNG was flowering, to avoid vegetative matter contamination of wool and to maintain animal welfare standards. At Goulburn, the paddock did not have to be destocked during CNG flowering since the inflorescences were mechanically topped in spring each year.

‘Strategic’ grazing plots were grazed by three or more sheep for up to 1 week on an ‘as needed’ basis to reduce the production of CNG flower heads and to limit grazing selectivity. The sheep used for the strategic grazing plots were selected from the commercial paddock mob and placed into the plots at 2500–3000 kg DM ha−1 and were removed when dry matter yields had been reduced to 800–1000 kg ha−1. Strategically grazed plots were typically grazed five times throughout the year at a minimum stocking rate of 300 DSE ha−1 equivalent with grazing timed to coincide with CNG phenology to minimize seed production. ‘Lock-up’ plots were ungrazed and totally excluded grazing stock.

The selection of species and cultivars, together with the time of sowing and fertilizer rates for plots sown with pasture, was chosen to suit the district conditions, aided by soil tests (Table 3). Only treatments that were to be sown down to pasture seed were fertilized, except at Glen Innes where all treatments received fertilizer and white clover seed (250 kg ha−1 single super and 2 kg ha−1 Trifolium repens cv. Huia) on 12 December 2003. In 2003, pasture seed was sown aerially at Goulburn and direct drilled at Greenvale and Toolleen. Due to seasonal conditions, pastures failed to establish and were resown at the above sites in the 2004 season (Table 3). This second attempt at pasture resowing was more successful than the first. Pastures were only sown at Glen Innes in 2004 (Table 3).

Table 3. Pasture seed mix, sowing rate and fertilizer application rate for the pasture resown treatments at each site
DateGreenvaleToolleenGoulburnGlen Innes
19 May 20036 April 200426 June 200317 June 20041 May 20035 July 20058 July 2004
  1. Sub-clover seed was lime coated and inoculated. Seed rate is expressed as equivalent sowing rate of bare seed.

  2. *Cocksfoot (Dactylis glomerata), Phalaris (Phalaris aquatica), Ryegrass (Lolium perenne); §Subterranean clover (Trifolium subterraneum); White clover seed (Trifolium repens); **Red clover (Trifolium pratense); ††Tall fescue (Festuca arundinacea).

Pasture sowing fertilizer (kg ha−1)
Nitrogen18·018·018·018·015·015·017·9
Phosphorus28·820·028·820·021·021·015·0
Potassium16·40·016·40·00·00·00·0
Sulphur12·61·612·61·67·07·013·1
Pasture seeds (kg ha−1)
Cocksfoot Kara*3·05·03·05·02·50·8 
Cocksfoot Porto*    2·50·82·0
Phalaris Australian    5·01·6 
Phalaris Holdfast 3 3·05·01·63·0
Phalaris Sirosa 3 3·0   
Ryegrass Lincoln    2·50·8 
Ryegrass Boomer     0·8 
Ryegrass Kingston AS    2·5 2·0
Ryegrass Victorian 2·0 2·0   
Subterranean clover Goulburn§6·02·56·02·52·50·8 
Subterranean clover Seaton Park§    2·50·8 
Subterranean clover Trikkala§ 2·5 2·5   
White clover seed      2·0
Red clover**       
Tall fescue Jessup††9·0 9·0    
Tall fescue Flecha††9·0 9·0    
Tall fescue Demeter††      5·0
Total pasture seed sown27·018·027·018·025·08·014·0

Measurements

Basal cover composition was recorded in each plot, at each site, on a seasonal basis over the period of 2004–2007 for Glen Innes and over 2003–2007 for the other three sites. For each site and year, there were three or four sampling occasions. The initial basal cover of CNG and desirable perennial grasses (Table 1) was obtained from the first sampling occasion of each site. Basal composition was measured at a fixed location within each plot using a 100 point 1 × 1 m quadrat, and recording 100 point basal observations (i.e. a basal observation is recorded if the plant extends into the soil at the point of measurement). The location was chosen to be in the same relative position within every plot of a site (1·4 m diagonally into plot from the same corner of every plot). The basal measurements recorded were CNG, perennial desirable grasses (e.g. Festuca arundinacea, Phalaris aquatica, Lolium perenne, Dactylis glomerata) and other (broadleaf plants, annual grasses, legumes, vegetative litter and bare soil). Plant biomass was estimated, not measured, in this experiment as the study investigated the effects of management on the extent of presence of CNG rather than feed available for grazing.

At each site, soil sampling to a depth of 100 mm was carried out at the first sampling occasion. The soil samples were then sent to a commercial laboratory for chemical analysis.

Statistical analysis

In most cases, the basal cover of CNG changed in a relatively smooth manner over time and only changed slowly over annual growth cycles (Grech, 2007). Thus, for each plot, annual basal cover of CNG and of desirable perennial grass was calculated by averaging over all samples in each year. For each combination of site and year, the annual basal cover of CNG was analysed using a general linear model analysis of variance with additive terms for replicate, grazing, herbicide spray type and with a final term for any treatment effect that was additional to the additive effects of grazing and spray type (noting that all treatment effects are orthogonal to the replicate effect). The term for additional effects includes the grazing and spray type interaction, as well as all effects involving pasture resown treatments. The experimental unit was a plot. This form of analysis of variance was chosen because preliminary analyses indicated that most treatment effects could be described by additive effects of grazing and herbicide type. All herbicide spray-type means are presented after adjusting for grazing effects and presented for set stock grazing. All grazing means are presented after adjusting for spray-type effects and presented for no herbicide application. All P values and standard errors are calculated using the residual mean square from the analysis of variance with saturated treatment effects. Angular transformations were examined for these analyses but, as there was little difference in the pattern of residual variation between the transformed and untransformed analysis, the untransformed analysis is reported.

With desirable perennial grass, the annual basal cover of each plot was log10(y + 1) transformed prior to similar general linear model analysis as that used for CNG basal cover. This transformation prevented the residual variation increasing as the treatment mean increased. One outlying plot, with very little desirable perennial grass despite being in the lock-up grazing treatment, was deleted from the 2005 analysis for Greenvale.

Sampling considerations

Pastures are notoriously heterogeneous, and although the sampling area used in this experiment was less than in certain other ecological studies of grazed pastures in south-eastern Australia (Lodge and Orchard, 2000; Hill et al., 2004), our plot size was also smaller than in these other studies. The statistics also appear to justify our levels of replication and sampling. Although there was only one quadrat per plot, it was 1 m2 in size, each quadrat had 100 points sampled and there were three plots per treatment, giving 300 basal point sampling on each occasion over an area of 3 m2. The results for each treatment were also analysed on a yearly basis for several years, and the non-spatial component of sampling variability was averaged from three to four sampling occasions per year. This sampling intensity provided good precision, as indicated by the relatively small standard errors of difference obtained in tables presented within the Results section of this paper (Tables 4-7). The analyses of variance also had 14–18 degrees of freedom, depending on site, which was greater than the minimum of 10 required to obtain a reasonable estimate of the residual variance (Stern et al., 2004).

Table 4. Effect of spray type, adjusted for grazing method, on the average basal cover (%) of Chilean needle grass at different sites and years. Values are predicted for the set stocking method
 No spray (a)Flupropanate (b)Glyphosate (c)Spraytop (a) (b)sed (standard error of difference)P value
(a) vs (b)(a) vs (c)(b) vs (c)
  1. *Bold values are statistically significant.

Goulburn
20083936324·35·05·60·37
20074439354·85·56·20·23
20063127224·95·66·30·30
20053631153·03·43·8 <0·001
20043327'173·94·45·0 0·007
20033421192·93·43·7 <0·001
Greenvale
20083222334·65·35·90·13
2007147131·82·12·4 0·004
20063415273·33·84·3 <0·001
20054216353·84·44·9 <0·001
200417983·13·64·1 0·022
200311991·82·12·30·45
Glen Innes
2008442241174·74·04·0 <0·001
2007421737115·04·34·3 <0·001
20062452133·12·72·7 <0·001
20051831122·82·42·4 <0·001
2004181495·64·94·9 0·024
Toolleen
20082924333·03·03·0 0·018
2007116102·32·32·30·17
20061814192·12·12·10·071
20053227273·03·03·00·18
200414241·21·21·2 <0·001
20032116161·91·91·9 0·030
Table 5. Effect of grazing method, adjusted for spray type, on the average basal cover (%) of Chilean needle grass at different sites and years. Values are predicted for the no-spray treatment
 Set (a)Strategic (b)Lock-up (b)sedP value
(a) vs (b)(b) vs (b)
  1. *Bold values are statistically significant.

Goulburn
20083929265·24·10·060
20074431265·84·6 0·021
20063119185·94·60·11
20053625253·62·8 0·014
20043327244·73·70·19
20033432343·52·80·76
Greenvale
20083228285·64·40·74
20071416152·21·70·63
20063424213·24·0 0·019
20054230284·63·6 0·022
20041716183·83·00·79
20031112132·21·70·74
Glen Innes
200844443·00·98
200742413·20·67
200624231·90·65
200518161·80·34
200418183·60·88
Toolleen
20082936313·52·60·086
20071111192·62·0 <0·001
20061820172·41·80·29
20053221223·42·6 0·012
20041414161·41·00·12
20032118192·21·70·54
Table 6. Effect of spray type, adjusted for grazing method, on the average basal cover (%) of desirable perennial grass at different sites and years. All values are predicted for the set-stocking method
 log (y + 1) transformedBack-transformed meanP value
No sprayFlupropanateGlyphosateSpraytopsed (range)No sprayFlupropanateGlyphosateSpraytop
  1. *Bold values are statistically significant.

Goulburn
20080·180·220·600·20–0·262260·20
20070·230·100·470·18–0·232140·28
20060·350·220·710·12–0·16326 0·020
20050·980·850·610·08–0·101185 <0·01
20040·350·400·970·11–0·143311 <0·001
20030·520·490·850·15–01954100·16
Greenvale
20080·470·860·530·16–0·2131040·080
20070·160·480·220·12–0·161310·051
20060·330·740·520·14–0·18263 0·028
20050·520·990·840·09–0·133107 <0·001
20041·051·051·070·07–0·091111120·97
20031·251·171·390·08–0·111915190·15
Glen Innes
20080·060·720·140·350·15–0·170401 0·0035
20070·351·100·450·460·12–0·1311222 <0·001
20060·711·280·620·720·10–0·1241834 <0·001
20051·371·511·241·100·08–0·0922311611 <0·01
20041·461·581·441·390·10–0·12283727240·44
Toolleen
2008−0·040·040·070·090000·50
2007−0·050·05−0·010·090000·50
20060·090·180·360·130110·13
2005−0·130·130·540·14002 <0·001
20040·410·580·980·09239 <0·001
20030·540·320·780·11315 <0·01
Table 7. Effect of grazing method, adjusted for spray type, on the average basal cover (%) of desirable perennial grass at different sites and years. All values are predicted for the no-spray treatment
 log (y + 1) Transformedsed (range)Back-transformed meanP value
SetStrategicLock-upSetStrategicLock-up
  1. *Bold values are statistically significant.

Goulburn
20080·180·210·280·19–0·242220·91
20070·230·400·550·17–0·222350·11
20060·350·500·460·12–0·153440·61
20050·980·961·030·08–0·101110120·64
20040·350·370·510·11–0·143340·33
20030·520·490·460·14–0·185440·95
Greenvale
20080·470·770·920·16–0·2038110·10
20070·160·550·930·12–0·151310 <0·001
20060·330·840·910·14–0·172810 <0·01
20050·521·041·050·09–0·1131112 <0·001
20041·051·141·010·07–0·091114100·17
20031·251·211·160·08–0·101917150·64
Glen Innes
20080·060·160·11000·37
20070·350·360·08110·83
20060·710·700·08440·87
20051·371·300·0622220·79
20041·461·390·0728230·32
Toolleen
2008−0·040·070·030·08–0·110000·61
2007−0·05−0·010·150·07–0·100000·065
20060·090·010·200·11–0·150010·23
2005−0·130·320·330·12–0·16011 0·016
20040·410·340·220·08–0·102110·16
20030·540·800·820·10–0·133550·11

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

The additional treatment effects were jointly tested first. These included the main effects of pasture resowing, the interaction effects between grazing and spray type, and all interactions between pasture resown and grazing/spray type effects. Treatment effects for the additive effects of spray type and grazing type were then tested.

There were few site/year combinations in which these additional effects (as listed above) were statistically significant at the 5% level. On the few times they were significant (P < 0·05), the effects were ephemeral, they were not associated with main effects for either spray type and grazing method, and the P value was greater than 0·01. Consequently, it is reasonable to only report main effects for spray type (adjusted for grazing method) and grazing method (adjusted for spray type) on the basal cover of CNG.

The strongest effects observed were due to spray type (Tables 4 and 5). At each site, glyphosate application reduced CNG basal cover compared with no spraying for about 2–3 years, but no effect was detectable after this time (Table 4). Flupropanate application only reduced CNG basal cover for 1 year at Goulburn and 2 years at Toolleen. At Greenvale and Glen Innes, flupropanate application reduced CNG basal cover for all the study period but, at both sites, the CNG basal cover was above 20% by 2008. Glyphosate spray topping at Glen Innes kept the CNG basal cover to levels similar to the application of flupropanate (Table 4).

It was common for the CNG basal cover to be lower with strategic grazing and lock-up than with set stocking (2004–2007 at Goulburn, 2005–2006 at Greenvale and 2005 at Toolleen, Table 5). Exceptions were at Glen Innes where strategic grazing had no effect on CNG in any year and at Toolleen in 2007 where nil grazing had higher CNG basal cover than set stocking.

By 2008, all treatments at all sites had substantial CNG basal cover, in the range of 20–50% (Tables 4 and 5).

In line with basal cover of CNG, there were only isolated additional effects to the additive effects of spray type and grazing type on the average basal cover of desirable perennial grasses (P < 0·05). Consequently, we only report main effects for spray type and grazing method on basal cover of desirable perennial grass.

Glyphosate led to some increase in desirable perennial grass basal cover at Goulburn in 2004 and 2006, and at Toolleen in 2003–2005 (Table 6). However, glyphosate was associated with a decline of desirable grass at Goulburn in 2005. Flupropanate led to some increase in desirable perennial grass basal cover at Greenvale in 2005 and 2006, and at Glen Innes in 2005–2008. Spray topping at Glen Innes led to a decline in desirable perennial grass basal cover in 2005.

From 2005 onwards, at Greenvale, there was more desirable perennial grass basal cover with strategic grazing and locking up than with set stocking (Table 7).

Nevertheless, with the exception of locking up at Greenvale, during 2007 and 2008 there were only negligible to small amounts (up to 15%) of desirable perennial grass basal cover at all sites and in all treatments.

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

None of the management options was able to reduce CNG basal cover to low levels and maintain reasonable levels of desirable perennial species by the end of 4–5 years.

Pasture sowing treatments were intended to provide a source of competition to CNG for light, soil resources and to resist weed invasion (Campbell, 1997b; Friend and Kemp 2000; Jackson and Caldwell, 1992; Moretto and Distel 1997). Unfortunately, pasture sowing had little to no effect in controlling CNG at any site. Nor did it increase the basal cover of desirable perennial grasses. Although each site had at least one reasonably successful pasture establishment, unseasonably dry conditions may have limited the vigour of introduced pasture species during the experimental period. Nevertheless, the lack of success of pasture sowing, with the methods used, places a question mark on the use of introduced pasture species to control established CNG infestations.

Strategic grazing has been advocated as a management tool to promote healthy pastures and reduce weed impacts (Lodge and Whalley, 1985; Kemp et al., 1996; Campbell, 1997b). For CNG, the theory is that heavy strategic grazing decreases selective grazing, so that all species are grazed evenly and the less palatable species are not left to run to seed. This would theoretically allow the palatable species to outcompete CNG, provided they have a faster growth rate than CNG or their phenology does not coincide with CNG. However, although strategic grazing provided some decrease in CNG basal cover and some increase in desirable grass basal cover at some sites and times, these decreases were insufficient to claim control of CNG. It is unlikely that the benefit could be improved by refining the strategic grazing regimes because, at the three sites with lock-up treatments, it was rare for the strategic grazing and lock-up treatments to not have similar CNG and desirable perennial grass basal cover. It seems that while the suggestion by Gardener (Gardener et al., 2005) that strategic grazing can preferentially disadvantage CNG is sometimes correct, it also seems that it is not a sufficiently strong management tool to enable CNG infestations to be controlled where the infestation is well established.

Generally, the most effective way to reduce CNG basal cover and increase desirable perennial grass basal cover was through the use of flupropanate. As a selective herbicide, flupropanate can be used to limit damage to non-target pasture species such as phalaris (Phalaris aquatica) and kangaroo grass (Themeda triandra) (Taskforce, 2006). However, our results were inconsistent between sites. For instance, at Greenvale and Glen Innes, there was benefit from using flupropanate for the whole experiment, whereas at Goulburn and Toolleen, there was no detectable benefit after 1–2 years.

Glyphosate is able to kill standing CNG plants and minimize seed production (Gaur et al., 2006). However, it is non-selective. At each site in our study, a reduction of CNG basal cover from glyphosate application could only be detected for 2–3 years after application.

Spray topping at Glen Innes provided a reduction of CNG basal cover for the whole study. However, there was no increase in desirable perennial grass basal cover at any time, implying that the benefit of spray topping will be marginal unless other techniques are used to increase the density of desirable perennial grasses. Pasture resowing is generally used to increase pasture density, although this technique was not effective in this study. Although spray topping may offer additional benefits such as reduced seed production or viability (Gaur et al., 2005), which may lead to consequent reductions in the contamination of fleece by vegetative -material, these were not measured in this experiment.

By the last year of the experiments (2008), all treatments had substantial CNG basal cover (>20%) and limited (≤15%) desirable perennial grass basal cover. Hence, repeated herbicide application every few years would be needed to maintain reasonable control of CNG in a grass pasture sheep system. From our results, it appears that it is not feasible to eradicate CNG where infestations are well established, while maintaining a grass pasture–based grazing system.

An alternative approach for extensive grazing enterprises is to develop management systems that utilize CNG while minimizing soil seedbank inputs. This is because CNG has a relatively low nutritional value, especially during reproductive stages, when compared with most desirable perennial grass species (Gardener, 1998; Grech et al., 2004). In addition, the panicle seeds produced during reproductive stages are sharp, causing injury to livestock as well as contamination in wool (Grech et al., 2006). Depleting the soil seedbank is challenging for CNG, as it is often found in land classes that cannot be managed by conventional means, such as cultivation. Nevertheless, we believe that developing appropriate management systems is worth investigating for a range of land classes. Strategies that should be considered, other than repeated herbicide use, may include management options such as forage cropping and changing the species of livestock.

Conclusion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

Herbicide application was the only treatment that provided reasonable control of CNG in these experiments, but the results were generally short term. For longer-term control, herbicides may need to be applied every few years. Given the potential for the development of herbicide resistance (McLaren et al., 2010), the dislike of chemicals by a significant proportion of graziers (Van der Meulen et al., 2007), the apparent persistence of CNG and the lack of effective management options for grass pasture–based grazing systems in south-eastern Australia, management systems may need to be developed that utilize CNG while minimizing the soil seedbank input.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

We thank the landholders and managers, Ian and Margaret Souter, John and Alan McKenzie, Roger and Brian Hickson, Peter Sastrom and Carolyn Muir. Financial support was provided by the Australian Commonwealth Government through the Natural Heritage Trust and the CRC for Australian Weed Management and DPI in Victoria and New South Wales. Our thanks also go to regional staff from DPI across New South Wales and Victoria, the University of New England, La Trobe University and the University of Melbourne for their assistance, especially David Chapman, Fred Fenn, Dale Chalker, Julio Bonilla, Natasha Baldyga, Brad Westhead and Jeff Wilkie. Thanks also to James Winters, Aaron Dodd, Caroline Ayres, Noel Campbell, Bruce McGregor, Tamara Threlfall and Michelle Whiteley for actively assisting in the management of the experiments.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References
  • ANRA (2011) Agriculture – sheep/wool industry – high rainfall zone. (online), http://www.anra.gov.au/topics/agriculture/sheep-wool/region-high-rainfall-zone.html (accessed 23 September 2011)
  • Bowman A.M., Alemseged Y., Melville G.J., Smith W.J. and Syrch F. (2009) Increasing the perennial grass component of native pastures through grazing management in the 400–600 mm rainfall zone of central western nsw. The Rangeland Journal, 31, 369376.
  • Campbell M.H. (1997a) Effect of low rates of flupropanate on selective removal of serrated tussock (Nassella trichotoma (Nees) Arech.) seedlings from a young improved pasture. Plant Protection Quarterly, 12, 175176.
  • Campbell M.H. (1997b) Pasture weeds. In: Lovett J.V. and Scott J.M. (eds) Pasture production and management, pp. 254266. Melbourne: Inkata Press.
  • Dowling P.M., Kemp D.R., Michalk D.L., Klein T.A. and Millar G.D. (1996) Perennial grass response to seasonal rests in naturalised pastures of central New South Wales. The Rangeland Journal, 18, 309326.
  • Dowling P.M., Michalk D.L. and Sindel B.M. (2000) Weed management in pasture systems. In: Sindel B.M. (ed.) Australian weed management systems. Melbourne: R.G. & F.J. Richardson.
  • Gardener M.R. (1998) The biology of Nassella neesiana (Trin. & Rupr.) Barkworth (Chilean needle grass) in pastures on the northern tablelands of New South Wales: Weed or pasture? PhD, The University of New England, Armidale, NSW.
  • Gardener M.R., Sindel B.M., Whalley R.D.B. and Earl J.M. (2005) Can Nassella neesiana, Chilean needle grass, be incorporated into a grazing management system in australia? Plant Protection Quarterly, 20, 3640.
  • Gaur S., McLaren D.A. and Butler K. (2005) Spraytopping as a management tool to reduce seed production in Chilean needle grass infestations. Proceedings of the Tussock Terminators Research Forum, Lake Hume Albury, New South Wales, pp. 3642.
  • Gaur S.K., McLaren D.A., Butler K. L. and Bonilla J. (2006) Broadacre chemical control for Chilean needle grass infestations. In: Preston C., Watts J.H. and Crossman N.D. (eds) Proceedings of the 15th Australian Weeds Conference, Adelaide Australia, pp. 864867.
  • Grech C.J. (2007) Grazing management for the long term utilisation and control of Chilean needle grass (Nassella neesiana). PhD. thesis, University of New England.
  • Grech C., McLaren C., Chapman D.F. and Sindel B. (2004) Effect of clipping and fertiliser addition on the feed value of Chilean needle grass (Nassella neesiana) during reproductive growth stages. In: Sindel B.M. and Johnson S.B. (eds) Proceedings of the 14th Australian Weeds Conference, Wagga Wagga, p. 110.
  • Grech C., McLaren D.A. and Sindel B.M. (2006) Effect of grazing management and stockclass on Chilean needle grass (Nassella neesiana) seed production. In: Preston C., Watts J.H. and Crossman N.D. (eds) Proceedings of the 15th Australian weeds conference, Adelaide, p. 868.
  • Hill J., Simpson R.J., Moore A.D., Graham P. and Chapman D.F. (2004) Impact of phosphorus application and sheep grazing on the botanical composition of sown pasture and naturalised, native grass pasture. Australian journal of Agricultural Research, 55, 12131225.
  • Jackson R.B. and Caldwell M.M. (1992) Shading and the capture of localized soil nutrients: nutrient contents, carbohydrates, and root uptake kinetics of a perennial tussock grass. Oecologia, 91, 457462.
  • Kemp D.R., Dowling P.M. and Michalk D.L. (1996) Managing the composition of native and naturalised pastures with grazing. Journal of Agricultural Research, 39, 569578.
  • Lodge G.M. and Orchard B.A. (2000) Effects of grazing management on sirosa phalaris herbage mass and persistence in a predominantly summer rainfall environment. Australian Journal of Experimental Agriculture, 40, 155169.
  • Lodge G.M. and Whalley R.D.B. (1985) The manipulation of species composition of natural pastures by grazing management on the northern slopes of New South Wales. Australian Rangeland Journal, 7, 616.
  • Lowien J., Gardener M.R. and Cameron A. (2001) Chilean needle grass (Nassella neesiana) management options. Proceedings of the Thirteenth Annual Conference, Grasslands Society of New South Wales, pp. 8485.
  • McLaren C. (1997) Dry sheep equivalents for comparing different classes of livestock, pp. 14. Attwood, Victoria: Department of Primary Industries.
  • McLaren D.A., Stajsic V. and Gardener M.R. (1998) The distribution and impact of South/North American stipoid grasses (Poaceae: Stipeae) in Australia. Plant Protection Quarterly, 13, 6270.
  • McLaren D.A., Grech C., J., Bonilla J., Butler K. and Ramasamy S. (2010) Serrated tussock resistance to flupropanate in australia – is the genie out of the bottle? In: Zydenbos S.M. (ed) Proceedings of the Seventeenth Australasian Weeds Conference, New Zealand, pp. 266269.
  • Popay I. and Field R. (1996) Grazing animals as weed control agents. Weed Technology, 10, 217231.
  • Pritchard G.H. (2004) Effect of herbicide application to emerged seed heads on the viability of panicle seed and cleistogenes of Chilean needle grass (Nassella neesiana). In: Sindel B.M. and Johnson S.B. (eds) Proceedings of the Fourteenth Australian Weeds Conference, Wagga Wagga Australia, pp. 274278.
  • Stern R.D., Coe R., Allan E.F. and Dale I.C. (2004) Good statistical practice for natural resources research. Wallingford, Oxfordshire, UK: CABI publishing.
  • Taskforce (2006) Taskforce herbicide frequently asked questions. (online), http://www.taskforceherbicide.com/faq (accessed on 30 May 06)
  • Taylor U. and Sindel B. (2000) The pasture weed management kit, a guide to managing weeds in southern australian perennial pastures. Adelaide: Douglas Press.
  • Thorp J.R. and Lynch R. (2000) The determination of weeds of national significance. Launceston: Weeds Strategy Executive Committee.
  • Van der Meulen A.W., Reeve I.J. and Sindel B.M. (2007) Weed management on grazing properties: a survey of livestock producers in new south wales and victoria. Australian Journal of Experimental Agriculture, 47, 14151421.