Skip to main content
NCBI home page
Annals of Botany logoLink to Annals of Botany
. 2005 Jul 26;96(5):799–809. doi: 10.1093/aob/mci230

Competition for Nitrogen between Australian Native Grasses and the Introduced Weed Nassella trichotoma

WARWICK B BADGERY 1,*, DAVID R KEMP 1, DAVID L MICHALK 2, WARREN M C G KING 2
PMCID: PMC4247044  PMID: 16046460

Abstract

• Background and Aims Nassella trichotoma is an unpalatable perennial grass weed that invades disturbed native grasslands in temperate regions of south-eastern Australia. This experiment investigated whether elevated N levels, often associated with disturbance, increases the competitiveness of N. trichotoma relative to C3 and C4 native Australian grasses.

• Methods A pot experiment investigated competitive interactions between four native grasses, two C3 species (Microlaena stipoides and Austrodanthonia racemosa) and two C4 species (Themeda australis and Bothriochloa macra), and N. trichotoma at three different N levels (equivalent to 0, 60 and 120 kg ha−1) and three competing densities (zero, one and eight neighbouring plants), using an additive design.

• Key Results All native grasses were competitive with N. trichotoma at low N levels, but only M. stipoides was competitive at high N. High densities of native grasses (8 : 1) had a major competitive effect on N. trichotoma at all N levels. The competitive ranking of native grasses, across all N levels, on N. trichotoma was: M. stipoides > A. racemosa > B. macra > T. australis. The C3 species were generally more competitive than the C4 species and C4 grasses were not inherently more productive at low N levels, in contrast to the results of other studies.

• Conclusion To resist invasion from N. trichotoma, these native grasses need to be maintained at a high density and/or biomass. The results do not support the theory that species such as N. trichotoma, with high tissues density, are always less competitive than those of low tissue density; in this case competitiveness depended on N levels. The ability of N. trichotoma to accumulate biomass at a higher rate than these native grasses, helps to explain why it is a major weed in disturbed Australian native grasslands.

Keywords: Australia, C3 and C4 competition, invasive species, Nassella trichotoma, native grasses, perennial grass weed, pot experiment, unpalatable grass

INTRODUCTION

Nassella trichotoma is a C3 perennial grass native to temperate South America. It is unpalatable to livestock because of its high leaf fibre content (Campbell and Irvine, 1966), low tissue nutrient content (Campbell, 1965) and because of its growth habit—a moderate-sized tussock usually containing much senesced material (Campbell, 1998).

Over the past century N. trichotoma has invaded grasslands in south-eastern Australia (Jones and Vere, 1998; McLaren et al., 1998), South Africa (Wells, 1974) and New Zealand (Healy, 1945), and has been recorded in the northern hemisphere (Ridley, 1930; Westbrooks and Cross, 1993). In Australia, N. trichotoma now occurs over more than a million hectares (Jones and Vere, 1998; McLaren et al., 1998) and in some situations completely dominates grasslands (Campbell and Vere, 1995). The presence of N. trichotoma reduces grassland plant species diversity, and livestock production declines in proportion to N. trichotoma abundance (Auld and Coote, 1981). The management of N. trichomata is complicated because much of the area invaded by N. trichotoma is steep and inaccessible, or site productivity is low (Jones et al., 2000).

The invasion of pastures by N. trichotoma is commonly associated with heavy grazing (Healy, 1945; Campbell, 1977). Continuous grazing combined with other disturbances has changed the composition of most native grasslands (Moore, 1970; Garden and Bolger, 2001); species that are less grazing-tolerant, e.g. Themeda australis, have been reduced in abundance or eliminated (Mott et al., 1992). The reduced native perennial grass biomass offers little competition to invading N. trichotoma plants (Campbell and Vere, 1995; Badgery, 2004).

Pasture composition is shaped by a complex of mechanisms including competition for more than one resource. In the generally nutrient-poor soils where N. trichotoma commonly invades, competition for nutrients is one of the more important mechanisms. Australian native grasses have evolved generally under lower soil fertility (especially N) than has N. trichotoma (Badgery, 2004). Moreover, the Australian flora contains a significant number of C4 perennial grasses and it has been postulated that these have the ability to grow at lower soil N concentrations than C3 grasses (Wedin and Tilman, 1996). In the absence of defoliation, and at low soil N concentrations, it might be expected that N. trichotoma would be less competitive than Australian native perennial grasses, especially C4 species, in the absence of grazing. If true this would be important for weed management.

The relationship between N. trichotoma and other perennial grass species and the interaction with grazing has been described from field studies (Badgery et al., 2002). However, little research has examined the mechanisms that underpin the responses between these species. Understanding these competitive relationships is considered central to the development of integrated weed management systems for N. trichotoma (Michalk et al., 1999).

In the study reported here (part of a wider programme investigating the ecology and management of N. trichotoma) competitive relationships between N. trichotoma and the Australian native perennial grass species (Austrodanthonia racemosa and Microlaena stipoides, C3 species; Bothriochloa macra and T. australis, C4 species) were investigated in a glasshouse experiment. The effects of soil N status, stage of plant growth and density were investigated in the absence of defoliation, to determine the competitive potential of native grasses with N. trichotoma for the development of integrated weed management systems for non-arable landscapes.

MATERIALS AND METHODS

This experiment was done in a temperature-controlled glasshouse at the Orange Agricultural Institute (33·39 °S 149·02 °E), Orange, New South Wales, Australia. The experiment had two different starting times: June 2000 (winter) for the two C3 native grasses (A. racemosa and M. stipoides) and January 2001 (summer) for the C4 native grasses (B. macra and T. australis). The different starting times ensured appropriate conditions for the growth of C3 and C4 grasses as glasshouse conditions did vary with seasons. Seed was obtained for each species from local grasslands. The temperature in the glasshouse was controlled using evaporative air-conditioning and heaters: the mean daily temperature was 18 °C, ranging from 11 to 24 °C from winter to summer.

Experimental design

The experiment was a factorial additive design where neighbouring species (A. racemosa, M. stipoides, B. macra or T. australis) were planted at three densities (zero, one and eight neighbouring plants) around a single N. trichotoma plant and fertilized with three N levels (equivalent to 0, 60 and 120 kg N ha−1). Pots were harvested at either 12 or 36 weeks to estimate seedling and mature plant competition separately. The experiment was replicated four times (72 pots per species per harvest).

The experiment used a variation of an additive series design in which N. trichotoma was always at the centre of the pot (150 mm diameter, 1·6-L pots) with the neighbouring plants at an equal distance from the target plant (∼37 mm) and from each other (∼30 mm for eight native plants per pot). Each of the neighbouring species was also grown as a monoculture (i.e. without any target plants), at both densities (one and eight), to enable separate assessment of the competitive effect of the target plant and intraspecific competition. The experimental layout was a completely randomized design and pot positions were re-randomized every 2 weeks to minimize the effects of variations in watering, temperature and light across the benches.

The pots were filled with a Ferrosol soil (Isbell, 1996), in which N. trichotoma commonly occurs. It was first passed through a 5-mm sieve and mixed (1 : 1) with sand to reduce the N level. Superphosphate (9 % P applied at rate equivalent to 250 kg ha−1) and lime (2·5 t ha−1) were added to each pot to ensure adequate pH and P supply to isolate N as the sole limiting nutrient from these other common deficiencies.

Nitrogen was applied in the form of NH4NO3 (Nitram®) fertilizer. The fertilizer was applied at 4-week intervals for plants harvested at 12 weeks and at 8-week intervals for plants harvested at 36 weeks. The first application was 1 week after the experiment began. The equivalent of 0, 20 or 40 kg of N ha−1 was applied three times to give a total N application of 0, 60 or 120 kg ha−1 for each treatment.

Seeds were germinated on pads in Petri dishes and transplanted to the pots after 1 week when they were approx. 5–10 mm high. Any seedlings that died after planting were replaced during the first 2 weeks. After this time they were not replaced.

Both root and shoot masses were measured at each harvest to assess the impacts of competition. Soil was washed from the roots first and then each plant was oven dried (60 °C for 48 h), cut to separate the roots from the shoots, and weighed. The root samples were retained and representative samples of different plant species, harvest dates and plant sizes were ashed to estimate the amount of residual soil that could not be washed from the roots. Analyses showed that only plant size appeared to affect the amount of residual soil retained on the roots. Linear regressions between root weight plus residual soil versus residual soil were used to correct root weights, which were then combined with shoot weights to give corrected weights for whole plants. The corrected total biomass (shoot + root) data for all plants was natural-log transformed prior to analysis.

As N. trichotoma was the only target plant and at times there was only one competing neighbour, pots were excluded when mortality occurred in one of these plants. Mortality may have been due to competition, but since plant death may have been caused by other factors these pots were excluded. In part of this experiment (not reported in this paper) Nassella trichotoma seedlings were transplanted into pots of four mature neighbouring native grasses, with very low success rates from several attempts. In this case, mortality was probably due to competition as the losses were greater than in other treatments. However, the large numbers of plant deaths meant that valid statistical analyses could not be done reliably.

Analyses

Differences between the biomass of plants of each species grown without competition were tested using ANOVA (Genstat® for Windows, 7th edition; Payne et al., 2003). For N. trichotoma, the two different start times, winter and summer, were analysed separately. N level and harvest time were used as factors in this analysis. The competitive response of N. trichotoma to neighbouring plant species and density was analysed at different harvests at each N level, using ANOVA. The competitive effect of N. trichotoma on each neighbouring species was then analysed at each harvest time by comparing its growth in the presence and absence of N. trichotoma. The interactions between N. trichotoma and neighbours were analysed for each individual species.

The competitive effects of the different neighbouring species on N. trichotoma was assessed using the yield-loss function:

graphic file with name M1.gif

where PC0 is the biomass of the N. trichotoma plant in the absence of competition, a is the co-efficient that describes the strength of the competitive interactions, Y is the density or biomass (g) of competing plants of species y and Px is the yield (g) of the N. trichotoma plant competing with species y (Freckleton and Watkinson, 1997).

Yield-loss equations were calculated for plants harvested at 36 weeks (Table 1 and Figs 6 and 7) and were calculated using density and biomass separately.

Table 1.

Competitive effects (a values with standard error) and R2 values of the yield-loss function at different N levels for each neighbouring species (the function was fitted to both biomass and density data)

Regression
 Neighbour species
 N (kg ha−1)
 a
 s.e.
 R2
Density A. racemosa 0 0·87 0·81 0·72
60 0·44 0·20 0·66
120 0·71 0·25 0·79
M. stipoides 0 2·48 1·52 0·78
60 0·41 0·28 0·43
120 3·38 2·40 0·80
T. australis 0 0·26 0·11 0·80
60 0·24 0·14 0·63
120 0·11 0·14 0·26
B. macra 0 1·82 1·30 0·66
60 0·35 0·21 0·71
120 0·14 0·04 0·79
Biomass A. racemosa 0 4·54 7·85 0·72
60 0·24 0·10 0·59
120 0·32 0·13 0·69
M. stipoides 0 3·38 2·21 0·79
60 0·38 0·18 0·70
120 0·53 0·33 0·82
T. australis 0 0·53 0·21 0·85
60 0·13 0·07 0·49
120 0·04 0·06 0·18
B. macra 0 1·59 0·65 0·92
60 0·31 0·15 0·75
120 0·12 0·04 0·64
Open in a new tab

Results

Comparison of species without competition

Harvest time had the largest influence on plant biomass (Fig. 1). At 12 weeks there was no significant difference in biomass between species at any N level. For plants harvested at 36 weeks, however, there was a significant difference in biomass among species (P < 0·001, d.f. = 5, F = 9·20) and biomass also increased with increasing N level (P < 0·001, d.f. = 2, F = 85·65). When averaged over the three nitrogen treatments, the rank order was (species sharing a superscript are not significantly different): N. trichotomaa (planted in winter) > M. stipoidesab > T. australisab > A. racemosab > N. trichotomab (planted in summer) > B. macrac.

Fig. 1.

Fig. 1.

Open in a new tab

Biomass (mean ± standard error) of single plants of each species grown without competition at different N levels (0, 60 or 120 kg ha−1) for each harvest (12 or 36 weeks). Nassella trichotoma, A. racemosa and M. stipoides were sown in winter and N. trichotoma, B. macra and T. australis were grown in summer.

The competitive response of N. trichotoma to neighbours

When harvested at 12 weeks, the biomass of N. trichotoma was always reduced (P < 0·05, d.f. = 2, F = 6·07 A. racemosa; 4·71 M. stipoides; 7·87 T. australis; 36·22 B. macra) in the presence of eight neighbouring plants irrespective of species (Fig. 2). Further, A. racemosa also significantly reduced the biomass of N. trichotoma with only 1 neighbouring plant (P < 0·01). On average, N level did not affect N. trichotoma biomass in the presence of neighbouring plants when harvested at 12 weeks.

Fig. 2.

Fig. 2.

Open in a new tab

Biomass (mean ± standard error) of N. trichotoma grown in the presence of each neighbouring species at different densities (zero, one or eight plants per pot) and N levels (0, 60 or 120 kg ha−1) when harvested at 12 weeks.

When harvested at 36 weeks, all four species caused a significant reduction (P < 0·01, d.f. = 2, F = 49·12 A. racemosa; 44·92 M. stipoides; 57·92 T. australis; 15·18 B. macra) in N. trichotoma biomass at a density of eight neighbouring plants (Fig. 3). Microlaena stipoides and B. macra also reduced (P < 0·001) the biomass of N. trichotoma at a neighbouring density of one plant. On average N. trichotoma biomass increased (P < 0·01, d.f. = 2, F = 11·83 A. racemosa; 10·14 M. stipoides; 10·26 T. australis; 27·96 B. macra) with increasing N application (Fig. 3).

Fig. 3.

Fig. 3.

Open in a new tab

Biomass (mean ± standard error) of N. trichotoma grown in the presence of each neighbouring species at different neighbouring plant densities (zero, one or eight plants per pot) and N levels (0, 60 or 120 kg ha−1) when harvested at 36 weeks.

The competitive effect of N. trichotoma on neighbours

The single ‘target’ N. trichotoma plant had, on some occasions, a competitive effect on neighbouring plants. At the 12-week harvest, N. trichotoma had no effect on any species at any N level (Fig. 4). The only significant effect observed was the decreased per plant biomass of A. racemosa and M. stipoides at densities of eight plants per pot.

Fig. 4.

Fig. 4.

Open in a new tab

Biomass (mean ± standard error) of neighbouring plants grown at different densities, with or without N. trichotoma, at different N levels (0, 60 or 120 kg ha−1), when harvested at 12 weeks.

For single-neighbour plants harvested at 36 weeks, the presence of N. trichotoma reduced (P < 0·01, d.f. = 1, F = 65·60 A. racemosa; 10·37 M. stipoides; 127·31 T. australis; 59·10 B. macra) plant biomass for neighbouring species (Fig. 5). The average per plant biomass for neighbouring species was always lower (P < 0·05, d.f. = 1, F = 59·37 A. racemosa; 4·84 M. stipoides; 125·75 T. australis; 59·10 B. macra) at a density of eight plants compared with one.

Fig. 5.

Fig. 5.

Open in a new tab

Biomass (mean ± standard error) of neighbouring plants grown at different densities, with or without N. trichotoma, at different N levels (0, 60 and 120 kg ha−1), when harvested at 36 weeks.

The biomass of individual neighbouring plants increased (P < 0·001, d.f. = 2, F = 103·95 A. racemosa; 7·65 M. stipoides; 129·13 T. australis; 10·69 B. macra) with increasing N application. However, two species showed an interaction between N application and the presence of N. trichotoma. While in the absence of N. trichotoma the average biomass of A. racemosa and T. australis plants increased with increasing N application rate, in the presence of N. trichotoma the application of 120 kg N ha−1 did not have a significantly greater effect on biomass of those species than application of 60 kg ha−1.

Yield-loss functions

Using density-derived data, the competitive effect of the neighbouring species was greater at low soil N levels (Table 1). The strength of the competitive interactions, measured by the mean a value (eqn 1) with no added N was 1·36, with 60 kg N ha−1 was 0·36 and with 120 kg N ha−1 was 1·09. At each level of soil N, similar rankings of species competitive effects were observed for either density or biomass-derived data: Pearson's rank correlation coefficients determined for pairwise comparisons of the rankings from each of the three N treatments averaged 0·60 for density-derived data and 0·67 for biomass-derived data. Averaged over the three N treatments, the rank order was the same using either density or biomass: M. stipoides > A. racemosa > B. macra > T. australis (Figs 6 and 7).

Fig. 6.

Fig. 6.

Open in a new tab

Fitted curves for the yield-loss equation using biomass as the independent variable, at different N levels (0, 60 or 120 kg ha−1). Nassella trichotoma was the target species grown with a single neighbouring species: T. australis, B. macra, A. racemosa or M. stipoides. The a values in Table 1 describe the curvature of the lines.

Fig. 7.

Fig. 7.

Open in a new tab

Fitted curves for the yield-loss equation using density as the independent variable, at different N levels (0, 60 or 120 kg ha−1). Nassella trichotoma was the target species grown with a single neighbouring species: T. australis, B. macra, A. racemosa or M. stipoides. The a values in Table 1 describe the curvature of the lines.

DISCUSSION

In its native range, N. trichotoma is widely distributed throughout the temperate Pampa of Argentina in regions ranging from semi-arid to sub-humid (Connor, 1960). It is not restricted by soil type, and soil fertility is generally high throughout its range. In Australia the distribution of N. trichotoma is likewise not strongly associated with any single environmental factor (Campbell and Vere, 1995). Rather, the invasion of N. trichotoma is often associated with soil disturbance (Healy, 1945; Wells, 1974). The rapid spread of N. trichotoma in Australian soils of low fertility (Jones and Vere, 1998) may be explained, in part, by increases in available soil nutrients as a consequence of soil disturbance (Vitousek and Walker, 1987). In addition, a reduction in native perennial grass density may provide increased opportunities for N. trichotoma to invade. This study provides support for those mechanisms.

Native Australian perennial grasses from the temperate region do not, in general, respond to increased soil fertility to the same extent as many introduced perennial grasses. This has been documented for T. australis (Fisher, 1974; Groves et al., 2003), Austrodanthonia spp. (Navas et al., 2002; Groves et al., 2003) and B. macra (Cook et al., 1976). However, in this experiment, at the highest fertility level, M. stipoides exceeded the biomass of N. trichotoma when grown in a 1 : 1 ratio. Thus, while the findings of previous work were supported for three native grass species, those studies had not included the rhizomatous M. stipoides. The biomass of N. trichotoma was significantly reduced by the presence of M. stipoides under these conditions. When the native grass to N. trichotoma density ratio was increased to 8 : 1, the total biomass of the native grasses exceeded that of N. trichotoma and severe effects on growth of N. trichotoma were recorded. This occurred at all fertility levels.

Yield-loss function analysis showed that the competitiveness of native grass species was least at the highest soil N level, with the exception of M. stipoides. At each fertility level, the rankings of competitiveness were consistent, with M. stipoides almost always the best competitor followed by A. racemosa then B. macra, with T. australis being the weakest competitor. The suggestion that C4 grasses are likely to be more competitive due to a greater ability to use and reduce soil N to a lower level of availability than C3 perennial grasses (Wedin and Tilman, 1996; Wedin, 1999) was not supported in this study. In fact, C4 grasses were almost always the weakest competitor irrespective of soil fertility. This finding contradicts suggestions from field observations (e.g. Hocking, 1998; Michalk et al., 1999) that T. australis competes strongly with N. trichotoma.

There could be a number of reasons why T. australis was a poorer competitor compared with field observations. Several mechanisms other than competition for N can affect species composition in the field. The population dynamics that enable T. australis to limit N. trichotoma invasion may not have applied in this experiment, since adult T. australis plants may compete well for N with N. trichotoma seedlings. Field observations may also have come from areas of established T. australis that had resisted invasion from N. trichotoma rather than been the outcome of competition between the species. The hypothesis that grasslands with a high density of established native grasses can resist invasion from N. trichotoma was supported in a section of this experiment that was not reported here. Where N. trichotoma seedlings were sown amongst adult native grass plants, survival was so low that the results could not be analysed.

Nassella trichotoma is an unpalatable species due to a high density of indigestible tissue in the leaves (Campbell, 1965). It has been proposed that species with high tissue density grow at a slower rate than those with lower tissue density (Grime, 1979; Craine et al., 2001; Craine, 2003), and they appear to be better adapted to low N soils. Species with high tissue density may compete successfully at low soil fertility because they accumulate biomass over a number of years, retaining limiting nutrients that species with low tissue density recycle back into the soil (Aerts and van der Peijl, 1993), as plant material takes longer to break down. Nassella trichotoma has relatively high tissue densities (digestibility 17–25 %; McManus et al., 1972), whereas the four native species have somewhat lower tissue densities (digestibility T. australis 54–75 %, B. macra 48–59 %, Austrodanthonia spp. 45–75 % and M. stipoides 55–80 %; Lodge and Whalley, 1983; Robinson and Archer, 1988). In this experiment, however, N. trichotoma was able to grow as quickly as or faster than the native grasses when there were no, or low density of, competing plants. The ability to grow quickly through the early stages of growth even at low soil fertility and be able to monopolize nutrients over a much longer period, perhaps explains partly why it is such a successful weed in temperate Australian native grasslands. This study has not supported the view that species of high tissue density are inherently less competitive.

During this experiment, plants harvested at 12 weeks showed very little inter- or intra-specific competition. The absence of a response to N indicates that it was probably not a limiting factor at that early stage of plant development, i.e. initial germination and seedling growth for these species was not dependent upon soil N level.

Competition for soil N will not be the only mechanism that influences N. trichotoma invasion into native grasslands in the field. Other factors such as competition for moisture and light, different seasonal growth patterns, or differential impacts of grazing events can play a role in regulating invasion by N. trichotoma. In this experiment, different starting times were used for each of the C3 and C4 competitor types. Under field conditions, however, it is unlikely that a single germination/recruitment event would occur unless there was severe disturbance. Examining the competitive effect of adult native perennial grasses on N. trichotoma seedlings in the field may provide further insight into competitive interactions. Once established, adult N. trichotoma plants have an extremely low natural mortality (Badgery, 2004), which indicates that mortality from the competitive effect of native grasses is probably minimal. Further experiments to determine the relative importance of these mechanisms in the field are certainly warranted.

Growing conditions

The experiment was done in a temperature-controlled glasshouse but there were different growing conditions for the two starting times (driven by temperatures varying with external conditions), highlighted by different growth patterns for individual N. trichotoma plants. Plants that were sown in winter grew slowly at first and had a lower biomass at 12 weeks compared with those that were sown in summer. Mean daily temperature across the experimental period fluctuated from 11 to 24 °C from winter to summer. Between 12 and 36 weeks the temperature for growth increased in the winter-planted treatments and decreased for the summer-planted treatments, resulting in larger N. trichotoma plants when planted in winter than those planted in summer. This was observed in the field by Badgery (2004) who found that autumn-germinating N. trichotoma plants had a higher chance of surviving then spring germinating seedlings because the autumn seedlings were larger when the period of maximum stress occurred in summer.

Method of determining competition coefficients

In the analysis of this experiment both density and biomass of neighbouring plants were used as independent factors in the yield-loss function to assess the impact of competition on N. trichotoma. Neighbouring species biomass is nonlinearly related to the density and is not generally used to predict competitive effects because of problems with independence (Freckleton and Watkinson, 2000). In this experiment, however, M. stipoides, a rhizomatous species, was found to show a different pattern of competition at high N levels. Microlaena stipoides had vegetative tiller recruitment from rhizomes within the experiment and one plant grew as much biomass as eight neighbouring plants when sown with N. trichotoma. This did not occur with any other species. As a consequence, consideration of biomass provided more insight than density for M. stipoides. It can be concluded that neighbouring species biomass may be better for calculating yield-loss competition coefficients for rhizomatous species than plant density, notwithstanding the assumptions about independence.

Practical outcomes

The more important outcome of this experiment is evidence that high native perennial grass density and/or biomass are essential for these native grasses to compete with N. trichotoma. This study has clarified that, as individual plants of three of these native grass species are not very competitive with N. trichotoma, a high density of perennial grass will be critical to provide sufficient competition against, and prevent invasion by, N. trichotoma, especially on more fertile soils. Higher soil fertility and low density of perennial grass are likely to favour N. trichotoma. One exception to this pattern found in this study was M. stipoides. It was competitive at low densities and higher fertility, in part because of its rhizomatous growth habit.

Current recommendations to farmers for N. trichotoma control commonly include the advice to apply fertilizer in the belief that this will make native perennial grass species more competitive. Such recommendations need to be qualified. To maximize competition of native grasses with N. trichotoma, management should focus on promoting a high density of actively growing native perennial grasses, which may be achieved by carefully managing grazing intensity and duration (e.g. Garden et al., 2000, 2003).

Acknowledgments

This research was funded by the Cooperative Research Centre for Weed Management Systems. We thank Jenni Tarleton, Geoff Wilson, Geoff Millar, David Pickering, Jenny Wickham and Karen Gogala for their valuable technical support, as well as Helen Nichol for her opportune statistical advice.

LITERATURE CITED

  1. Aerts R, van der Peijl MJ. 1993. A simple model to explain the dominance of low-productive perennials in nutrient-poor habitats. Oikos 66: 144–147. [Google Scholar]
  2. Auld BA, Coote BG. 1981. Prediction of pasture invasion by Nassella trichotoma (Gramineae) in south east Australia. Protection Ecology 3: 271–277. [Google Scholar]
  3. Badgery WB. 2004.Managing competition between Nassella trichotoma (serrated tussock) and native grasses. PhD Thesis, University of Sydney. [Google Scholar]
  4. Badgery WB, Kemp DR, Michalk DL, King WM. 2002. Competition between native grasses and serrated tussock (Nassella trichotoma) at low fertility—initial results. Plant Protection Quarterly 17: 95–98. [Google Scholar]
  5. Campbell MH. 1965.Investigations for the control of serrated tussock (Nassella trichotoma (Nees.) Hack.), on non-arable land. Masters Thesis, University of Sydney. [Google Scholar]
  6. Campbell MH. 1977. Serrated tussock. Part 1. Life history, identification. Division of Plant Industry Bulletin. Bathurst, NSW: Department of Agriculture, Agricultural Research Station. [Google Scholar]
  7. Campbell MH. 1998. Biological and ecological impact of serrated tussock (Nassella trichotoma (Nees.) Arech.) on pasture in Australia. Plant Protection Quarterly 13: 80–86. [Google Scholar]
  8. Campbell MH, Irvine JH. 1966. Block supplementation of sheep grazing a serrated tussock (Nassella trichotoma)—sown pasture association. Agricultural Gazette of New South Wales 77: 564–571. [Google Scholar]
  9. Campbell MH, Vere DT. 1995.Nassella trichotoma (Nees) Arech. In: Groves RH, Shepherd RCH, Richardson RG, eds. The biology of Australian weeds, Vol. 1. Melbourne: RG and FJ Richardson, 189–202. [Google Scholar]
  10. Connor HE. 1960. Nassella tussock in Argentina. New Zealand Journal of Agriculture 100: 18–21. [Google Scholar]
  11. Cook SJ, Lazenby A, Blair GJ. 1976. Comparative responses of Lolium perenne and Bothriochloa macra to temperature, moisture, fertility and defoliation. Australian Journal of Agricultural Research 27: 769–778. [Google Scholar]
  12. Craine J. 2003. The role of nitrogen in grasslands: from ecophysiology to ecosystems and competition to herbivory. In: Allsopp N, Palmer AR, Milton SJ, Kirkman KP, Kerley GIH, Hurt CR, Brown CJ, eds. Proceeding of the VIIth International Rangelands Congress, Durban, 6–13. Irene, South Africa: Document Transformation Technologies. [Google Scholar]
  13. Craine JM, Froehle J, Tilman DG, Wedin DA, Chapin III FS. 2001. The relationships among root and leaf traits of 76 grassland species and relative abundance along fertility and disturbance gradients. Oikos 93: 274–285. [Google Scholar]
  14. Fisher HJ. 1974. Effect of nitrogen fertilizer on a kangaroo grass (Themeda australis) grassland. Australian Journal of Experimental Agriculture and Animal Husbandry 14: 526–532. [Google Scholar]
  15. Freckleton RP, Watkinson AR. 1997. Measuring plant neighbour effects. Functional Ecology 11: 532–534. [Google Scholar]
  16. Freckleton RP, Watkinson AR. 2000. Designs for greenhouse studies of interactions between plants: an analytical perspective. Journal of Ecology 88: 386–391. [Google Scholar]
  17. Garden DL, Bolger TP. 2001. Vegetation changes in southeastern Australian temperate grasslands. In: Gomide JA, Mattos WRS, da Silva SC, eds. Proceedings of the XIX International Grassland Congress, 905–906. Piracicaba, Brazil: Fundação de Estudos Agrários Luiz de Queiroz (FEALQ). [Google Scholar]
  18. Garden DL, Ellis NJS, Rab MA, Langford CM, Johnston WH, Shields C, et al. 2003. Fertiliser and grazing effects on production and botanical composition of native grasslands in south-east Australia. Australian Journal of Experimental Agriculture 43: 843–859. [Google Scholar]
  19. Garden DL, Lodge GM, Friend DA, Dowling PM, Orchard BA. 2000. Effects of grazing management on botanical composition of native grass-based pastures in temperate south-east Australia. Australian Journal of Experimental Agriculture 40: 225–245. [Google Scholar]
  20. Grime JP. 1979.Plant strategies and vegetation processes. Chichester: John Wiley and Sons. [Google Scholar]
  21. Groves RH, Austin MP, Kaye PE. 2003. Competition between Australian native and introduced grasses along a nutrient gradient. Austral Ecology 28: 491–499. [Google Scholar]
  22. Healy AJ. 1945.Nassella tussock. Field studies and their agricultural significance. Department of Scientific and Industrial Research New Zealand 91. Wellington: Ferguson & Osborn. [Google Scholar]
  23. Hocking C. 1998. Land management of Nassella areas—implications for conservation areas. Plant Protection Quarterly 13: 86–91. [Google Scholar]
  24. Isbell RF. 1996.The Australian soil classification. Collingwood: CSIRO Publishing. [Google Scholar]
  25. Jones RE, Vere DT. 1998. The economics of serrated tussock in New South Wales. Plant Protection Quarterly 13: 70–76. [Google Scholar]
  26. Jones RE, Vere DT, Campbell MH. 2000. The external costs of pasture weed spread: an economic assessment of serrated tussock control. Agricultural Economics 22: 91–103. [Google Scholar]
  27. Lodge GM, Whalley RDB. 1983. Seasonal variations in the herbage mass, crude protein and in-vitro digestibility of native perennial grasses on the north-west slopes of New South Wales. Australian Rangeland Journal 5: 20–27. [Google Scholar]
  28. McLaren DA, Stajsic V, Gardener MR. 1998. The distribution and impact of South/North American stipoid grasses (Poaceae: Stipeae) in Australia. Plant Protection Quarterly 13: 62–70. [Google Scholar]
  29. McManus WR, Manta L, McFarlane JD, Gray AC. 1972. The effects of diet supplements and gamma irradiation on dissimilation of low-quality roughages by ruminants. 1. Studies on the terylene-bag technique and affects on supplementation of base ration. Journal of Agricultural Science 79: 27–40. [Google Scholar]
  30. Michalk D, Kemp DR, Campbell MH, McLaren DA. 1999. Control of serrated tussock—problems in developing IWM systems. In: Bishop AC, Boersma M, Barnes CD, eds. Proceeding of the Twelfth Australian Weeds Conference, 20–24. Davenport, Tasmania: Tasmanian Weeds Society. [Google Scholar]
  31. Moore RM. 1970. South-eastern temperate woodlands and grasslands. In: Moore RM ed. Australian grasslands. Canberra: Halstead Press, 169–190. [Google Scholar]
  32. Mott JJ, Ludlow MM, Richards JH, Parsons AD. 1992. Effects of moisture supply in the dry season and subsequent defoliation on persistence of the savannah grasses Themeda triandra, Heteropogon contortus and Panicum maximum Australian Journal of Agricultural Research 43: 241–260. [Google Scholar]
  33. Navas ML, Garnier E, Austin MP, Viaud A, Gifford RM. 2002. Seeking a sound index of competitive intensity: application to the study of biomass production under elevated CO2 along a nitrogen gradient. Austral Ecology 27: 463–473. [Google Scholar]
  34. Payne RW, Baird DB, Cherry M, Gilmour AR, Harding SA, Kane AF, et al. 2003.GenStat®for Windows, 7th edn. Oxford: VSN International. [Google Scholar]
  35. Ridley HN. 1930.The dispersal of plants throughout the world. Kent: L. Reeve and Co. [Google Scholar]
  36. Robinson GG, Archer KA. 1988. Agronomic potential of native grass species on the Northern Tablelands of New South Wales. I. Growth and herbage production. Australian Journal of Agricultural Research 39: 415–423. [Google Scholar]
  37. Vitousek PM, Walker LR. 1987. Ecosystem level interactions. In: Gray AJ, Crawley MJ, Edwards PJ, eds. Colonization, succession and stability. Oxford: Blackwell Scientific. [Google Scholar]
  38. Wedin DA. 1999. Nitrogen availability, plant-soil feedbacks and grassland stability. In: Eldridge D, Freudenberger D, eds. Proceedings of the VI International Rangeland Congress, 193–197. Aithkenvale, Australia: VI International Rangeland Congress Inc. [Google Scholar]
  39. Wedin DA, Tilman D. 1996. Influence of nitrogen loading and species composition on the carbon balance of grasslands. Science 274: 1720–1723. [DOI] [PubMed] [Google Scholar]
  40. Wells MJ. 1974.Nassella trichotoma (Nees) Hack. in South Africa. Proceedings of 1st South African Weeds Conference. [Google Scholar]
  41. Westbrooks RG, Cross G. 1993. Serrated tussock (Nassella trichotoma) in the United States. Weed Technology 7: 525–528. [Google Scholar]

Articles from Annals of Botany are provided here courtesy of Oxford University Press

RESOURCES