Abstract

Keywords

Introduction

Short and long term pasture phases are needed in southern Queensland for rejuvenating degraded cropping soils by rebuilding soil nitrogen levels, increasing soil organic matter and restoring soil structure and infiltration. Pastures also provide greater cover than annual cropping systems thus reducing run-off and consequent erosion. Profitable pasture phases that contribute to sustaining the natural soil resources are dependent on success during their establishment, production and persistence. This study assessed the relative performance (establishment, cover, biomass production and persistence) of several commercially available tropical pasture species to quantify further their role in crop-pasture rotations in mixed farming systems in southern Queensland. Owens et al. (2008) report on water use efficiency for 4 of the species (Rhodes grass, buffel grass, butterfly pea and annual lablab) was also measured within this trial.

Methods

Experimental design.

The trial ran over three growing seasons from Jan 2006 to Feb 2008 on a brown Vertosol soil at the Roma research station in southern Queensland (26.58^oS, 148.76^oE, elevation: 303 m). Nine tropical grasses and 6 tropical legumes were sown on the 31^st of January 2006 in 3 replicate plots of 5 x 20 m at the sowing rates outlined in Table 1. Recommended sowing rates (Johnson et al. 2006) were adjusted according to germination test results (e.g. lower for stylo) and to allow for the weight of the seed coating (e.g. buffel grass). The legume species were planted to moisture (3-6 cm) using 20 cm row spacing, with the exception of the two lablab treatments which were planted using 50 cm row spacing. All the grasses were broadcast and lightly harrowed. Irrigation was used to assist establishment and strategic irrigations were intended to normalise the season to an average rainfall year if required but this was prevented due to a limitation on available irrigation water. The whole trial was cut mechanically at the end of each growing season and the biomass removed to simulate grazing.

Table 1. The species and cultivars of the 9 grasses and 6 legumes used in the trial and the planting rates, determined by previous germination test results.

Common name/cultivar	Scientific name	Sowing rate (kg/ha)	Germination rate (%)	Germinable seed/m2
Buffel grass	Cenchrus ciliaris	16	not tested
Creeping bluegrass cv. Bisset	Bothriochloa inscupta	8	20	120
Gatton panic	Panicum maximum	8	32	340
Rhodes grass cv. Fine cut	Chloris gayana	8	16	430
Digit grass cv. Premier	Digitaria eriantha	8	30	600
Silk sorghum	Sorghum spp. hybrid	8	56	67
Floren bluegrass	Dicanthium aristatum	8	34	220
Purple pigeon grass	Setaria incrassata	8	not tested
Bambatsi panic	Panicum coloratum	8	34	450
Burgundy bean	Macroptilium bracteatum	5	70	21
Perennial lablab cv. Endurance	Lablab purpureus	12	90	5.4
Butterfly pea cv. Milgarra	Clitoria ternatea	8	80	15
Desmanthus cv. Marc	Desmanthus virgatus	3	30	22
Annual lablab cv. Highworth	Lablab purpureus	15	100	6
Caatinga stylo	Stylosanthes seabrana	12	35	136

Measurements

Plant counts, a visual estimate of groundcover and biomass cuts were taken every 4-6 weeks over the three growing seasons from three quadrants (0.25 m²) placed along a stratified transect diagonally across each plot. All biomass was removed to a height of 2 cm above the soil surface and subsequently dried at 80⁰C and weighed.

Results

In year 1 a full soil profile at planting and effective rain during early autumn (Feb-March) enabled good production. In the second year extremely dry conditions (decile 2-3 rainfall) greatly reduced plant growth and plant numbers in the drought sensitive species. The summer of year 3 was above average with 455mm falling between Oct and Feb (Table 2).

Table 2. Mean and measured rainfall (mm) over the three growing seasons.

	Oct	Nov	Dec	Jan	Feb	Mar	Apr	Total
Mean	60	64	71	74	83	44	38	435
05/06	98	63	44	2	45	41	46	339
06/07	7	52	65	51	60	30	17	282
07/08	49	167	81	168	68	2	0	457

Grass DM production and persistence.

Silk sorghum and purple pigeon proved most productive in year 1 producing > 5 t DM/ha by the end of autumn (Figure 2). Other grasses were less productive in the first year (<3.5 t DM/ha), with digit and buffel grass slow to establish (Table 3). Floren bluegrass failed to establish in good numbers and significant seed contamination from other species was obvious. Bisset bluegrass was also slow to establish but produced the most growth during autumn in year 1 as previously found by Johnson et al. (2006). Bambatsi panic, buffel grass, and Gatton panic persisted well through the low rainfall of year 2 and were most productive in year 3 (>4 t DM/ha). Silk sorghum numbers were greatly reduced after the first year (as expected) and Rhodes grass showed poor persistence and production after the dry second year due to its greater drought sensitivity than other grasses (Figure 2, Table 3). Silk sorghum may have performed better in subsequent years if fertilised with N as shown by Ross (1999). The poor survival of Bisset bluegrass and Floren blue grass was somewhat surprising as they are believed to be a persistent grass but their lower suitably to this soil type may have reduced persistence.

Figure 2. Annual dry matter production of the 9 tropical grasses and 6 tropical legumes over three growing seasons between Jan 05 and Feb 08 at Roma research station. NP – Not planted

Table 3. Average % grass cover at selected summer dates during the three growing seasons.

Species	7-Mar-06	15-Mar-07	30-Jan-08
Buffel grass	17	32	88
Bambatsi panic	83	69	80
Digit grass	33	45	78
Purple pigeon grass	77	42	62
Gatton panic	63	35	60
Rhodes grass	43	49	48
Silk sorghum	90	9	26
Creeping bluegrass	8	85	0
Floren bluegrass	50	38	0

Legume DM production and persistence.

Annual lablab, perennial lablab and burgundy bean proved most productive in year 1 producing > 2 t DM/ha by the end of autumn (Figure 2). Perennial lablab and burgundy bean were again the most productive in year 2 but burgundy bean did not recruit seedlings in year 3 presumably due to the low rainfall in year 2 which reduced seed set (Table 4). Annual lablab was resown in the second season (1-Nov 06) but not the third. Butterfly pea persisted well but production was lower than the best legume in all years. Stylo and desmanthus were slow to establish but regeneration and DM production increased in year 3 (Fig 2, Table 4).

Table 4. Average legume plant numbers/m² at selected summer dates during the three growing seasons.

Species	23-Mar-06	15-Mar-07	5-Dec-07
Caatinga stylo	3.7	3.5	62.6
Desmanthus	1.7	50	36
Butterfly pea	8.8	21	14.6
Burgundy bean	7.5	55	0
Perennial lablab	3.5	2.6	0
Annual lablab	4	5.3	NP

NP – not planted

Conclusion

Gatton panic, Bambatsi panic and buffel grass were the most persistent, productive and most suitable for longer term leys (3+ years). Caatinga stylo and desmanthus showed that they are the best accompanying legumes for longer pasture sequences. Annual lablab, perennial lablab and burgundy bean appeared best suited to shorter term ley pasture rotations (1-2 years). Silk sorghum was highly productive in year 1, but in order to maintain productivity would have to be managed intensively by grazing and fertilisation.

References

Johnson B, Lloyd D, O’Brien S, Williams R and Boschma S (2006). Leygrain – Ley Pastures in the Northern Grainbelt. p 60, DPI&F Queeensland and GRDC February 2006.

Owens J, Bell L, Rodriguez D, Whitbread A, Lawrence J, and Mann M (2008). Comparing the water efficiency of tropical pasture grasses and legumes used in Queensland’s mixed farming systems. In ‘Proceedings of the 14^th Australian Agronomy Conference’, Adelaide, Australian Society of Agronomy.

Ross B (1999) Agnote 784 No: E67 Silk sorghum. www.primaryindustry.nt.gov.au accessed April 2008.

• "Global Issues. Paddock Action." Edited by M. Unkovich. Proceedings of 14th Agronomy Conference 2008, 21-25 September 2008, Adelaide, South Australia.