Search

Patch formation is common in grazed grasslands but the mechanisms involved in the formation and maintenance of patches are not clear. To increase our knowledge on this subject we examined possible reasons for patch formation and the influence of management on changes between patch states in three experiments in native pasture communities in the Crows Nest district, south-east Queensland. In these communities, small-scale patches (tall grassland (dominated by large and medium tussock grasses), short swards (dominated by short tussock grasses and sedges), and lawns (dominated by stoloniferous and/or rhizomatous grasses)) are readily apparent. We hypothesized that the formation of short sward and lawn patches in areas of tall grassland was due to combinations of grazing and soil fertility effects. This was tested in Experiment 1 by applying a factorial combination of defoliation, nutrient application and transplants of short tussock and stoloniferous species to a uniform area of tall grassland. Total species density declined during the experiment, was lower with high nutrient applications, but was not affected by defoliation. There were significant changes in abundance of species that provided support for our hypotheses. With light defoliation and low nutrients, the tall grassland remained dominated by large tussock grasses and contained considerable amounts of forbs. With heavy defoliation, the pastures were dominated by medium tussock grasses and there were significant decreases in forbs and increases in sedges (mainly with low nutrients) and stoloniferous grasses (mainly with high nutrients). Total germinable seed densities and those of most species groups were significantly lower in the heavy defoliation than the light defoliation plots. Total soil seed numbers were not affected by nutrient application but there were fewer seeds of the erect forbs and more sedge seeds in plots with high nutrients. The use of resting from grazing and fire to manage transitions between patches was tested. In Experiment 2, changes in species density and abundance were measured for 5 years in the three patch types with and without grazing. Experiment 3 examined the effects of fire, grazing and resting on short sward patches over 4 years. In Experiment 2, total species density was lower in lawn than short sward or tall grassland patches, and there were more species of erect forbs than other plant groups in all patch types. The lawn patches were originally dominated by Cynodon spp. This dominance continued with grazing but in ungrazed patches the abundance of Cynodon spp. declined and that of forbs increased. In the short sward patches, dominance of short tussock grasses continued with grazing but in ungrazed plots their abundance declined while that of large tussock grasses increased. The tall grassland patches remained dominated by large and medium tussock species. In Experiment 3, fire had no effect on species abundance. On the grazed plots the short tussock grasses remained dominant but where the plots were rested from grazing the small tussock grasses declined and the large tussock grasses increased in abundance. The slow and relatively small changes in these experiments over 4 or 5 years showed how stable the composition of these pastures is, and that rapid changes between patch types are unlikely.

The land and pasture development practices of 375 northern Australian beef properties in 8 regions were surveyed during 1996–97. These properties represented a broad cross-section of the beef industry in terms of geographical location, enterprise and herd size, and ownership structures. Both tree clearing and killing were more common in Queensland than in the Northern Territory or northern Western Australia. In all regions where trees were poisoned, native pasture was more widely used than sowing introduced grass and/or legume species. In contrast, tree clearing was most often accompanied by sowing pastures (either an introduced grass only or introduced grass and legume species together), rather than using native pastures. Central coastal Queensland had the highest use of poisoning trees for pasture development. Tree clearing and using native pasture was most important in central Queensland regions and the Maranoa South West. Sowing introduced pasture species under live trees was more commonly practiced in northern Queensland, the Northern Territory and northern Western Australia than in other regions. A considerable number of introduced grass and legume species were sown by producers. Most of the sown species were grasses. Many of the sown grass and legume species were spreading naturally. Buffel grass was spreading in all areas with < 1000 mm average annual rainfall, but most sown species were spreading only in wetter regions. Stylosanthes spp. were the most commonly spreading legume species in regions with > 500 mm average annual rainfall. The results are discussed in relation to contemporary natural resource management issues and how this may affect land and pasture development activities in the future.

Breeding herd performance and husbandry were surveyed on 375 northern Australian beef properties in 8 regions during 1996 and 1997. Mean branding rates ranged from 62.6% in the Northern Territory to 77.1% in the Maranoa South West regions of Queensland with considerable variation within regional herds. The proportion of herds using controlled mating was highest in the Central Coastal, Central Highlands and Maranoa South West. Mean mating period across all regions varied from 5.6 to 11.8 months. Calving was seasonal with peak activity in the August–December period. Calving commenced earlier in the south (August) than it did in northern regions (September–November). Rainfall influenced the timing of commencement of calving and peak calving activity. The use of pregnancy testing was widespread but selective and often not all females were pregnancy tested. About 97% of properties used weaning strategies with a peak in calf weaning in April–July, and a minor peak in September–October. Although mean weaner ages were similar across regions (5.9–6.9 months), mean weaner liveweight varied markedly with weaners in the more northern regions being lightest (200 kg. Culling criteria for cows and heifers focused on temperament, conformation and reproductive failure; age was also important for cows with a common culling age of 10 years. Bulls were more commonly culled at 7–8 years of age. In addition to age, bulls were culled for physical defects, reproductive problems, temperament and poor quality/performing calves. Four to 5 criteria were commonly used to select bulls. Structural soundness and temperament ranked highest followed by conformation, weight for age, Breedplan and colour. Producers using Breedplan tended to use structural soundness and temperament also as selection criteria. The results suggest that producers associated increasing turn-off weight or decreasing turn-off age more with pasture improvement than with bulls of higher genetic merit for growth. Supplementation of the breeding herd increased in the last half of the calendar year. Supplements containing molasses, urea, phosphate source, salt, and sulphate of ammonia were the most commonly supplied supplementary nutrients. Vaccination for botulism was quite common (>30% of regional survey groups) in all regions except Central Coastal and Maranoa South West regions of Queensland. Comparisons with previous surveys show that there have been marked improvements in breeding herd performance and management over an extended period. The results are discussed in relation to scientific developments over the last 2 decades and future development of the northern Australian beef industry.

Natural resource management practices were surveyed on 375 northern Australian beef properties in 8 regions during 1996–97. These producers represented a broad cross section of the beef industry in terms of geographical location, size of enterprise, and herd and ownership structures. More than 48% of producers reported land degradation (erosion, salinity, weeds, etc.) and >68% reported woody weeds on their properties. Eucalypt (Eucalyptus and Corymbia spp.) and wattle (Acacia spp.) regrowth were universally common native woody weeds. Parkinsonia (Parkinsonia aculeata) was widespread across all survey regions occurring with the greatest frequency in northern regions. A range of native and exotic woody weed species were regionally significant. Brigalow (Acacia harpophylla) regrowth was mainly reported in the central Queensland and Maranoa South West regions. Prickly acacia (Acacia nilotica) was mainly reported in central and north-west Queensland. Rubber vine (Cryptostegia grandiflora) was mainly reported in northern Queensland with a significant presence in central coastal and north-west Queensland. Infestations with multiple weed species were common. Although many producers reported the presence of problem woody weeds, only 3–29% indicated that weed management was practised. Most producers (67–100%) used fire. The intended purpose of fire varied across survey regions but fire was commonly used to reduce rank material, reduce fire risk, control woody weeds and for grazing management. Seasonal conditions and variability restricted fire use. Assessment by visual observation was a commonly used practice for determining stocking rate. Between 4–20% of producers were using visual observation combined with some form of calculation of stocking rate at the end of the growing season to determine stocking rate. A majority (>67%) of producers used some form of paddock spelling or regeneration practice. Many producers (47–89%) considered that they used conservative stocking rates. Carrying sufficient stock to meet income requirements was significantly (P66% of producers had increased the number of stock watering points, and more than 85% of producers had erected fencing. Most fencing was for new paddocks and laneways, or to replace old fences, and was not associated with new stock watering points. However, the installation of new stock watering points was associated with the creation of new paddocks. Planned future fencing focused on creating new paddocks or laneways or replacing old fences. The results are discussed in relation to contemporary natural resource management issues and the future approaches to and challenges of implementing sustainable rangeland management.

The soil seedbanks in three patch types (tall grassland, short sward and lawn) were measured in native pastures near Crows Nest, south-east Queensland in two experiments. In the first experiment, seedbanks were measured at four sites, and in the second, the variation in seedbanks during the year was measured at one site. In each experiment the size and composition of the seedbanks were determined by germinating the seeds in soil samples. In the first experiment, total seed numbers ranged from ~6000 to 12,000 per m2. There were no significant differences between sites for total seeds or for seeds of any species group except legumes which comprised only a small portion of the seedbanks. There were significant differences between patch types for total seeds and for the following species groups, medium tussock grasses, short tussock grasses, stoloniferous grasses, legumes and forbs but not large tussock grasses and sedges. Total seed numbers, and those of medium tussock grasses, stoloniferous grasses and forbs were highest in the lawn patches and lowest in the short sward patches. Legumes had higher numbers in the tall grassland patches than in other patch types. In the second experiment, there were large differences between total seed numbers at the different sample dates (January, May, September and November). Numbers were highest in January and then declined throughout the year. This pattern was largely a reflection of the changes in numbers of forb seeds, the species group with the most seeds. There were no significant differences between patch types for total seeds but there were for medium tussock grasses, stoloniferous grasses, sedges and forbs. There was no relationship between the composition of the pasture sward and the composition of the seedbank for any of the three patch types. Twenty-eight species were allocated to persistent and transient seedbank types; all seedbank types occurred in all three patch types. The major species in the seedbanks were sedges (Cyperus gracilis, C. sesquiflorus), forbs (Gamochaeta spp., Paronychia brasiliana, Daucus glochidiatus) and Eragrostis spp.