Fire in Semi-Arid Shrublands and Woodlands: Spatial and Temporal Patterns in an Australian Landscape

Semi-arid landscapes are of interest to fire ecologists because they are generally located in the climatic transition zone between arid lands (where fires tend to be rare due to lack of fuel, but are enhanced following large rainfall episodes) and more mesic regions (where fire activity tends to be enhanced following severe rainfall deficits). Here we report on the characteristics of the contemporary fire regimes operating in a semi-arid region of inland south-western Australia with rainfall averaging around 300 mm per annum. To characterize fire regimes, we analyzed a geodatabase of fire scars (1960–2018) to derive fire preferences for each major vegetation type and fire episode and used known fire intervals to model fire hazard over time and calculate typical fire frequencies. We also used super epoch analysis and correlations to explore relationships between annual fire extent and rainfall received before the fire. We found fires strongly favored sandplain shrublands, and these tended to experience hot crown fires once every 100 years (median fire interval), with fire hazard increasing linearly over time. In contrast, fires were rare in eucalypt woodland and other vegetation types, with a median interval of 870 years and broadly consistent fire hazard over time. Annual fire extent was most strongly linked with high rainfall in the year prior to fire, and this was particularly so for eucalypt woodlands. Large-scale fires in shrublands tended to favor areas burnt in previous large fires, whereas in woodlands they favored edges. In conclusion, we found divergent fire regimes across the major vegetation types of the region. Sandplain shrublands were similar to Mediterranean shrublands in that they experienced intense stand-replacing wildfires which recovered vigorously although slowly, meaning burnt shrublands did not experience fires again for at least 25 and 100 years on average. In contrast, eucalypt woodlands were fire sensitive (trees readily killed by fire) and experienced fires mostly around the edges, spreading into core areas only after large rainfall events elevated fuel levels. Overall, both vegetation types subscribed to typical arid-zone fire regimes where elevated rainfall, and not drought, promoted fires, although the role of fuel accumulation over time was more important in the shrublands.

Grazing pressure and tree competition affect cattle performance and native pastures in Eucalypt woodlands of Queensland, north-eastern Australia

Context Well managed grazing pressure will optimise animal and pasture production, and preserve the soil to maintain a viable beef business on native pastures in eucalypt woodlands. Aims A cattle grazing experiment was established to measure animal and pasture performance under management practices used in the Aristida/Bothriochloa native pastures in central Queensland. Methods Performance of Brahman-cross steers and pastures were measured in an experiment with three grazing pressures by two tree densities in a Eucalyptus populnea woodland in north-eastern Australia over 8 years in paddocks of 4–18 ha. Key results At low grazing pressure with trees killed by herbicide (‘cleared’), stocking rate increased 35% as pasture composition and biomass improved over 8 years. At low grazing pressure where treed, stocking rate remained constant, however, at high grazing pressure where treed, it was reduced after 4 years. The annual liveweight gain increased from 0.37 to 0.45 to 0.51 kg/head.day as grazing pressure was reduced from high to medium to low grazing pressure respectively, and across grazing pressures it decreased from 0.49 where cleared to 0.39 kg/head.day where treed. Liveweight gain per hectare increased under low grazing pressure and declined at medium and high pressures. Body condition scores responded positively to lower grazing pressure and a lack of tree competition to pastures. This treatment combination also produced higher animal sale values. Pasture biomass, basal area and ground cover were all affected negatively by increasing grazing pressure. Conclusions Grazing 25% of autumn pasture improved dry matter production, species composition and land condition, and increased steer growth rates, body condition and their market value. This grazing pressure produced an increasing trend in stocking rates relative to the decline at higher grazing pressures. Higher liveweight gain/ha was produced initially at high grazing pressure (75% utilisation), however, after 4 years animal condition and pastures deteriorated, requiring a reduction in stocking rate to maintain the condition of both the remaining animals and the pastures. Managing tree competition to pastures is necessary to maintain the higher animal production potential. Implications This objective information demonstrates the benefits for cattle, pastures and long-term economic outcomes of managing for conservative grazing pressure and controlling tree competition to pasture in this woodland. Applying these findings can improve beef business outcomes and provide management groups with objective educational resources.

Standing dead trees contribute significantly to carbon budgets in Australian savannas

Previous estimates of greenhouse gas emissions from Australian savanna fires have incorporated on-ground dead wood but ignored standing dead trees. However, research from eucalypt woodlands in southern Queensland has shown that the two pools of dead wood burn at similar rates. New field data from semiarid savannas across northern Australia confirmed that standing dead trees comprise about four times the mass of on-ground dead wood. Further, the proportion of total woody biomass comprising dead wood increases with decreasing fire frequency and a decreasing proportion of late dry season (August to December) fires. This gives scope for increasing the carbon stock in the dead wood pool with a reduced fire frequency. Following a previously published approach to quantify total dead wood loads in savannas, new and previously collected data on tree stand structures were used across the whole savanna zone to quantify dead wood loads in equilibrium with historic fire regimes. New parameters are presented for calculating dead wood dynamics including dead trees in Australia’s savannas.

The legacy of pasture improvement causes recruitment failure in grassy eucalypt woodland conservation reserves in the Midlands of Tasmania

Australia’s most fragmented and least reserved landscapes are the grassy eucalypt woodlands of the south-east. Two hundred years of agricultural disruption have transformed these landscapes, and agricultural enterprises continue to expand and develop, meaning the threats to these landscapes have not abated. The Tasmanian Midlands is primarily privately owned, with very little area devoted to conservation of biodiversity. In this landscape, conservation covenants have been enacted on many private properties with the intention of encouraging tree recruitment and conservation of threatened plant communities and rare species. Evidence of the effectiveness of these covenants in protecting overstorey tree population health is lacking. This study compared the demographic structures of overstorey Eucalyptus species and midstorey tree genera on public and private properties with contrasting land use histories. Reserves on private lands had little tree recruitment, probably because exotic pasture species were common, whereas tree recruitment was abundant in public reserves, where pasture improvement has not occurred. Active measures are needed to restore ecological structure and function in grassy woodland conservation reserves on private land by encouraging regeneration of Eucalyptus and Acacia species as well as returning the ground layer to a functionally native state. This will entail reinstating fire disturbance, reducing exotic pasture species cover and managing domesticated, feral and native herbivores.

Spring fire effects on two Aristida/Bothriochloa native pastures in central Queensland, Australia

Controlled burns are commonly used to suppress woody plant regrowth and to remove accumulated unpalatable pasture from rangelands and occasionally to alter pasture composition in native pastures in central Queensland, Australia. Outcomes can be somewhat unpredictable and short-term, and reliable evidence is needed to confirm the likely long-term efficacy of such fires. We imposed a regime of repeated spring burns on native Aristida/Bothriochloa pastures growing in two contrasting eucalypt woodlands of central Queensland to determine the effects on pasture composition, ground cover, landscape stability and woody plant recruitment, all in the absence of grazing. The sites selected were a silver-leaved ironbark (Eucalyptus melanophloia F.Muell.) woodland and a poplar box (E. populnea F.Muell.) woodland. Weather conditions precluded spring burns in 3 years out of 7 at the silver-leaved ironbark site and in 2 years out of 8 at the poplar box site. The burn intensity was variable, and frequent fires produced a marked change in abundance of only a few pasture species. Depending on the site, fires significantly increased the frequency of Enneapogon spp., Bothriochloa bladhii (Retz.) S.T.Blake and Dichanthium sericeum (R.Br.) A.Camus and reduced the frequency of some minor components such as Cymbopogon spp., Panicum effusum R.Br., Cenchrus ciliaris L. and, ephemerally, that of some forbs. Contrary to expectation, only Aristida calycina R.Br. declined in abundance among the many Aristida species present, and the abundance of Heteropogon contortus (L.) P.Beauv. ex Roem. & Schult. barely increased under regular spring fires. The total germinable seeds of herbaceous species in the soil each spring was significantly reduced by burning in the previous spring. Repeated spring fires rarely reinforced any initial change induced by burning, and slightly lowered average ground cover as well as various indices of landscape stability and ecosystem functionality. Changes produced were not always consistent across the two communities. Though prescribed burning is often important for maintaining grazing productivity and landscape values, very regular use is not recommended.

Experimental reintroduction of three grassland forbs to assess climate-adjusted provenancing, grazing protection and weed control

We explored climate-adjusted provenancing as a strategy for reintroduction of three perennial forbs: Microseris walteri, Bulbine bulbosa and Stackhousia monogyna into native grassland with a history of livestock grazing. Seed source populations were sampled over a temperature gradient of 2°C, located up to 120km from the receiving site. Two receiving-site management treatments were applied: hand removal of exotic species, and exclusion of macropod grazing and trampling. There were no effects of weeding the receiving grassland and we concluded the low fertility of the site provided protection from exotic competition. After eight growing seasons, protection from grazing resulted in double to triple the numbers of established plants, depending on the species. Amongst populations within a species, larger seeds and early emergence reflected later establishment success for Bulbine only. At the species level, early performance was not indicative of later numbers of reproductive plants established. We found no long-term evidence of home-site advantage in any species, and no apparent disadvantage of non-matching lithology. All source populations were represented at the end of the experiment (Season 6). Counts after eight seasons’ growth indicated a continuing trend of population increase in all three species. The persistence of all populations suggests that admixture or climate-adjusted provenancing may be a realistic option to restore depleted populations of herbaceous plants in grassy eucalypt woodlands in a warming climate.