Semi-arid Woodland Research Articles

Influence of Landscape Structure on Kangaroo Abundance in a Disturbed Semi-Arid Woodland of Queensland.

This paper develops a hierarchical landscape approach for investigating how landscape structure influences the abundance of eastern grey kangaroos, red kangaroos and common wallaroos on properties in a partially cleared semi-arid woodland of Queensland. This approach is applied to examine the extent to which a mosaic of spatial factors at a hierarchy of scales influences kangaroo abundance at the property level. the level of management interest. The analysis concludes that the structure of the property habitat mosaic, coupled with pasture productivity, is the most important influence for each species. Access to water was not a limiting factor. although it may be important locally. Grey kangaroos prefer an even mosaic of all habitat types on properties with productive grass-dominated pastures. Red kangaroos are positively associated with run-on areas and large-shrub regrowth patches. reflecting their foraging preferences for forbs and short grasses and their ability to use more open habitats. Wallaroos have a clumped density distribution associated with a heterogeneous mosaic of open habitats interspersed with fragmented forest patches and small to medium-grained shrub-regrowth patches. The research. \+bile not replicated. identifies linkages between tree clearing practices at the property level and increased large kangaroo abundance in the region, These linkages have been previously overlooked in thc kangaroo Inanagement debare. Therefore. any tree clearing guidelines dt.\eloped at the propert) level need to be sensitive to key ecological processes influencing kangaroo populations at both the landscape and property levels. If not, the sustainable management of total grazing pressure (livestock plus kangaroos) and biodiversity conservation will never become a reality. Key words: kangaroos, landscape structure, landscape change, pattern, process, scale

Effects of Woody Plants on Microclimate in a Semiarid Woodland: Soil Temperature and Evaporation in Canopy and Intercanopy Patches

Effects of Woody Plants on Microclimate in a Semiarid Woodland: Soil Temperature and Evaporation in Canopy and Intercanopy Patches

Effects of Woody Plants on Microclimate in a Semiarid Woodland: Soil Temperature and Evaporation in Canopy and Intercanopy Patches

The canopies of woody plants in semiarid ecosystems modify the microclimate beneath and around them, with canopy patches usually having lower soil temperatures than intercanopy patches. However, lacking are studies that have evaluated how heterogeneity in soil temperature, induced by woody plant canopies, influences soil evaporation rates and the consequent effects on plant‐available water. Soil temperatures were measured and soil evaporation rates were estimated for canopy and intercanopy patches in a semiarid pinon‐juniper woodland (Pinus edulis and Juniperus monosperma) in northern New Mexico. Soil temperature was measured at 2‐cm depths in four canopy and four intercanopy locations during 1994. Maximum soil temperature in intercanopy patches was greater than in canopy patches between May and September, by as much as 10°C, while soil temperatures in intercanopy patches were lower than in canopy patches during colder parts of the day in the fall and winter months. Equations for soil drying rates for san...

Runoff and micromorphological properties of a grazed haplargid, near Cobar, NSW, Australia

We investigated the effects of 2 different grazing regimes on the surface soil properties of a dunefield land system in the semi-arid woodlands of eastern Australia. Sandy siliceous, thermic Xeric Haplargids (Siliceous Sands, Uc1·23) occur on the sandy, 2–4-m-high longitudinal dunes. Fine-loamy, siliceous, thermic Xeric Haplargids (Massive Red Earths, Uc2·13) occur in the swales between the dunes. We compared very high-intensity grazing (approx. 1 year) by feral goats with low-intensity grazing (approx. 4 years) by sheep. A rainfall simulator, applying water at the rate of 30 mm/h, measured the hydraulic properties of the surface soils formed under the 2 different grazing regimes. We examined undisturbed samples of the upper 5-cm layer of the soil surface using micromorphological techniques. In the swales, there were no differences in the effects of the 2 grazing regimes on soil properties. At low-intensity sheep grazing (0·2–0·3 sheep/ha), the soil surface on the dunes remained in an excellent condition. The surface had a good vegetative cover and consisted of either loosely packed sand grains, or areas where the sand grains were bonded together by clay and organic matter to form an organic crust. The total carbon content of the 0–2 cm depth of soil was 0·86%. Both soil surfaces have a high infiltration rate (i.e. >30 mm/h) and also appear to contain stable microaggregates of parna material distributed among the eolian sand grains. Very high-intensity goat grazing (up to 4·0 goats/ha) rapidly depleted the perennial grasses, killed most of the shrubs, and converted the soil surface on the dunes to one highly susceptible to erosion by wind. The low total carbon content (depth 0–2 cm) of 0·3% and absence of iron-stained clay coatings on the sands further support this view. The surface soil on the dunes in the very high-intensity goat-grazing plots consisted of either loosely packed sand grains or areas where poorly orientated clays coated the sand grains to form a strong, physical crust. We suggest that the physical crust may cause a change in the hydrology of the land system which may enhance the conditions for recruitment of unpalatable shrubs in the dune–swale interface. This increase in unpalatable shrubs and decrease in perennial grasses caused by the very high intensity goat grazing is therefore detrimental to the long-term productivity of these semi-arid lands.

Overstory-Imposed Heterogeneity in Solar Radiation and Soil Moisture in a Semiarid Woodland

Overstory-Imposed Heterogeneity in Solar Radiation and Soil Moisture in a Semiarid Woodland

Differential Use of Spatially Heterogeneous Soil Moisture by Two Semiarid Woody Species: Pinus Edulis and Juniperus Monosperma

1 Soil moisture in semiarid woodlands varies both vertically with depth and horizontally between canopy patches beneath woody plants and the intercanopy patches that separate them, such that shallow soil layers in intercanopy locations are wettest, yet few studies have considered both dimensions of spatial variability in testing for acquisition of resources by plants. 2 Three hypotheses were tested relative to the use of shallow water in intercanopy locations by two coexisting semiarid-woodland tree species, Pinus edulis (a pinion) and Juniperus monosperma (a juniper): (i) both P. edulis and J. monosperma can use shallow water from intercanopy locations;. (ii) J. monosperma is able to obtain more shallow water from intercanopy locations than P. edulis, and (iii) the spatial arrangement of the trees influences the amount of water they obtain. Soil moisture and plant water potential (i.e. plant water stress) were measured before and after the addition of water to shallow depths (0-30 cm) of intercanopy locations for trees of both species in two spatial arrangements: isolated and paired with a contiguous tree of the other species. 3 Both species responded to the addition of shallow water in intercanopy locations, as measured by plant water potential. The response of J. monosperma was significantly greater than that of P. edulis, as measured by depletion of shallow soil moisture in intercanopy locations and by change in plant water potential per unit change in soil water potential (the difference was not detectable on the basis of plant water potential alone); in addition, the amount of depletion was correlated with basal area of J. monosperma but not of P. edulis. The responses were not influenced by spatial arrangement (isolated vs. paired with a contiguous tree of the other species). 4 The results of this study are consistent with differences in the relative abundances of the two species across locations, suggesting that species differences in ability to use shallow water in intercanopy locations is important in structuring semiarid woodlands. Further, the results suggest that current theoretical concepts for semiarid ecosystems, which ignore either vertical or horizontal variability in soil moisture, may be inadequate for predicting changes in the ratio of woody to herbaceous plant biomass, particularly for plant communities with co-dominant woody species that differ in ability to acquire spatially heterogeneous resources.

Grass-mediated Capture of Resource Flows and the Maintenance of Banded Mulga in a Semi-arid Woodland

It has been proposed for semi-arid and arid regions that the scarce resources of water and nutrients normally limiting plant growth are distributed patchily in the landscape in order to maintain both stability and productivity (Tongway 1990). ‘Fertile patches’ are maintained by the concentration of resources during flows and when resource control is weakened their persistence is threatened. This proposition was tested by comparing water relations of perennial life-forms in the nutrient and water rich ‘island-bands’ of mulga (Acacia aneura F.Muell. ex Benth.) shrubs, with and without perennial grasses. Dawn xylem water potential (dawn ψx) and stable hydrogen isotope composition data showed that A. aneura and companion grasses competed for soil water, whereas the poplar box (Eucalyptus populnea F.Muell.) trees accessed additional sub-soil water. During a drying cycle, dawn ψx of A. aneura on the upslope portion of ‘island-bands’ was –2 MPa where grass was abundant, and –4.8 MPa where grasses were killed by heavy grazing. On the downslope portions, dawn ψx of A. aneura was similar at –4.2 MPa. Recruitment of A. aneura was occurring in the upslope portion, but was less where there were no grasses. There were more dead A. aneura shrubs on the downslope portion and in grass-free ‘island-bands’. The results suggest that removal of perennial grasses by grazing shortens the period of water supply to A. aneura shrubs in ‘island-bands’ and they prematurely die during low rainfall periods. Furthermore, continued grazing pressure from domestic and non-domestic herbivores will prevent restoration of landscape processes necessary for the survival of these shrubs and associated flora.

Emissions of CO2, CO, and hydrocarbons from fires in diverse African savanna ecosystems

Emissions of CO2, CO, and hydrocarbons from 13 savanna fires (∼7 ha) were investigated in Zambia and South Africa in August and September 1992. The experiments were conducted at two moist woodland savanna sites, a moist grassland savanna site, and a semiarid woodland savanna site in Zambia, and nine semiarid woodland savanna sites in South Africa. The hydrocarbons measured were CH4, C2–C6alkanes, alkenes, and alkynes, and aromatic compounds. The linear relationship between the emission factor of CH4and the combustion efficiency is similar regardless of the phase of combustion (flaming or smoldering), the climatic zone, the herbaceous species, or the amount of aboveground biomass. Concentrations of emitted nonmethane hydrocarbons are linearly correlated with those of CH4. The emission ratio of a trace gas to CH4is constant for either phase of combustion in these ecosystems. Biomass burning in African savannas is a significant source of CO, ethyne, propene, and benzene in the atmosphere. The amount of CO, ethene, ethane, ethyne, propene, propane, and benzene emitted per year from these fires is about 20–95% of the amount released from global industrial activities.

Carbon isotope discrimination in three semi-arid woodland species along a monsoon gradient.

Leaf carbonisotope discrimination (Δ) was measured for three dominant, semi-arid woodland species along a summer monsoon gradient inthe southwestern United States over a 2-year period. We tested the hypothesis that decreased humidity levels during the growing season along this gradient resulted in lower leaf Δ values. Sites of similar elevation along the transect were selected and the range in monsoon contribution to overall annual precipitation varied from 18 to 58%, while total annual precipitation differed by a maximum of only 25% across this gradient. Leaf Δ values in Quercus gambelii were negatively correlated with ϖ, a seasonally-weighted estimate of the evaporative humidity gradient, suggesting that stomatal conductance declined as transpiration potential increased. For two other trees that co-occurred along this gradient, Pinus edulis and Juniperus osteosperma, Δ remained relatively constant despite large variation in ϖ. These woodland species represent the full spectrum of responses of carbon isotope discrimination to increases in evaporative potential; that of decline where c i /c a (ratio of internal to ambient CO2 concentration) and presumably stomatal conductance decrease, and that of constancy where whole plant internal adjustments allow c i /c a to remain stable.

The Influence of Recent Grazing Pressure and Landscape Position on Grass Recruitment in a Semi-Arid Woodland of Eastern Australia.

This study examined the effects of previous grazing pressure, position in the landscape and apparent seed trapping capability of soil surface micro-sites on recruitment of the perennial grass Monachather paradoxa (mulga oats) in a semi-arid woodland. Seedling emergence was counted on small plots which had been kept moist for one month. The plots were on bare ground, or at grass tussocks, or at log mounds, sited in the run-off, interception and run-on zones of paddocks that had been grazed for six years at 0.3 and 0.8 sheep equivalent/ha. Few naturally occurring perennial grass seedlings emerged on any of the sites. The level of previous grazing pressure influenced the recruitment of grasses from natural sources as well as from seed of M. paradoxa broadcast on the soil surface; significantly more grass seedlings recruited in paddocks stocked at 0.3 than at 0.8 sheep/ha. Emergence of the sown grass did not differ significantly between the three zones in the landscape, but trends in the data suggest the interception zone may have been the most favourable. Recruitment from in situ grass seed was highest in the mulga grove (run-on) zone. Most seedlings of the sown grass emerged around the bases of existing perennial grass tussocks, but recruitment of volunteer perennial and annual grasses was more evenly distributed between the mulga log-mounds and perennial grass tussocks. It is concluded that very low levels of readily germinable seed of perennial grasses remained in the soil at the end of the drought and that areas with a history of high grazing pressure have less probability of grass recruitment when suitable rain occurs.

Characteristics of Arboreal Marsupial Habitat in the Semi-Arid Woodlands of Northern Queensland.

The distribution of the koala (Phascolarctos cinereus) and the common brushtail possum (Trichosurus vulpecula) within the Prairie-Torrens Creek Alluvials province of the Desert Upland region of north-western Queensland was examined. The optimum habitat for each species as indicated by the occurrence of faecal pellet groups was found to be that associated with creek-lines. However, other land types were also used by each species to varying degrees. The relationship between various habitat variables and pellet group counts was investigated using Multiple regression and a Generalised linear model. Proximity to creek-bed, total basal area of trees, species richness and Acacia basal area (negative) best explained the occurrence of koalas. Proximity to creek-bed, Acacia basal area (negative), total basal area of trees, and available water (negative) best explained the occurrence of brushtail possums. In contrast to studies of arboreal species in the moist-south-eastern forests of Australia no relationship was found between foliar nutrient concentrations and the occurrence of koalas or brushtail possums. However, a significant relationship was found between leaf water concentration and the occurrence of koalas. It is suggested that water availability is the paramount factor defining preferred arboreal habitat in arid and semi-arid woodlands.

Dispersal of Microphytes by Water Erosion in an Australian Semi-Arid Woodland

Dispersal of Microphytes by Water Erosion in an Australian Semi-Arid Woodland

Landscape and Microsite Influences on Shrub Recruitment in a Disturbed Semi-Arid Quercus-Juniperus Woodland

Seedling demography of the leguminous shrubs Leucaena pulverulenta and Acacia berlandieri was monitored to determine the functional role of microsites in the restoration of disturbed semiarid woodlands. Seeds were sown on replicated landscapes during 1989 and again in 1990 into four microsites after brush clearing: 1) Rocky, with shallow soils ( 80%), intermediate on Rocky (65%), and lowest on Herbaceous (36%) microsites. Desiccation appeared to be the primary cause of first-season mortality. Differences between the microsites in seedling height and number of leaves were significant the first planting but not the second. Overwinter mortality of Leucaena and Acacia seedlings was lowest on Cleared microsites (55 and 7%, respectively) and greatest on microsites dominated by Herbaceous vegetation (100 and 87%, respectively). Survival of established seedlings during the second growing season exceeded 80% for both species on all microsites. A. belandieri had the highest establishment potential and the greatest ecological amplitude, with recruitment after two years being comparable on Cleared, Rocky and Tree microsites (72-78 seedlings). L. pulverulenta showed marked preferences for Cleared microsites (47 seedlings established) relative to Rocky (13 seedlings), Tree (12 seedlings) and Herbaceous microsites (0 seedlings). Spatial variability in emergence between replicated landscapes (range 35 to 78%) was greater than the variability between microsites within a landscape (48 to 62%). Differences in seed disappearance associated with surface runoff from high intensity storms and losses to granivory on Tree microsites were the primary causes of differences in recruitment between landscapes. Inter-landscape variability in granivory and precipitation runoff therefore overshadowed the effects of within-landscape seed placement among first year seedlings. Generalizations of seedling safe site characteristics based solely on short-term (one year), pseudoreplicated, within-landscape studies may therefore be misleading. Failure to account for variability in important processes at greater spatial scales may significantly influence the robustness of experimental microsite study results derived from small-scale research.

Sediment loads in an Australian dust storm: implications for present and past dust processes

On 1 December 1987 a dust storm driven by a strong cold front entrained soil from 100m000 to 200m000 km z of the Simpson Desert, Channel Country and semi-arid woodlands of Western Queensland for eighteen hours. Between 6.3 and 5.5 million t of soil with average particle sizes of 6.75 to 10.5 μm were lost from the dust source, 3.4 to 1.9 million t was lost off the Australian continent, and from this some 3 to 1.7 million t might have traversed the Tasman sea towards New Zealand. A mathematical model is used to estimate dust lost from the region. This approach provides estimates of Total Suspended Particulate (TSP) that are within a factor of 3 of observations. From statistics of the frequencies of dust storms 3.8 to 6.8 million t of dust per year is estimated to fall into the South Pacific and contribute to marine sediments. These are the largest estimates of soil loss by wind in Australia and support quantitatively earlier hypotheses of the significance of dust process in the evolution of the Australian landscape.

Spatial organisation of landscapes and its function in semi-arid woodlands, Australia

The spatial organisation of three major landscape types within the semi-arid woodlands of eastern Australia was studied by a detailed analysis of gradient-oriented transects (gradsects). The aim was to characterise the spatial organisation of each landscape, and to account for that organisation in functional terms related to the differential concentration of scarce resources by identifiable processes. Terrain, vegetation and soils data were collected along each gradsect. Boundary analysis was used to identify the types of landscape units at a range of scales. Soil analyses were used to determine the degree of differential concentration of nutrients within these units, and to infer the role of fluvial and aeolian processes in maintaining them. All three major landscape systems were found to be highly organised systems with distinctive resource-rich units or patches separated by more open, resource-poor zones. At the largest scale, distinct groves of trees were separated by open intergroves. At smaller-scales, individual trees, large shrubs, clumps of shrubs, fallen logs and clumps of grasses constituted discrete patches dispersed across the landscape. Our soil analyses confirmed that these patches act as sinks by filtering and concentrating nutrients lost from source areas (e.g., intergroves). We suggest that fluvial runoff-runon and aeolian saltation-deposition are the physical processes involved in these concentration effects, and in building and maintaining patches; biological activities also maintain patches. This organisation of patches as dispersed resource filters (at different scales) has the overall function of conserving limited resources within semi-arid landscape systems. Understanding the role of landscape patchiness in conserving scarce resources has important implications for managing these landscapes for sustainable land use, and for the rehabilitation of landscapes already degraded.

Natural Regeneration of Ground Storey Vegetation in a Semi-Arid Woodland Following Mechanical Disturbance and Burning. 2. Response of Individual Species.

A study was made over a period of 12 years of the natural regeneration of vegetation along a corridor cleared for the construction of a natural gas pipeline in a semi-arid woodland in central-western New South Wales. Total cover, proportion of grasses, and species composition were assessed on the infilled trench and areas from which topsoil had been bladed, as well as on adjacent undisturbed areas or areas burned by a wildfire. All areas were grazed continuously by sheep under normal station management. The effects of mechanical disturbance and of burning on individual species were measured in terms of species dominance and occurrence. Some insight was also gained in to the successional process on bared sites and in to the effects of the timing of seasonal rainfall on species composition. Of the 75 major species recorded, 15 were found to have greater occurrence on trenched sites, 46 decreased in occurrence and 14 showed no clear trend. Perennial grasses and small annual plants were the main decreasers, whilst Medicago spp., Erodium crinitum, Hordeum leporinum and certain 'weedy' annuals were the main increasers. Some of the early colonisers lasted only a few years or even less in any quantity, thereafter to be replaced by other increasers. Burning had little effect on the long-term occurrence of Stipa variabilis, Wahlenbergia spp., Helipterum spp., Calotis cuneifolia and some minor species. Eragrostis lacunaria appeared to decrease, while several annual plants were more common on burned areas.

Natural Regeneration of Ground Storey Vegetation in a Semi-Arid Woodland Following Mechanical Disturbance and Burning. 1. Ground Cover Levels and Composition.

A study was made over a period of 12 years of the natural regeneration of vegetation along a corridor corridor cleared for the construction of a natural gas pipeline in a semi-arid woodland in central- western New South Wales. Total cover, proportion of grasses, and species composition were assessed on the infilled trench and on areas from which topsoil had been bladed, as well as on adjacent undisturbed areas and areas burned by a wildfire. All areas were grazed continuously by sheep under normal station management, and by native animals. After four years of about average rainfall, total cover on disturbed areas had regained levels as high or higher than on adjacent areas, and was maintained thereafter. On many individual sites this period was as short as eighteen months. The proportion of grasses remained much lower on disturbed areas, except for some periods of annual grass dominance. Species composition remained very different between disturbed and undisturbed areas, the main difference being in perennial species. However, composition also varied markedly between years, according to rainfall seasonality and competition from previously established plants. Little difference was apparent between undisturbed areas and similar areas which had burned twelve months before the first measurements. Mechanical disturbance caused by trenching and blading had a much greater effect on vegetation than did fire, and the effects were still obvious after twelve years.

The Effects of Grazing Exclusion and Blade-Ploughing on Semi-Arid Woodland Vegetation in North-Western New South Wales Over 30 Months.

Inedible native shrubs are considered to be a primary cause of lost pastoral productivity in the semi- arid woodlands of north-western New South Wales. Most rehabilitation programs have focussed on reducing shrub density. In recent years blade-ploughs have been tested for this purpose. Results have been variable and most programs have brought neither lasting shrub control nor improved production for pastoralism over the medium to long term. Failure to control grazing after initial treatment is suspected as one reason for the low success rate. An experiment involving blade-ploughing and grazing exclusion was established in July 1990 in sandplains supporting semi-arid woodlands, near Bourke New South Wales (NSW). The treatment factors were crossed and each was imposed at two levels: present and absent. Shrub density, pasture composition and pasture biomass variables were selected and measured before, and over a 30 month period following treatment. Data were analysed using repeated measures ANOVA. The densities of Eremophila sturtii, E. deserti, E. gilesii and Dodonaea viscosa ssp. angustissima as well as that of 'Total shrubs', all increased after ploughing, irrespective of the presence or absence of grazing. In all cases except that of E, sturtii, there was no significant difference between shrub density in ploughed and unploughed plots 30 months after treatment. While E. sturtii density increased significantly to begin with, it remained significantly lower in ploughed plots compared to unploughed plots at the end of the study period. This suggests that blade-ploughing offered only temporary control of these shrubs on this land type. Pasture biomass was significantly greater in ploughed/ungrazed plots (1300 kg/ha) compared with other treatments at the end of the study. Ploughed/ungrazed plots were the only ones where fuel had accumulated to a level which offered any possibility of burning to control shrub regeneration. The same treatment also contained a significantly greater proportion of desirable pasture species than any other treatment at the end of the study period.

Changes in Grass Density in Australian Semi-Arid Woodlands.

Changed grazing regimes since European settlement have been widely proposed as the cause of a decline of indigenous perennial grasses in the semi-arid woodlands of eastern Australia. A five year experiment using exclosures examined the effects of grazing on densities of perennial grasses. Short- lived Stipa spp. and Aristida spp. were dominant at most sites. Their densities fluctuated greatly with season and reached over 200 plants/m2 during climatically favourable periods. The long-lived Eragrostis eriopoda occurred at densities that were generally less than 5 plants/m2 and its populations were relatively stable. The response of Enneapogon avenaceus was distinctive. Though its density fluctuated considerably, successive peaks in density were higher and the species increased more in ungrazed areas than in destocked or unfenced areas. The differences between grazed and ungrazed populations became greater with successive peaks in density. Within the short periods that pastoralists are likely to be willing or able to apply such treatments, destocking or even removing all herbivores is unlikely to have a large effect on the density of many palatable perennial grass species. The rate of response to resting pastures will depend on seasonal conditions.

Pasture management in semi-arid tropical woodlands: effects on herbage yields and botanical composition

The effects of pasture management options (introduced legumes and grasses, superphosphate, timber treatment, cultivation before sowing, stocking rate) on the yield and botanical composition of pastures were measured from 1982 to 1992 at 2 sites, Hillgrove (extractable phosphorus, 50 mg P/kg) and Cardigan (extractable P, 6 mg/kg), near Charters Towers, northeastern Queensland. Despite generally poor establishment and growing conditions, some sown species (Stylosanthes hamata, S. scabra, Cenchrus ciliaris, Urochloa mosambicensis) established and persisted at both sites. The initial sown grass populations were small but yields increased markedly towards the end of the experiment. There were no yield responses by the native pastures to superphosphate, but the sown pastures responded at Cardigan, although not at Hillgrove. There were large yield responses to tree killing. The percentage increase in herbage yield following tree killing increased as the available water in the soil during the growing season decreased, and was greater at the higher fertility Hillgrove site. Overall, the native pastures were dominated by perennial grasses, but annual grasses and forbs increased on plots with live trees and high stocking rates towards the end of the trial. sown grasses, particularly C. ciliaris, became dominant (and are likely to remain so) at Hillgrove and on plots with superphosphate at Cardigan, especially where the trees were killed. Productive, mixed pastures with reasonable proportions of both grass and legume were maintained under live trees.