Keywords

1 Introduction

Australia (113°08′E–153°38′E, 10°41′S–43°38′S), has a terrestrial land area of almost 7.7 million km2, which includes the Australian continent mainland, the island of Tasmania, and numerous small islands. Australia is the driest inhabited continent in the world (Commonwealth of Australia 2012). The arid Australian climate can be attributed to the subtropical anticyclonic zone which covers the center of the continent; the Great Dividing Range which blocks water vapour from the east coast; and the West Australia Current, a cold current that significantly reduces precipitation in Western Australia. The climate in Australia is highly variable, with frequent drought events throughout the country. The climatic fluctuation and extreme climates in Australia are mainly driven by ocean currents, including the Indian Ocean Dipole and the El Niño–Southern Oscillation. Australia is dominated by drylands (aridity index < 0.65; 733.9 × 104 km2; 95.4% of terrestrial Australia), with water-sufficient areas existing only on Tasmania Island, in the eastern and northern coastal areas, and in the southwest corner. More than half of the country (65.2%) is composed of arid regions, followed by semiarid (25.5%), dry subhumid (4.7%), and humid (4.6%) regions, while hyper-arid areas are 0.007% of Australia area (Fig. 11.1). The arid Australian climate gives rise to a specialized and quite vulnerable terrestrial ecosystem which can be characterized by pervasive deserts and sparse grasslands. Approximately 80% of terrestrial Australia is classified as rangelands, where land use is dominated by extensive grazing of sheep and cattle (Feng et al. 2020; Foran et al. 2019).

Fig. 11.1
A heat map of Australia and a pie chart present the distribution of drylands. They indicate the following. Arid, 501.48. Semiarid, 196.28. Dry sub-humid, 36.12. Humid, 35.26. Hyper arid, 0.06. All values are 10 to the power of 4 kilometer squared.

Distribution of drylands in Australia. The standard global aridity index (AI, the ratio of long-term average annual precipitation to average annual potential evapotranspiration) map is from the Global Aridity and Potential Evapotranspiration (PET) Database v2 developed by the Consultative Group for International Agriculture Research-Consortium for Spatial Information (CGIAR-CSI) (Trabucco and Zomer 2019). AI ranges are 0–0.03, 0.03–0.2, 0.2–0.5, 0.65–0.65 and >0.65 for hyper-arid, arid, semiarid, dry subhumid, and humid regions, respectively

Full size image

In spite of the water-limited climate conditions and vulnerable terrestrial ecosystems, Australia has the highest average wealth, and the GDP per capita was approximately 5.74 × 104 US dollars in 2018. Australia has a population of nearly 26 million, equalling an average population density of 3.4 per km2. The population is highly concentrated in cities on the eastern seaboard. Population distribution pattern outside the main cities are of a few medium size towns and then many very small communities. Dryland social-ecological systems (SES) in Australia are threatened by the degradation of rangelands due to more arid climates and excessive grazing. Moreover, agricultural expansion, especially poor irrigation activities in areas with high potential evapotranspiration but limited rainfall, has led to dryland salinity, which is a key problem contributing to land degradation in southern Australia (Clarke et al. 2002; Lambers 2003). Human society in terms of population distribution, economic development, and the livelihoods of local communities, is greatly affected by water deficits and drought-induced ecosystem degradation. Therefore, in this chapter, we would like to provide an overview of the spatiotemporal dynamics of climate, ecosystems, and human society in Australia, especially during recent decades, and explore the relationships among these three key components. Multiple datasets on environmental conditions, vegetation cover, and human society or activities are analysed, and published studies are referenced.

2 Major Characteristics of Dryland Social-Ecological Systems in Australia

2.1 Climate Conditions

The mean annual temperature (MAT) is the lowest in the southeastern part of Australia, e.g., Tasmania Island. In tropical areas in northern Australia, the weather is perennially hot, whereas in the interior of the continent, which is covered by the arid anticyclone, summers are extremely hot, and winters are cool. Extreme temperatures influence vegetation, animals, and even humans (Cheng et al. 2018; Ebi et al. 2021; Hoffmann et al. 2019). Annual precipitation is the lowest in central Australia and high in northern Australia and some coastal regions. The precipitation seasonality declines from north to south in Australia and is higher on the southwestern coast than in the southeastern regions. For most parts of Australia, precipitation is the highest in summers and the lowest in winters. However, western and southern Australia showed the opposite pattern of seasonal precipitation variation. Therefore, over the southwest corner and Spencer Bay (including Kangaroo Island) located in southern Australia, where precipitation seasonality is large, typical Mediterranean climates are present. Mean annual solar radiation is generally higher in the north than in the south, but the seasonality of radiation increases with latitude. Desert areas receive higher radiation than relatively humid places (Fig. 11.2).

Fig. 11.2
6 heatmaps of Australia are labeled a to f. The precipitation seasonality, mean annual precipitation, temperature, and solar radiation are high over the northern regions, while temperature seasonality and solar radiation seasonality are high over the central, western, and Victoria regions, respectively.

Spatial patterns of temperature, precipitation and solar radiation and the seasonality in the drylands of Australia. The data are from the Australian Gridded Climate Data (AGCD), which are newly published by the Bureau of Meteorology Australia at the national computational infrastructure (NCI), as the successor of the Australian Water Availability Project (AWAP). The monthly precipitation data are available from 1900 until 2019 (Australian Bureau of Meteorology 2020b), whereas the monthly means of daily minimum and maximum air temperature data are available from 1910 to 2019 (Australian Bureau of Meteorology 2020a). Both temperature and precipitation data have a spatial resolution of 0.05°. Solar radiation data are from the updated Breathing Earth System Simulator (BESS)-RAD dataset from 2000 to 2019 and have a high accuracy (Ryu et al. 2018)

Full size image

The dryland climates in Australia (hereinafter ‘DRY AUS’) can be classified into 10 types (Fig. 11.3) according to the world map of the Köppen-Geiger climate classification (Kottek et al. 2006; Rubel et al. 2017). The central and western parts of Australia are dominated by a tropical desert climate (BWk), the northern coasts are hot year-round and dry in winter (Aw), most parts of the eastern coasts have warm and humid weather (Cfa and Cfb), and Mediterranean climates (Csa and Csb, warm and dry summers) dominate the southwest corner and some areas in southern Australia. The mean annual air temperature increases with the aridity level, but precipitation declines with the aridity level (Fig. 11.6a). The interrelationships among the interannual variations in solar radiation, temperature, and precipitation are all stronger in more arid regions (according to data from 2000 to 2019). Precipitation is negatively correlated with solar radiation (p < 0.01) in all regions. Temperature is positively correlated with radiation and negatively correlated with precipitation in all drylands in Australia, but these relationships are not significant in dry subhumid areas.

Fig. 11.3
A heatmap of Australia categorized the regions based on 10 different Koppen-Geiger climate classifications, namely, A w, B W k, B S k, C s b, C f a, B W h, B S h, C s a, C w a, and C f b. Most of the central and southern regions of Australia fall under B W h, while some northern regions fall under A w.

Köppen-Geiger climate classification in the drylands of Australia (In the symbol in the legend, the first letter indicates the main climates: A: equatorial; B: arid; C: warm temperate; the second letter indicates precipitation: W: desert; S: steppe; f: fully humid; s: summer dry; w: winter dry; and the third letter represents temperature: h: hot arid; k: cold arid; a: hot summer; b: warm summer)

Full size image

2.2 Soil and Topography

Australia has the lowest and flattest topography among all continents. However, eastern Australia is marked by the Great Dividing Range, which stretches more than 3500 km and has widths from 160 km to more than 300 km. The heights of the range are typically 300–1,600 m. The southern Great Dividing Range contains the highest place in mainland Australia: Mount Kosciuszko (2228 m above sea level). Except for eastern Australia, where the silt or clay fraction is relatively high in soils, sand dominates the surface soil in the drylands of Australia (Fig. 11.3). Soil organic carbon is high in the eastern coast and southwest corner of Australia but is quite low in the interior parts of the continent, especially in the desert areas (Fig. 11.4).

Fig. 11.4
2 heatmaps of Australia labeled a and b. A. Most of the western and central regions are dominated by sand, while the eastern and northern regions have silt and clay fractions. B. Higher soil organic carbon percent is noted in the eastern and south-western regions, while other regions have lower values.

Soil properties of drylands in Australia. a Soil texture and the fractions of sand, silt and clay; b spatial pattern of soil organic carbon. Soil properties are from the national soil attribute maps produced by the Soils and Landscape Grid Facility of Commonwealth Scientific and Industrial Research Organization (CSIRO) and soil mapping (Odgers et al. 2015a, b; Viscarra Rossel et al. 2015). Bulk density, pH, available water capacity, total nitrogen, and total phosphorus of the soil layer (0–5 cm) are provided in this dataset (Viscarra Rossel et al. 2014)

Full size image

2.3 Land Use/Cover in Dryland Regions in Australia

Drylands in Australia are dominated by sparse and scattered grasses and shrubs (37.1%), followed by open shrublands (10.5%) and sparse trees (9.8%), all of which are typical ecosystem types in arid climates. Shrublands and grasslands are representatives of arid and semiarid regions in Australia and rarely exist in more humid places. On the other hand, closed forests mainly exist in humid and dry subhumid regions, while open forests can be found in semiarid, dry subhumid, and humid regions (see Fig. 11.5). Vegetation cover is much denser in more humid coastal areas. From the humid coasts to the dry interior lands, the ground cover changes from forest to grass and finally to bare ground (Fig. 11.6).

Fig. 11.5
2 parts. A. A map of Australia plots 22 different kinds of land use cover from 2002 to 2015. The open and closed trees are noted over the eastern and south western coasts. B. A stacked bar graph of land use cover type versus area. The shrubs and grasses, sparse and scattered, are highest in arid areas.

Land cover types in drylands in Australia: a most of the land cover in the drylands of Australia during 2002–2015; b annual mean total area of 22 land use cover types in different regions distinguished by aridity levels in Australia. Please note that the drylands of Australia, denoted by DRY AUS, consist of arid, semiarid, and dry subhumid regions. The data are from the National Dynamic Land Cover Dataset (NLCD) v2.1, published by the Australian Bureau of Agricultural and Resource Economics and Science (Lymburner et al. 2015)

Full size image
Fig. 11.6
2 heatmaps of Australia. A. The mean annual L A I value is high over the eastern and southwestern coastal regions, while lower values are noted over the interior regions. B. Most of the regions have short vegetation, with bare ground prevailing in the central region.

Vegetation cover in drylands in Australia. a Map of the mean annual leaf area index (LAI) during 2001–2018; b mean vegetation continuous fields in 1982–2016. The LAI data are from the Global Land Surface Satellite (GLASS) product suite (Liang et al. 2021; Tang et al. 2013; Xiao et al. 2014; Xu et al. 2018). Vegetation continuous field (VCF) datasets are from the National Aeronautics and Space Administration (NASA) Making Earth System Data Records for Use in Research Environments Project (Hansen and Song 2018; Song et al. 2018)

Full size image

2.4 Socioeconomic Factors

Australia’s population was 25,704,340 on 31 March 2021 according to the Australian Bureau of Statistics. The population in Australia is concentrated in humid regions, especially urban areas located on the southeast and southwest coasts (Fig. 11.7a). The average population densities in the arid, semiarid, dry subhumid, and humid regions of Australia were 0.12, 4.05, 15.98, and 21.81 individuals per km2, respectively. The population in the drylands (arid, semiarid, and dry subhumid regions) of Australia increased from 12.33 million in 2000 to 16.52 million in approximately 2020. The mean gross domestic production (GDP) values per area were 0.41, 14.40, 52.93, and 66.05 × 104 US$/km2 in the arid, semiarid, dry subhumid, and humid regions of Australia, respectively (Fig. 11.7b).

Fig. 11.7
2 maps of Australia. A. Most of the interior regions have less than 0.1 population density, while the eastern and southwestern coasts have about 0.1 to 0.5 population density. B. It indicates G D P in 10 to the power of 4 U S dollars per kilometer square. Most of the interior regions fall between 0 and 1.

Socioeconomic factors in the drylands of Australia: a population and b GDP. Gridded population maps are from the Gridded Population of the World (GPW) v4 from the Socioeconomic Data and Applications Center (SEDAC). The Gridded GDP is from Kummu et al. (2018)

Full size image

Grazing (beef cattle/sheep) is a key industry in Australia. Meat and wool production contributed almost 30% of the gross agricultural production value in 2009–2010, and the total number of grazing businesses (farmers) during that period was 88,945 (Commonwealth of Australia 2011). According to National Scale Land Use version 5, in 2010–2011, the total grazing land area was 415.1 × 104 km2, including 344.9 × 104 km2 of grazing native vegetation land and 70.2 × 104 km2 grazing-modified pastures, which accounted for 44.8 and 9.1% of terrestrial Australia (Commonwealth of Australia 2016). Grazing native vegetation land is mainly located in the arid region (71.5%), and grazing-modified pastures are mostly distributed in the semiarid region (66.5%) (Fig. 11.21a, b). Thus, the fraction of grazing area to the total area in the semiarid region is the highest, reaching 66.4%, followed by 51.8%, 42.7% and 26.6% for the arid, dry subhumid, and humid regions, respectively.

The indigenous lands, the next largest land use in Australia, is albeit diverse and with its own economy. Land use for mining and tourism is also important as a major part of economic activity besides the extensive grazing in the indigenous lands.

3 Changes in Ecosystem Structures

The sparse tree areas have decreased, while the areas with open and closed trees have increased, indicating that forests in the drylands of Australia have probably become denser since 2002. However, for shrublands, the closed areas have declined, while the areas with open shrublands and sparse or scattered shrubs and grasses have increased, implying that the shrubland canopy density in the drylands of Australia may have decreased. Grasslands have expanded, especially in arid areas, which may have also resulted from the degradation of closed shrublands. In addition, some rainfed crop areas in semiarid regions may have been replaced by rainfed pastures (Fig. 11.8).

Fig. 11.8
A positive-negative stacked bar chart plots land use cover type versus area trend in kilometers squared per year. It indicates that sparse trees in semiarid and closed shrubs in arid are decreased, and sparse and scattered shrubs and grasses, open shrubs, and open and closed tussock grasses are increased.

Interannual trends of different land use types in Australia’s drylands during 2002–2015

Full size image

Leaf area index (LAI) increased significantly in the relatively humid coastal forest regions but decreased in many arid areas that were dominated by shrubs or grasses during 2001–2018 (Fig. 11.9a). This finding is consistent with the land use cover changes in Australia, namely, expanded closed forests and degraded shrublands, and can be attributed to the ‘drier drylands, wetter wet areas’ climate change pattern. Accordingly, significant LAI gains (p < 0.01) occurred in the dry subhumid regions, whereas both the arid and semiarid regions experienced no significant LAI changes in recent decades (Fig. 11.9b).

Fig. 11.9
2 parts. A. A heatmap of Australia plots the L A I trend from 2001 to 2018. A color bar ranges between negative 0.12 and 0.15. Most of the central regions have mid-range values. B. A multiline graph of the annual mean L A I from 2001 to 2018 plots 4 curves. The fluctuating values of dry subhumid is the highest.

LAI changes in the drylands of Australia. a Spatial pattern of the interannual LAI trend during 2001–2018; shaded areas represent statistical significance at the 95% confidence level; b interannual variation in the LAI within different regions in the drylands of Australia

Full size image

During a longer period of 35 years (i.e., 1982–2016), the Advanced Very High Resolution Radiometer (AVHRR)-based vegetation continuous field (VCF) maps showed an increase in bare ground in most desert areas in Australia. On the northeastern coasts, tree cover declined, but short vegetation cover increased, whereas in southern Australia, tree cover greatly expanded (Fig. 11.10a). In the dry subhumid regions, bare ground significantly transitioned into short vegetation after 2006, but in drier areas (arid and semiarid regions), VCF changes were not significant throughout the study period (Fig. 11.10b–d).

Fig. 11.10
A map of Australia plots the V C F trend, and 3 multiline graphs of annual mean tree cover, mean short vegetation, and mean bare ground cover from 1982 to 2016 are labeled a to d, respectively. A. A few short vegetation pockets are noted over the eastern coast. B to D. All curves exhibit a fluctuating trend.

VCF changes in the drylands of Australia during 1982–2016. a Map showing the interannual trends of tree cover, short vegetation cover, and bare ground cover; b–d interannual variations in b tree cover, c short vegetation cover, and d bare ground cover in different dryland regions in Australia

Full size image

The drylands of Australia are dominated by bare soil (BS) and nonphotosynthetic vegetation (NPV), especially in the arid interior of the continent (Fig. 11.11a). During 2001–2018, in arid and semiarid regions, NPV, or both photosynthetic vegetation (PV) and NPV generally decreased, while the bare soil area expanded, indicating gradual vegetation degradation that was probably driven by the drier climate in the typical drylands of Australia. Conversely, on the eastern coasts and in southern Australia, where vegetation cover is denser owing to the relatively humid climates, PV, or both PV and NPV, increased, which agreed with the recent precipitation gains in most of those relatively wet areas (Fig. 11.11b). The interannual variations in PV, NPV, and BS are shown in Fig. 11.11c–e. In 2011, due to high precipitation, PV was the highest during the study period, while both NPV and BS were quite low in the drylands of Australia.

Fig. 11.11
A map of Australia plots vegetation fractional cover and its trends, and 3 multiline graphs of P V, N P V, and B S from 2001 to 2020 are labeled a to e, respectively. A. A few P V pockets are noted over the eastern coast. B to E. All curves, arid, semiarid, dry subhumid, and dry A U S, exhibit a fluctuating trend.

Vegetation fractional cover and its changes in the drylands of Australia. a Map showing the spatial distributions of photosynthetic vegetation (PV), nonphotosynthetic vegetation (NPV) and bare soil (BS); b spatial pattern of the interannual trends of vegetation fraction cover; c–e interannual variations in c PV, d NPV, and e BS in regions with different aridity levels in the drylands of Australia

Full size image

Eucalypts—often called gum trees—are iconic Australian flora. According to Australia’s State of the Forests Report 2013, ninety-two million hectares of the eucalypt forest type occur in Australia and form three-quarters of the total native forest area. The eucalypt native forest is distributed in all states and territories and across all except the most arid deserts (Fig. 11.12). Eucalypt plants provide a wide variety of resources that include food, shelter, refuge, and breeding sites for animals (Bennett 2016), as well as medicinal materials and wood for humans. Moreover, native eucalypt forests are important for the conservation of Australia’s rich biodiversity because they support many forest-dwelling or forest-dependent species of flora and fauna.

Fig. 11.12
A political map of Australia plots the distribution of native eucalypt forests. The coastal regions of North and West Australia, Queensland, New South Wales, and Victoria are scattered with eucalypt open forest and eucalypt woodland forest, while the coast of South Australia has eucalypt mallee forest.

Distribution of native eucalypt forests (Commonwealth of Australia 2016)

Full size image

4 Changes in Ecosystem Services

Ecosystem services in this chapter include ecosystem carbon sequestration, water yield, soil conservation, and crop production. In particular, ecosystem carbon sequestration is indicated by net primary production (NPP). Water yield is the difference between annual precipitation and ecosystem evapotranspiration. The average annual per-area NPP in the humid region is the largest at 950.1 gC/m2, followed by that in the dry subhumid area (723.3 gC/m2) and the semiarid region (405.3 gC/m2), and finally, it is the lowest in the arid region (142.0 gC/m2, Fig. 11.13a, b). The total annual NPP is, however, the highest in the semiarid region, at 808.7 TgC, followed by the arid region (714.3 TgC). Due to their small areas in Australia, dry subhumid and humid regions produced NPP values of only 279.1 TgC/yr and 367.8 TgC/yr, respectively, during the period of 2001–2018. In the drylands of Australia, no region experienced significant NPP gains over the whole period, but the significance (p = 0.07) and rate (k = 2.66 ± 2.86 TgC/yr2) of NPP increases in the dry subhumid region were larger than those in the semiarid region (p = 0.57; k = 0.49 ± 1.80 TgC/yr2) and arid region (p = 0.38; k = 1.76 ± 4.11 TgC/yr2) in Australia. Figure 11.13c shows that the total NPP in the drylands of Australia reached its highest in 2011, when precipitation was the largest during the study period and was significantly lower in dry years (e.g., 2002 and 2018). As shown in Fig. 11.13d, recent NPP increases are concentrated in the relatively humid coastal areas of Australia, whereas in the arid interior part, the NPP declines in most places, which can also be explained by the climatic temporal pattern.

Fig. 11.13
2 maps of Australia, a multiline graph, and an area chart. A. The average annual N P P is high over the eastern and southwestern coasts. B. The annual mean N P P in dry subhumid is the highest. C. The regional total N P P is low for dry subhumid. D. The N P P trend value is high over the northeastern coast.

Spatiotemporal pattern of NPP in the drylands of Australia: a spatial map of average annual NPP during 2001–2018; b mean per-area NPP changes in different regions in Australia distinguished by aridity; c interannual variation in total NPP in regions with different aridity levels in the drylands of Australia; and d map of the interannual trend of NPP during 2001–2018. Shaded areas indicate that the NPP trends are significant at the 95% confidence level. NPP data comes from the GLASS product suite (Li et al. 2013; Yuan et al. 2007, 2010, 2014)

Full size image

The annual water yield service was the highest in the tropical region located in northern Australia, mainly due to the high precipitation in that region (Fig. 11.14a). Evapotranspiration data in Australia are provided by the Australian Landscape Water Balance website (www.bom.gov.au/water/landscape). The average annual water yields in the arid, semiarid, and dry subhumid regions of Australia during 2000–2020 were 15.47 mm, 112.4 mm, and 245.5 mm, while the coefficients of variation (i.e., the ratios of the standard deviation to the average value) in these three regions were 1.93, 0.68, and 0.48, respectively. Accordingly, the water yield in Australia decreased as the aridity increased, but the interannual variability was stronger in more arid areas. In all drylands in Australia, the mean and standard deviation values of water yield were 59.9 and 45.4 mm, respectively. During 2000–2020, the drylands’ water yield was the highest in 2000, 2011, and 2020, which were the years with much rainfall, while the values were the lowest and even negative in the dry years of 2002 and 2019 (Fig. 11.14b). Over the whole study period, water yield declined significantly in northern Australia due to the reduced rainfall and decreased slightly and nonsignificantly in most of the interior dry areas of Australia (Fig. 11.14c).

Fig. 11.14
3 maps of Australia, and a multiline graph. A. The average annual water yield is high over a few pockets of the northern coast. B. The annual mean water yield in dry subhumid is the highest. C and D. The trend of annual water yield and annual soil loss by water erosion is high over the north and east coasts.

Water yield and water erosion in the drylands of Australia. a Spatial pattern of average annual water yield; b interannual variation in water yield in regions with different aridity levels in the drylands of Australia; c spatial map of the annual water yield trend from 2000 to 2020; and d map of annual soil loss induced by water erosion in the drylands of Australia

Full size image

It may be worth noting that the pulse in 2011 related to flooding that stored so much water that it was detectable in sea level globally, indicating the global significance of dryland in Australia. Australia contributed uniquely and substantially to the intensity and persistence of the global land hydrologic mass increase during the 2011 sea level drop. The persistence of Australia’s mass anomaly was attributed to the continent’s unique surface hydrology, which includes expansive arheic and endorheic basins that impede runoff to ocean (Fasullo et al. 2013).

By using the Australian soil loss map that shows water erosion calculated through the revised Universal Soil Loss equation (RUSLE) model (Teng et al. 2016; Viscarra Rossel et al. 2016), it is found that the soil erosion is high on the eastern coast and in the northwestern part of Australia and is the lowest in southern Australia (Fig. 11.14d). Moreover, the harvested area and production of primary crops increased at rates of 0.10 × 106 ha/yr and 0.54 × 106 t/yr during 2000–2018, according to food statistics from the Food and Agriculture Organization (FAO) of the United Nations (Fig. 11.15).

Fig. 11.15
2 column charts, labeled a and b, plot harvested area in 10 to the power of 6 hectares and production in 10 to the power of 6 ton, from 2000 to 2018, respectively. Both graphs exhibit a fluctuating trend, with the highest bars noted in 2017 with values of about 26 and 105, respectively.

Harvested area and production in the drylands of Australia: time series of harvested area (a) and production (b) during 2000–2018

Full size image

5 Driving Forces of Change in Dryland Social-Ecological Systems in Australia

5.1 Climate Trend

The temperature rose in the drylands of Australia after 1970, but precipitation showed a limited trend during 1900–2019. According to a 20-year period moving window, the variance of temperature followed a cycle of 20–25 years approximately before 1990 in the drylands of Australia, decreasing during 1919–1925, 1948–1965 and 1970–1990 (note: these are the middle times of 20-year periods), while increasing during 1925–1937, 1937–1948 and 1965–1970. However, after 1990, these cycles disappeared and were replaced by a very strong and consistent warming trend. In recent decades, i.e., 2000–2019, the arid region in Australia experienced a warming of 0.053 °C/yr, and for all drylands in Australia, the trend was 0.048 °C/yr, which were the fastest rates since 1910.

The 20-year period precipitation trends increased abruptly after 1940 and slightly declined after 1988. Recently (i.e., 2000–2019), precipitation declined at rates of −4.9 mm/yr and −5.2 mm/yr in arid regions and in all drylands of Australia, respectively. In Australia, precipitation and temperature are highly correlated, their coupling becoming stronger. Accordingly, the accelerated rise of temperature and reduced precipitation in the twenty-first century suggest that Australia is becoming drier. In addition, the solar radiation in the drylands of Australia has increased somewhat during the twenty-first century (Fig. 11.16a–f).

Fig. 11.16
9 line graphs, a to g, i, and j. A bar graph, h. A to J plot trends for annual temperature, trend of annual temperature and precipitation in each 20 year period, annual precipitation, interannual correlation, annual solar radiation, annual S P E I, number of droughts, and extreme high and low-temperature events.

Change in the Australian climate. a Interannual variation in mean temperature in different regions; b dynamics of the trend of annual temperature in each 20-year moving window in different regions. The x-axis labels are the middle years of the 20-year periods; c interannual variation in precipitation in different regions; d dynamics of the trend of precipitation in each 20-year moving window; e dynamics of the interannual correlation between temperature and precipitation in every 20-year period (note that the interdecadal Pacific oscillation (IPO) curve represents the average filtered interdecadal Pacific oscillation index in each period; see Sect. 11.4); f interannual variation in mean solar radiation in different regions; g interannual variation in the standardized precipitation evapotranspiration index (SPEI) in arid Australia; h number of drought years and drought quarters with different degrees during each 20-year period from 1920 to 2019; i average number of extreme hot noontimes and hot nights from 1910 to 2019 in different regions; note that the ‘-POP’ means the values are averaged and weighted by population; and j average number of extreme cold nights and extreme high precipitation days from 1910 to 2019 in different regions of Australia

Full size image

In Western Australia, the temperature rose significantly during 2000–2019, while in northern Australia, both solar radiation and temperature increased, and precipitation significantly declined in some areas. Moreover, southeastern Australia has recently experienced climate warming. Solar radiation increased significantly in northern Australia and in some parts of the central desert areas (Fig. 11.17a–c). By combining these three basic meteorological variables with the aridity index of the SPEI, most parts of dryland Australia have become drier, especially the northern areas where the drying trends were significant during 2000–2019 (Fig. 11.17d). Compared to the significant coupling between annual mean temperature and precipitation over drylands in Australia, the central-southern parts exhibiting negative correlations between annual temperature and precipitation were fewer in the last 20 years (Fig. 11.17e, f).

Fig. 11.17
8 heat maps of Australia, labeled a to h, plot the trends of climate change. Each map has a color gradient bar with a different range of values.

Spatial patterns of the climate changes in the drylands of Australia. a Map of temperature trends; b map of precipitation trends; c map of solar radiation trends; d map of standardized precipitation evapotranspiration index (SPEI) trends; e, f spatial patterns of the correlation between annual mean temperature and precipitation during e 1910–2019 and f 2000–2019; g, h trends of annual number of extreme (99%) g hot noontimes; and h hot nights from 1970 to 2019. The shaded areas indicate significant trends or significant correlations at the 95% confidence level

Full size image

Precipitation in drylands in Australia had a significant impact on that year’s LAI. Air temperature was negatively correlated with the annual mean LAI/NPP in the drylands of Australia, but the significance of the relationships was lower than that between the LAI and precipitation in all regions (Fig. 11.18a, b). In addition, wetting in the drylands of Australia promoted the LAI in the next year, suggesting a ‘lag effect’. For tree cover (TC), only in the arid region did the corresponding year’s precipitation exhibit a significant positive effect. However, precipitation rose very significantly (p < 0.01), promoting short vegetation (shrubs and grasses) cover in both arid and semiarid regions, indicating that short vegetation in the drylands of Australia is more easily influenced by precipitation changes than are trees. Owing to this strong relationship, bare ground cover in arid and semiarid areas was negatively correlated with precipitation (p < 0.01). Precipitation in the previous year also showed some effects, though not significant (Fig. 11.18c). Increases in precipitation can significantly promote photosynthetically active vegetation (PV) in all regions of Australia, especially in more arid areas. However, precipitation in the current year had a significant negative impact on senescent or dead vegetation (NPV) cover in semiarid and dry subhumid regions. Specifically, in arid regions and in all drylands of Australia, although the precipitation in a given year showed no significant effects, higher precipitation in the previous year could significantly (p < 0.01) promote the NPV cover, which can be explained by the fact that the drylands of Australia are largely covered by annual herbaceous plants. Finally, both the precipitation in the current year and that in the previous year had a strong negative effect on the bare soil cover in all regions of Australian drylands, especially in arid and semiarid places (Fig. 11.18d).

Fig. 11.18
4 grouped positive-negative charts. The bars for T E M current year and the previous year have negative values in arid, semiarid, and dry A U S, in both graphs a and b. The bar values of P R E current and previous year B G are negative in all regions in both graphs c and d.

Climatic impacts on the ecosystem in Australia: a interannual correlation coefficients between LAI and precipitation or air temperature in different regions distinguished by the aridity level; b correlation between NPP and precipitation or temperature; c correlation between VCFs (tree cover, short vegetation cover, and bare ground indicated by TC, SC, and BG, respectively) and precipitation in the current year and that in the previous year; and d temporal correlation between fractional cover (PV, NPV, and BS) and precipitation

Full size image

5.2 Changes in Extreme Climate Events

Not only extreme droughts but also the numbers of extreme (99%) hot noontimes and hot nights have significantly increased in the drylands of Australia since the late 1970s (Fig. 11.16i). Moreover, the number of extreme cold nights with the lowest 1% of all night temperatures did not change significantly throughout the 110 years, but in highly populated areas, there were fewer extreme cold nights after 1960 (Fig. 11.16j). The number of extreme high daily precipitation events in all drylands of Australia also increased rapidly, especially after the 1970s, indicating increases in both droughts and floods. However, in urban areas with high populations, extreme high precipitation events decreased after the 1970s. Eastern Australia experienced significant increases in annual extreme (99%) hot noontimes and nights over the past 50 years (1970–2019, Fig. 11.17g, h).

5.3 Dynamics of Fire Disturbance

Fire is a frequent and prevalent and usually the most important natural disturbance to ecosystems in the drylands of Australia (Bradstock et al. 2002). As the consequences of very different seasonalities in north and south Australia (shown in Fig. 11.2b), fire in northern tropical grasslands is certainly the most extensive (Fig. 11.19a), but the nature of fire is totally different in the north (annual grass fires) from the south (dominated by longer return time forest fires). Fires in the drylands of Australia have clear seasonal variation patterns. In 27% of the total dryland area in Australia, the largest fire events occur in January, when air temperature reaches the maximum in a year, while for 81% of the total dryland area, the largest area of burning occurs during spring to summer, i.e., September to January of the following year (Fig. 11.19b). The burned area percentage is the highest in the dry subhumid region of Australia, with an average monthly value of 1.21%, followed by the semiarid region (0.84%), the humid region (0.60%), and finally, the arid region (0.32%), indicating that burning in Australia relies on vegetation, the raw material for combustion, and the relatively arid weather, which are environmental requirements for fires.

Fig. 11.19
2 maps of Australia. A. It plots the average monthly burned area percentage, which ranges from 0 to 0.063. The values decrease from the north to the central regions. B. It indicates the month with the largest burned area. The eastern and southwestern regions have the most burned areas in March and April.

Fire distribution in the drylands of Australia. a Spatial pattern of the average monthly burned area percentage; b spatial pattern of the month with the largest burned area in a year averaged during 1996–2016. The monthly burned area is averaged during 1996–2016 using the Global Fire Emissions Database (GFED) version 4.1 (Randerson et al. 2018) and is further converted into the burned area fraction (i.e., the ratio of burned area to the total grid cell area)

Full size image

Burned area in the drylands of Australia showed a nonsignificant decrease during 1996–2016, yet large burned areas were observed in 2000–2002 and 2011–2012. It was also reported that between September 2019 and early January 2020, when droughts occurred, Australian mega forest fires led to unprecedented burned areas (Boer et al. 2020; Ward et al. 2020). The temporal variation in fire in the drylands of Australia was attributed to climate change, especially moisture conditions (Clarke et al. 2019; Kelley et al. 2019; Phillips and Nogrady 2020) and human controls (Liu et al. 2021). Precipitation can significantly promote burned areas in the following year within arid and semiarid regions, probably by increasing the fuel for combustion, whereas precipitation significantly reduces fires in humid regions and nonsignificantly inhibits fires in dry subhumid regions within the same year. In contrast, a nonsignificant positive impact of precipitation on fire occurred in arid regions in the same year (Fig. 11.20).

Fig. 11.20
2 graphs. A. A multiline graph plots the average monthly burned area percentage in each year from 1996 to 2016. It has fluctuating lines for arid, semiarid, dry subhumid, humid, and dry A U S. B. A grouped bar chart indicates that the correlation coefficient of pre-current and previous years is lowest in humid.

Fire dynamics in the drylands of Australia. a Interannual variation in average monthly burned areas in regions with different aridity levels in Australia; b correlation coefficients between burned areas and the precipitation in the current year and that in the previous year in different regions in Australia

Full size image

5.4 Ecosystem Management

The Australian government has invested abundant funds in pasture management (Barson et al. 2011) with advanced grazing and fire management systems, water resource management, systematic mechanisms of carbon farming and agroecological approaches to simultaneously support agricultural biodiversity and promote sustainable livelihoods. Moreover, these investments have achieved successes in ecosystem management methods of community engagement.

Different grazing modes were used to improve productivity, maintain desirable pasture species, and reduce land degradation. Native pastures can be managed through a number of grazing strategies, including continuous grazing, rotational grazing, cell grazing, time control grazing, and spell grazing (O’Reagain et al. 2014). Continuous grazing requires minimal labour and can deliver good production but is often accompanied by overgrazing, with livestock habitually revisiting preferred areas. Rotations are often organized around the plant growth cycles with the aim to optimize pasture utilization, which prevents uneven grazing and allows perennials to replenish their root reserves and better withstand dry periods, thereby benefiting both soil structure and land conditions (McDonald et al. 2019). Cell grazing and time-controlled grazing are similar to rotational grazing but are more intensive and involve more paddocks or ‘cells’ (McCosker 2000). Spell grazing involves locking up pastures at critical times in their growth cycle to allow plants to replenish root reserves and set seeds. This reduces the risk of overgrazing and encourages pasture plant recruitment through seed setting. Additionally, some feed additives are used to inhibit the microorganisms that produce methane in the rumen and subsequently reduce methane emissions. Overall, a successful grazing system should manage pasture utilization effectively (carrying capacity and timing of spelling), reduce uneven grazing, and match the stocking rate to the diet quality required by animal production targets (Hunt 2008).

To prevent soil degradation, soil and water erosion and plant/animal biodiversity losses caused by overgrazing (Hansen et al. 2019), the Australian government invested more than US$ 442 million to improve soil and biodiversity management practices on farms by November 2011 (Barson et al. 2011) and has supported the PROGRAZE™ educational systems (Martindale and Marriott 2004). These education and management practices (e.g., controlling the stocking rate, rotational grazing, preventing invasive plants and controlling pests (Hacker et al. 2019)) are intended to prevent soil acidification, maintain soil nutrients and protect ground vegetation cover, etc. (Waters et al. 2017). At least 50–70% ground cover, depending on the location, is recommended (Barson et al. 2011). Under grazing controls, the total grazing area in Australia has declined over the past 10 years, which was approximately 341 × 104 km2 in 2016–2017 but only 325 × 104 km2 in 2020 according to the Australian Bureau of Statistics. According to Fig. 11.21c, the head of livestock, especially sheep, has significantly decreased since 1991, indicating strong management of grassland grazing in the drylands of Australia.

Fig. 11.21
3 parts. A. A map of Australia indicates that the east and southwest coasts have grazing-modified pastures and non-grazing in the central parts. B. A stacked bar chart denotes that native vegetation in arid areas is the highest. C. An area chart denotes that the head value of sheep is the highest and falls over years.

Grazing extent in the drylands of Australia: a spatial distribution of grazed native vegetation and grazing-modified pastures; b areas of grazed native vegetation and grazing-modified pastures in different regions of Australia; and c heads of beef cattle and sheep in Australia from 1961 to 2019, according to official statistics

Full size image

Grazing and wildfires play important roles in tropical savanna management and industries in Australia, yet they both contribute to greenhouse gases in Australia. Fires consume vegetation and, in doing so, produce CO2 and other greenhouse gases. Greenhouse gases emitted from savanna fires average 3% of Australia’s emissions. Due to the large area percentage of rangelands in Australia (~80%), direct livestock emissions account for approximately 70% of greenhouse gas emissions in the agricultural sector and 11% of total national greenhouse gas emissions. These values make Australia’s livestock the third largest source of greenhouse gas emissions after the energy and transport sectors. Livestock are the dominant sources of methane (CH4) and nitrous oxide (N2O), accounting for 56% and 73% of Australia’s emissions, respectively.

Australia government has implemented climate change mitigation policies to reduce emission caused by the changing agricultural practices. Carbon farming is one of the most important policies contributing to the climate change mitigation, which aims to increase the amount of carbon stored in the soil and vegetation (sequestration) and to reduce greenhouse gas emissions from livestock, soil, or vegetation by changing agricultural practices or land use (Evans 2018). Carbon farming is important in Australia because agriculture accounts for 13.5% of Australia’s greenhouse gas emissions. To help reduce these emissions, the Carbon Farming Initiative in Australia was implemented in December 2011 to encourage land projects. Carbon farming potentially offers landholders financial incentives to reduce carbon pollution, which often generates economic and environmental ancillary (co)benefits (Tang 2016; Tang et al. 2018), such as improved soil quality, erosion prevention, better protection for stock, improved livestock production, native habitat creation for threatened species (Dumbrell et al. 2016; Kragt et al. 2016). Moreover, there still exist some barriers for carbon farming due to the lack of information and the government policy uncertainties (Kragt et al. 2017).

Prospective projects in Australia include savanna fire management and rangeland management (Kelly and Brotons 2017; Rolfe et al. 2021). The timing and intensity of fire, and consequently its various ecological impacts, can be significantly influenced by fire management activities. Effective fire management can reduce greenhouse gas emissions, protect fodder and infrastructure, and potentially attract payment for stewardship activity (Steffensen 2020). For thousands of years, indigenous Australians expertly managed the tropical savannas of northern Australia. They use fire to shape the landscape and to achieve social, economic, and spiritual well-being for the country. The fire management practices help protect the biodiversity, balance trees and grasses, mitigate emission, reduce infrastructure damage, and improve livelihoods (Bradstock 2010; Kelly and Brotons 2017; McKemey et al. 2022; Nikolakis and Roberts 2020; Steffensen 2020).

The Murray-Darling Basin (MDB) is an example of a complex river system undergoing substantial water reform to balance the needs of human and the environment. The basin extends across 4 states in south-eastern Australia, occupying 14% of Australia’s total surface area. Much of the basin is semiarid and contains 50% of Australia’s irrigated agriculture. Multiple efforts, such as the 2007 Water Act and 2012 Murray–Darling Basin Plan (MDBP), were issued to sustainably optimize social and environmental outcomes in relation to water use in the basin (Bischoff-Mattson and Lynch 2017). The basin plan includes five parts (i.e., balancing/sharing, monitoring, review, revision, and adaptation) and guides all stakeholders to use water in a sustainable way (Productivity-Commission 2018). Most importantly, the MDBP manages the basin as one system, enabling the river systems to adapt to climate changes and continue to support all water stakeholders in the long term. Although crisis-driven management to some extent prevents ‘economic and environmental decline’, the Plan makes no direct allowance for climate change, setting the scene for a future crisis that will trigger further reform (Colloff and Pittock 2022; Pittock 2019).

Australia claims to be pursuing a ‘green growth’ model in response to the global economic crisis and climate change. An agroecological approach supports agricultural biodiversity while promoting sustainable livelihoods (Lanka et al. 2017). Indigenous people living in Australia’s tropical savanna landscapes are increasingly searching for income opportunities from environmental services or ecosystem services as an avenue for economic development and improvement of socioeconomic conditions (Greiner 2010). The sustainable livelihood in Australia can better cope with and recover from stresses and shocks and promote its capabilities and assets both now and in the future (Davies et al. 2008; Moran et al. 2018).

5.5 Social and Economic Development

Population ageing is probably the most important social problem for Australia in recent decades, since the proportion of the elderly population nearly doubled from 8.24% in 1970 to 16.21% in 2020 (Fig. 11.22a), and was still increasing at a high rate (0.29%/yr during 2010–2020), bringing about challenges to social and economic development (Kendig et al. 2016; O’Loughlin et al. 2017), 2.23 million women (17.2% of all women) for the ages over 65 years, more than men in the same age (1.96 million, which is 15.4% of all men), were now living in Australia. Spatially, the percentage of elderly people was generally higher in southern Australia than in northern Australia and is the highest on the southeast and southwest coasts (Fig. 11.22b). The elderly proportion was similar among regions with different aridity levels, but the proportions of children and juveniles in arid regions were lower, resulting in fewer youth labourers, especially 20–29-year-old people, than in other regions. Rapid growth of the Indigenous Population is expected, with population momentum, identification change, and mixed partnering and childbearing shown to contribute more to growth than above-replacement fertility and increasing life expectancy. Since 1971 the indigenous population of Australia has trebled. From 1991 to 1996 numbers grew by 33%. The future growth of Australia’s Indigenous Population is thus intimately connected to its interaction with the Non-Indigenous Population (Wilson 2016).

Fig. 11.22
2 parts. A. A line graph denotes that the elderly population percentage falls to about 8 in 1970, and it exhibits an upward trend. B. A map of Australia plots the percentage of people older than 65, with values ranging from 0.02 to 0.42. Some pockets of the southwest and southeast regions have the highest values.

Change in the population structure in the drylands of Australia: a temporal change in the proportion of the elderly (>65 years old) population in Australia; and b the percentage of people older than 65 in all local government areas in Australia in gridded 2020 population maps are from the Gridded Population of the World (GPW) v4 dataset from the Socioeconomic Data and Applications Center (SEDAC). The land area per pixel values were resampled from ~1 km resolution to 0.05°

Full size image

In Australia, the population is now generally moving out of Western Australia (WA) and the Northern Territory (NT) into eastern and southeastern Australia, including New South Wales (NSW), Victoria (Vic.), and Queensland (Qld). Therefore, the general direction of internal migration in Australia is from relatively arid regions to more humid regions.

The GDP in the drylands of Australia has more than doubled since 1990, increasing from US$ 3.3 × 1011 to US$ 7.0 × 1011 in 2015 (Fig. 11.23). The GDP increase rates were somewhat higher in the semiarid and dry subhumid regions (3.3% and 3.2%/yr) than in the arid regions (2.5%/yr).

Fig. 11.23
An area chart of G D P in 10 to the power of 10 U S dollars versus years. The values of arid are the lowest, and they are almost constant, while the values of semiarid and dry subhumid increase.

GDP changes from 1990 to 2015 in regions with different aridity levels in the drylands of Australia. The gridded GDP is from Kummu et al. (2018)

Full size image

According to nighttime lights (NTL), socioeconomic activities in dry humid areas in Australia remained relatively stable during 2000–2010 but rose rapidly later, indicating that urban development in the drylands of Australia significantly accelerated after 2010 (Fig. 11.24b) but remained unchanged in drier regions.

Fig. 11.24
A. A map of Australia plots the average nighttime light. Few spots over the southwest and southeast coast have the highest values, ranging from 5 to 56.3, while most of the interior region has 0 values. B. A multiline graph of mean nighttime light versus years. The highest plot values are for dry subhumid.

Nighttime lights (NTL) in the drylands of Australia: a spatial pattern of NTL averaged during 2000–2020; and b mean NTL changes from 2000 to 2020 in regions with different aridity levels in the drylands of Australia

Full size image

6 Summary and Conclusion

Dryland social-ecological systems in Australia are facing the accelerated warming and rapid socioeconomic developments since the twenty-first century, including population and GDP increases and urban development since the twenty-first century, but with great diversity in space. In terms of spatial variance, forests in the drylands of Australia have become denser since the twenty-first century, but shrubs may have degraded. This result is consistent with the NPV, or both the PV and NPV, which generally decreased in arid and semiarid regions and vice versa in dry subhumid areas. Increases in the LAI/NPP were concentrated in the relatively humid coastal areas of Australia, whereas in the arid interior part, the LAI/NPP generally declined. Precipitation changes dominated the variation in vegetation in the drylands of Australia (legacy effects exist), where short vegetation is more easily influenced by precipitation changes than trees. Reductions in fire have significant impacts on emission mitigation and air purification but may have adverse effects on endemic biodiversity. Fire management (i.e., proactive burning) is necessary both to conserve biodiversity and to reduce the negative impacts on socioeconomic systems (fight fire with fire). The roles of livestock grazing/fencing in biodiversity are heterogeneous. Both grazing and fencing can be useful management tools to achieve conservation objectives and can also be threats to biodiversity conservation. Australia has invested considerably in improving biodiversity since the late 1980s. Integration of policy makers, funding agencies, and the general public are essential for the next step of dryland social-ecological system conservation in Australia.