Land degradation affects 10–20 % of drylands globally. Intensive land use and management, large-scale disturbances such as extractive operations, and global climate change, have contributed to degradation of these systems worldwide. Restoring these damaged environments is critical to improving ecosystem services and functions, conserve biodiversity, and contribute to climate resilience, food security, and landscape sustainability. Here, we present a case study on plant species of the mining intensive semi-arid Pilbara region in Western Australia that examines the effects of climate and soil factors on the restoration of drylands. We analysed the effects of a range of rainfall and temperature scenarios and the use of alternative soil materials on seedling recruitment of key native plant species from this area. Experimental studies were conducted in controlled environment facilities where conditions simulated those found in the Pilbara. Soil from topsoil (T) stockpiles and waste materials (W) from an active mine site were mixed at different proportions (100 % T, 100 % W, and two mixes of topsoil and waste at 50 : 50 and 25 : 75 ratios) and used as growth media. Our results showed that seedling recruitment was highly dependent on soil moisture and emergence was generally higher in the topsoil, which had the highest available water content. In general, responses to the climate scenarios differed significantly among the native species which suggest that future climate scenarios of increasing drought might affect not only seedling recruitment but also diversity and structure of native plant communities. The use of waste materials from mining operations as growth media could be an alternative to the limited topsoil. However, in the early stages of plant establishment successful seedling recruitment can be challenging in the absence of water. These limitations could be overcome by using soil amendments but the cost associated to these solutions at large landscape scales needs to be assessed and proven to be economically feasible.
Land degradation affects nearly 2 billion hectares of land globally, with 25 % of the total global land considered degraded to some extent (Barbero-Sierra et al, 2015; Bisaro et al., 2014; Brevik et al., 2015; Miao et al., 2015; Stanturf et al., 2015; Torres et al., 2015; Wang et al., 2015). Restoring these damaged environments is critical to improving ecosystem services and functions, conserve biodiversity, and contribute to climate resilience, food security and landscape sustainability at the local, regional and global level (de Moraes Sá et al., 2015; Minnemeyer et al., 2011; Perring et al., 2015; Prosdocimi et al., 2016; Roa-Fuentes et al., 2015; Zucca et al., 2015). Drylands, which include semi-arid and arid environments, are particularly vulnerable to land degradation with estimates suggesting 10–20 % of these ecosystems are degraded globally (Millennium Ecosystem Assessment, 2005; Safriel et al., 2005), and continue to be degraded across millions of hectares every year (Brauch and Spring, 2009; Wang et al., 2015; Yan and Cai, 2015). Intensive land use and management, large-scale disturbances such as extractive operations (e.g. mining), and global climate change have contributed to degradation of these systems worldwide (Anaya-Romero et al., 2011; Keesstra et al., 2016a; Kildisheva et al., 2016; Muñoz-Rojas et al., 2015).
When attempting to restore degraded, arid ecosystems challenges include limited rainfall, high temperatures, and soils with low nutrient levels and water holding capacity (Anaya-Romero et al., 2015; Muñoz Rojas et al., 2016a). Thus, despite the efforts and investments to restore these systems worldwide (Keesstra et al., 2016a), restoration of drylands has low rates of success (James et al., 2013; Sheley et al., 2011). To improve our capacity to reinstate biodiverse, viable plant communities, there is a strong need to advance our understanding of how these systems function and the effects that environmental and edaphic factors have on processes such as seedling emergence and plant growth and survival (Perring et al., 2015). For example, changes in soil water availability as a consequence of reduced rainfall and evaporation, or increases in temperature due to global warming, may affect restoration outcomes through influencing seedling recruitment (Cochrane et al., 2015; Lloret et al., 2004) or the composition and distribution of plant species (Lai et al., 2015). But the impact of environmental factors on restoration can be also compounded by unfavourable edaphic conditions (Audet et al., 2013; Thomas et al., 2015). Thus, improving soil physical and chemical properties can be decisive for successful revegetation (Machado et al., 2013), which is important in extractive industries operating in dryland environments.
During open-cut and strip mining operations, the top layer of soil is commonly removed and stockpiled before starting the extraction process and then respread before seeding the target sites for restoration (Lamb et al., 2015; Rivera et al., 2014). This topsoil is an important source of seeds, nutrients and microorganisms (Erickson et al., 2016a; Golos and Dixon, 2014; Koch, 2007; Muñoz-Rojas et al., 2016b) but its use in restoration is often limited by its scarcity and the detrimental conditions that topsoil stockpiling can have on soil functionality (Keipert et al., 2002). Waste materials produced in mining operations provide alternative substrates that are currently being used as growth media in restoration (Machado et al., 2013; Muñoz-Rojas et al., 2016b; Thomas et al., 2015). These substrates can integrate coarser materials that help to reduce slope instability and prevent erosion processes, but they are often highly deficient in organic matter which can reduce soil water retention (Shrestha and Lal, 2006). In addition, developing appropriate soil structures for restoration, for example technosoils, can be expensive and demanding in terms of time and natural resources (Rivas-Pérez et al., 2016).
Where topsoil is limiting and waste materials form the substrate for plant growth, direct seeding is the most feasible means of reinstating biodiverse plants communities, particularly at larger scales (Ceccon et al., 2016; Erickson et al., 2016a; James et al., 2011; Perring et al., 2015; Porensky et al., 2014). However, direct seeding is inefficient in terms of the proportion of seeds that produce an established seedling; in arid ecosystems it is common for only 2–7 % of seeds to establish (Chambers, 2000; James et al., 2011; Larson et al., 2015), although the use of biochar has shown to increase these percentages (Drake et al., 2016). The early developmental life-stages of plants are usually more sensitive to environmental or edaphic constraints than the adult stages (Standish et al., 2012) and the transition from germinated seed to emerged seedling has been identified as the life-stage transition most limiting the success of direct seeding (James et al., 2011). As these first stages of plant regeneration fundamentally influence the composition of the future plant community (Jiménez-Alfaro et al., 2016), characterising abiotic factors of the edaphic environment and their effects on seeds and seedlings is necessary for developing seeding practices that can achieve the desired restoration outcomes. For example, highly erodible soils have proved to be an additional challenge for seed germination and consequently for obtaining an adequate plant cover (Bochet, 2015; Cerdà and García-Fayos, 1997, 2002; Wang et al., 2014).
With the numerous potential drivers of global change comes a wide range of
potential climate change scenarios (IPCC, 2014). This hinders the
incorporation of future climate predictions into restoration programs
(Standish et al., 2012). In this context, more experimental studies are
needed to accurately evaluate the effects of altered climatic conditions on
seedling recruitment and subsequent vegetation community structure and
function, all of which, in turn, are strongly linked to soil conditions
(Audet et al., 2013). Such experimental approaches can be effectively
addressed by manipulation of combinations of climate and soil factors under
controlled conditions (Lloret et al., 2004). Here, we present a case study
on plant species of the Pilbara region in the northwest of Western
Australia, where we assess the effects of climate and soil factors on the
restoration of semi-arid ecosystems. The Pilbara (22
This study was conducted between August and December 2014 in a controlled
environment room (CER) at The University of Western Australia (UWA) and a
glasshouse facility at Kings Park and Botanic Garden in Perth, Western
Australia. Five native plant species from five families were selected as
representative of a diverse range of life-forms (e.g. perennial grass, shrub
and tree components) that commonly contribute to the mature plant
communities found throughout the mining intensive Pilbara region of Western
Australia (Erickson et al., 2016b). These study species comprised
Two experimental studies were carried out to test different climate and soil
scenarios. The climate in the Pilbara region is semi-arid with mean annual
rainfall ranging between 250 and 400 mm, mostly concentrated in the summer
months (December to March), accounting for approximately 72 % of the total
annual rainfall. This rainfall originates from sporadic summer convection
thunderstorms and tropical cyclones. Mean annual temperatures range between
19.4 and 33.2
Simulated rainfall (watering) treatments utilised in this study. Pulse durations and rainfall amounts were selected from interrogating long-term weather data accessed from the Newman Airport weather station (CSIRO and Bureau of Meteorology, 2007). Simulated rainfall treatments (R1–R4) comprised repeat daily applications of water for either 6 or 3 days and two different rainfall amounts (20 or 10 mm). The total irrigation amount of 50 or 25 mL matched the pot sizes used in this study and rainfall amount treatments required to simulate the desired simulated rain conditions.
Soil physicochemical and hydraulic properties of growth
media types (mean
For the soil scenario experiment, a range of growth media blends were
evaluated to assess the feasibility of using growth media mixes in
restoration sites. These growth media consisted of four different blends of
the soil materials collected from the mining sites: 100 % topsoil (
Topsoil and waste material from the mine site were collected and transported
to the CER facilities at UWA and Kings Park in 200 L drums. To create each
growth media combination, one drum of topsoil (
Soil hydrological parameters (Table 2) were determined according to Conant
et al. (2004) using a pressure plate device at four tensions between
saturation (
Seeds for each species were obtained from commercially collected seeds
supplied to the mining industry for use in Pilbara restoration programs.
Upon receipt at Kings Park and Botanic Garden, seeds of
To maximise the germination potential of each batch and accommodate seeds
with primary dormancy, seed pre-treatments followed pre-treatment
recommendations in Erickson et al. (2016a). Seeds of
The climate scenario experiment was conducted in the CER at UWA, where
temperature, CO
For both experiments, pots of 25 cm
Seedling emergence was recorded daily in each pot for 16 days. Final emergence (%) was
determined as the average emergence per pot after 16 days divided by
five (the number of seeds per pot) and mean emergence time (MET) was
calculated using the following equation adapted from Ellis and Roberts (1981):
Total seedling emergence (%, mean
Differences in seedling emergence (final proportion of emerged seedlings
among climate and soil scenarios) and time to emergence among treatments
were tested using analysis of variance (ANOVA). Comparisons between means
were performed with the Tukey's HSD (honestly significant difference) test
(
Our results showed that seedling emergence of the Pilbara native plant
species was highly dependent on soil water content in the topsoil growth
media (Table 3). Total emergence varied significantly across plant species
and water treatments (
Seedling emergence for
Effects of climate factors (temperature and water) and
plant species types, and interactive effects of these factors on total
emergence and mean time to emerge. Statistical significance levels: NS: not
significant,
Mean time to emergence (days, mean
Overall, our results showed that rainfall patterns had a large influence on
seedling emergence across the five native species and suggest that seedling
recruitment of these native plants may decrease in a climate scenario of
increasing drought. These results are broadly consistent with other similar
studies conducted in seasonally dry environments. For example, Lewandrowski (2016)
found that seedling emergence of
In our study, seedling emergence responses to the watering regimes differed
significantly among the five species. We found significantly decreased
emergence of seedlings of
The mean time for emergence of the five plant species was significantly
different across temperature and rainfall treatments with slightly shorter
times recorded under higher temperatures, particularly in
Regardless of plant species or temperature conditions, our results showed significantly higher rates of emerged seedlings with longer pulses of simulated rainfall (6 days compared to 3 days) with the same amount of accumulated water during the treatment (60 mL over the irrigation phase). Semi-arid ecosystems are particularly influenced by precipitation patterns, and water availability in these environments can be highly pulsed with discrete rainfall events followed by drought periods (Miranda et al., 2001). Therefore, changes in precipitation frequency, such as rainfall pulses, can have a stronger effect than rainfall quantity in these environments (Woods et al., 2014).
Total seedling emergence (%, mean
Another factor that might affect plant production in global climate change
scenarios is the elevated concentration of atmospheric CO
Seedling emergence differed significantly between growth media types,
watering treatments and plant species, but the effect of water inputs seemed
to be a larger driver of emergence than growth media type (
Effects of soil or growth media type, water treatments and
plant species, and interactive effects of these factors on total emergence
and mean time to emerge. Statistical significance levels: NS: not
significant,
The analyses of soil physio-chemical properties showed lower contents of
sand in the topsoil growth media (70.5
Our study showed that seedling emergence across the five plant species was higher in the topsoil growth media which might be explained by the greater water availability as a consequence of larger amounts of organic C content (Table 2). Although additional factors, such as adequate nutrient levels in the soil, can be necessary for plant establishment in degraded soils (Valdecantos et al., 2006; Brevik et al., 2015), water availability seems to be more critical at early plant life stages, particularly in semi-arid environments (Cortina et al., 2011; Miranda et al., 2001).
The use of growth media such as waste materials has proved to be a competent alternative to the original soil (i.e. topsoil) in restoration of degraded semi-arid areas (Machado et al., 2013; Muñoz-Rojas et al., 2006b; Rivera et al., 2014). Muñoz-Rojas et al. (2016b) showed that soil functions in a rehabilitated area of northwest Western Australia, with the use of mine waste material, can reach levels of microbial activity and organic C similar to those of topsoil once vegetation was established. However, here we show that at the early stages of plant recruitment, the use of alternative substrates depleted of organic materials can be challenging for successful seedling recruitment in the absence of water. Low contents of soil OC have been commonly associated to the loss of soil structure, which as a consequence, diminishes water holding capacity, increases bulk density, and accordingly produces soil compaction (Lal, 2004; Willaarts et al., 2016).
Overall, the results obtained in this study provide evidence that the availability of water in the soil system is a key determinant factor for increasing seedling recruitment and, therefore, optimising restoration of semi-arid lands such as the Pilbara. The application of irrigation has been proposed in restoration of semi-arid systems to control watering inputs (Bainbridge, 2002). There are several types of irrigation systems available that could effectively increase seedling recruitment, particularly in plant species most sensitive to water limitations (Padilla et al., 2009). However, there are higher costs associated with this alternative that makes its use impractical at the landscape level (Cortina et al., 2011).
Degraded soils – frequently infertile and depleted of organic materials – can respond positively to the addition of amendments (Cortina et al., 2011; Keesstra et al., 2016b; Lozano-García et al., 2011; Valdecantos et al., 2006). Soil amendments have been commonly used in restoration to improve soil structure, restore the hydrological balance and increase the mineral nutritional capacity (Hueso-González et al., 2014; Jordán et al., 2011). Inorganic amendments (e.g. fertilisers) are usually applied to overcome plant nutritional deficiencies or physical limitations. However, the use of organic amendments such as mulch or manure has proved to increase soil water retention in soils with poor structure with a consequent increase of plant survival in mine restoration (Benigno et al., 2013). Even low doses of composted organic waste applied in degraded soils have shown to support seedling response for long periods (Fuentes et al., 2010; Yazdanpanah et al., 2016). Nevertheless, the application of organic amendments can have several implications such as competition with existing species which is compounded by the high costs of these practices at large scales in mine restoration (Cortina et al., 2011).
Since seedling establishment from seeds can be challenging in restoration
(James et al., 2011), increasing seed input, or enhancing the availability
of suitable micro-sites for seedling emergence through modifying the soil
environment or alternatively improving the regenerative capacity of seeds
represent alternative strategies for those species with limited recruitment
(e.g.
Seedling recruitment of the five native plants was highly dependent on soil
moisture and temperature did not have a significant effect on the number of
emerging seedlings. Emergence across the five plant species was higher in
the topsoil growth media compared to the other soil materials, most likely
due to its larger available water content as a consequence of increased
amounts of organic C. Overall, under drought scenarios total seedling
emergence was below 40 % for all species and growth media types. In
general, responses to the climate scenarios differed significantly among the
five native species suggesting that future climate scenarios of increasing
drought might affect not only seedling recruitment, but also diversity and
structure of native plant communities. In particular, we found significantly
decreased emergence rates in seedlings of
This research was supported by a BHP Billiton Iron Ore Community Development Project (contract no. 8600048550) under the auspices of the Restoration Seedbank Initiative, a partnership between BHP Billiton Iron Ore, The University of Western Australia, and the Botanic Gardens and Parks Authority. Edited by: S. Keesstra