Native plant species and growth substrates were sourced from a magnetite mining operation in the mid-west region of Western Australia, located 400 km northeast of Perth. This area experiences a semi-arid Mediterranean climate with mild wet winters (mean monthly maximum 19°C) and hot dry summers (mean monthly maximum 37°C). The study area receives 311 mm annual rainfall, about 65% of which falls between May and September (winter; Australian Bureau of Meteorology, http://www.bom.gov.au/climate/data/, Station 10195). This area also experiences 2400 mm of potential annual evapotranspiration, leading to an annual water balance of approximately -2100 mm.
The mining area is located in the Yalgoo Bioregion according to the Interim Biogeographic Regionalisation for Australia (version 7; http://www.environment.gov.au/land/nrs/science/ibra). The local plant communities are typical Eremaean sclerophyll shrublands (Beard 1990). Plant communities are generally low to open woodlands, predominantly comprising shrubs or trees of Acacia spp. (e.g., A. sibina, A. ramulosa var. ramulosa), Eucalyptus spp. (e.g., E. leptopoda, E. kochii), Melaleuca leiocarpa, Allocasuarina acutivalvis, Callitris columellaris, and Hakea recurva subsp. recurva, with an understorey of shrubs, grasses and herbaceous annuals (Markey and Dillon 2008).
Natural topsoil at the study area comprises highly-weathered red earth, from stony red loamy sand to loamy clay soil (Leptic Rudosol, Orthic Tenosol or Red Kandosol, according to Australian Soil Classification) containing abundant fragments (usually 2–20 mm) of ironstone gravel (Payne et al., 1998). All rocky material (>40 mm) and all large woody debris (branches and twigs) were removed prior to experimental use. Stockpiled topsoil used in the present study was sourced from stripped natural topsoil (top 15 cm) but was stockpiled for ca. 4–5 years. Dry-stacked magnetite tailings (hereafter referred to as tailings) were generated as fine-textured (processed by high-pressure grinding to <4 mm) waste materials at the end of magnetite-ore processing. Selected physical and chemical profiles for topsoil and tailings are presented in Cross et al. (2021b), with mineralogical characteristics presented in Wu et al. (2019).
Plant species selection and growth conditions
The experiment was conducted in a glasshouse at the University of Western Australia, Perth, between May and December 2018. Six plant species native to the region were selected for the study, including Acacia ramulosa (Fabaceae), Allocasuarina acutivalvis (Casuarinaceae), Austrostipa scabra (Poaceae), Eucalyptus loxophleba (Myrtaceae), Hakea recurva (Proteaceae), and Maireana georgei (Chenopodiaceae). These species are common components of native vegetation assemblages adjacent to the storage facility of the tailings at the study site, and represent a wide range of taxonomic and functional diversity including a variety of nutrient-acquisition strategies and symbiotic associations with soil biota (Table S1). These species were selected based on their performances in a study of 40 native plant species grow on Fe ore mine tailings (Cross et al. 2021a).
Seeds were purchased from Nindethana Seed Service Pty Ltd (Albany, Western Australia) and sown in plastic tapered square pots (18 cm x 18 cm x 44 cm) containing the experimental substrates. For A. ramulosa, H. recurva and M. georgei, 15 seeds were sown per pot, whereas equal weights of seeds were sown for the smaller-seeded species, A. acutivalvis (0.10 g per pot, ~35 seeds); A. scabra (0.10 g per pot, ~20 seeds); and E. loxophleba (0.02 g per pot, ~100 seeds). To yield comparable biomass between species, same numbers or weights of seeds from each species were sown initially. Plants of each species were grown in three treatment substrates including capped tailings, tailings, and stockpiled topsoil with 5 replicate pots for each substrate treatment. Capped tailings pots contained about 4.8 kg of tailings overlain by 1.2 kg of mixed topsoil and tailings (1:1 w/w), so as to mimic the proposed substrate design in the mine-closure plan. Topsoil and tailings pots contained about 6 kg pure stockpiled topsoil and tailings, respectively.
Within each substrate treatment, half of the pots were provided with a microbial inoculum (hereafter, referred to as inoculated). Each inoculated pot was treated with 100 ml of inoculant containing 2 g of a commercial freeze-dried microbial mix (Langley Fertilizers Troforte® Microbe Blend – Cropping, Sunpalm Australia Pty Ltd, Wangara, WA, Australia; Appendix 1 and 2 in Supplemental Materials), reconstituted in deionised (DI) water, at one and two months after seeds were sown. The fertilisation effect of the added inoculant can be neglected, given that it was diluted first by 100 ml of DI water and then by 6 kg of substrate. Additional unseeded pots were established as controls to compare changes over time and in contrast with seeded substrates and inoculation treatment. Water content was maintained at approximately 15-20% (water to filled pot weight) for all pots by manual watering through irrigation spikes (Products of Excellence Pty Ltd, Brookvale, NSW, Australia) placed in the middle of each pot.
Seedling emergence and the number of seedlings that survived to the end of the experiment were recorded for all pots. Plants were harvested at six months after sowing. Shoots were harvested by severing at about 0.5 cm above the soil surface, and lightly washed to remove any soil. Roots were retrieved by washing away the substrates under gently running water over a mesh grid (2 mm) to prevent loss of biomass. Roots were not harvested for treatment groups with poor survival, mainly in tailings and mix pots. Extra attention will be paid when interpreting the effect of plant growth on chemical changes of mined substrates. Harvested shoot and root mass was determined after drying the plant material to a constant weight at 70°C for at least 72 hours. Dry biomass is expressed on a per-pot basis.
Soil samples of ca. 500 g were collected after plant biomass was harvested, and sieved to remove large debris, roots and small gravel (2 mm stainless steel sieve) before being stored in zip-lock polyethylene bags at 4°C for up to one week prior to analysis of microbial biomass and mineral N measurements. For each sample, about 250 g subsamples were air-dried (35°C for up to one week) and stored in zip-lock polyethylene bags at room temperature for analysis of other chemical properties.
Soil microbial biomass carbon (MBC) and nitrogen (MBN) was measured by CHCl3 fumigation and 0.5 M K2SO4 extraction of fresh soil as described in Vance et al. (1987) and Brookes et al. (1985). Soluble organic carbon (C) in extracts was analysed using an Aurora O.I 1030W wet oxidation total organic carbon analyser (College Station, Brazos, TX, USA). A conversion factor (Kc) of 0.45 was applied to the MBC results as described in Wu et al. (1990). A conversion factor (Kn) of 0.54 was applied to the MBN results as described in Brookes et al. (1985). Soil microbial biomass P (MBP) was measured by CHCl3 fumigation and 0.5 M NaHCO3 extraction of fresh soil, as described in Brookes et al. (1982). Phosphorus concentrations were determined colorimetrically using a UV160A spectrophotometer at 880 nm (Shimadzu, Kyoto, Japan) after reaction with molybdate blue (Murphy and Riley 1962; Blakemore et al. 1987). A conversion factor (Kp) of 0.4 was applied to the MBP results as described in Brookes et al. (1982). Soil mineral N (NH4-N and NO3-N) was extracted with 2 M KCl from fresh soil (Clough et al. 2001), and determined by LACHAT FIA QuikChem 8500 Series 2 (Loveland, CO, USA).
Soil pH (both in DI water and in 0.01 M CaCl2) and electrical conductivity (EC) (deionised water) were measured for each sample using soil pH and EC probes calibrated with pH 4 and 7 buffer solutions (Orion 720a, Beverly, MA, USA). Soil total organic matter content was determined by the loss on ignition (LOI) method. Air-dried soil was dried in a crucible at 105°C in an oven for 12 hours, re-weighed when cool, placed in a muffle furnace, held at 500°C for four hours, and then re-weighed when cool. Soil organic C concentration was estimated by dividing the total organic matter (measured as LOI) by 1.72 (Blakemore et al., 1987). Total N was measured by the combustion method via a Leco analyser (FP628, St. Joseph, MI, USA).
Exchangeable cations were determined by extraction in 0.1 M BaCl2 (2 h, 1:30 soil-to-solution ratio), with detection by inductively coupled plasma optical-emission spectrometry (ICP–OES) (Thermo iCAP 6000 series ICP-OES, Freemont, CA, USA) following the methods of Blakemore et al. (1987). Effective cation exchange capacity (ECEC) was calculated as the sum of Al, Ca, K, Mg and Na concentrations. Colwell-P and -potassium (K) were extracted with 0.5 M NaHCO3 (adjusted to pH = 8.5) from air-dried soil (16 h, 1:40 soil-to-solution ratio). Phosphorus and K concentrations ([P] and [K]) in the extracts were determined colorimetrically (Rayment and Lyons 2010), using a UV160A spectrophotometer at 880 nm (Shimadzu, Kyoto, Japan) and atomic absorption spectroscopy (Varian SpectrAA 55A AAS, Palo Alto, CA, USA), respectively.
All soil data are presented as changes per total biomass using the difference of treated pots between unplanted controls divided by total dry biomass per pot.
Three-way analysis of variance (ANOVA) with plant species, inoculation, substrates, and the interaction of them included as fixed effects were used to test for the effects of species, inoculation, and substrate types on soil parameter changes per total biomass. This was followed by a nonparametric analysis (Wilcoxon test) by substrates to test the concatenated effects of plant species and inoculation on soil parameters. Student’s t tests were conducted between inoculated and non-inoculated plant biomasses within substrate type according to the data sets’ distribution normality and variance homogeneity. For each species, the differences in plant biomass produced in different substrates were not tested statistically because some species had less than three surviving individuals in tailings or capped tailings; also the greatest plant biomass was produced in the stockpiled topsoil. Data and statistical analyses were performed using JMP® 15 (SAS Institute Inc., Cary, NC, USA). All model results are summarised in Tables S2 and S3 in the Supplementary Materials.