Fauna has long been neglected in the monitoring of ecological restoration, despite the key role they play in ecosystem function. Vertebrate surveys can be time consuming and costly, often requiring multiple methodologies and taxonomic expertise, making comprehensive monitoring cost prohibitive. Here we evaluate a new method of assessing mammal and bird diversity through the genetic identification of scat collections. Using DNA metabarcoding of scat collections from three bioregions we generated bird and mammalian assemblage data and distinguished between sites with different restoration histories. However, scat detectability was affected by environmental conditions (e.g. rainfall and soil), suggesting that our approach is most applicable at certain times of year or in arid (or semi-arid) environments with rocky soils, where conditions are favourable for scat preservation. Taken together these data provide a pathway to: plan, monitor and establish best-practice when restoring landscapes and add to the growing body of literature on the value of DNA metabarcoding in biomonitoring applications.

1. Study Sites

Samples were collected at three locations. The Swan Coastal Plain (SCP) and Jarrah Forest (Eucalyptus marginata) (JF) Bioregions are both within the Mediterranean-climate Southwest Australian Global Biodiversity Hotspot (Myers et al. 2007), while the third is in the semi-arid Pilbara Bioregion (PB). At each location, we sampled two restoration sites under nine years old, two restoration sites over 9 years old, and two reference sites (Figure 1). Therefore, a total of 18 sites were targeted for scat collection. 

The Banksia Woodland of the Coastal Plain is a biodiverse ecosystem with a Mediterranean climate with cool, wet winters and hot dry summer. The area has a mean annual minimum temperature of 12.8°C, a mean annual maximum of 24.7°C, and a mean annual rainfall of 757 mm (Australian Bureau of Meteorology). The dominant tree species are Banksia attenuata and B. menziesii, with less dominant Eucalyptus todtiana and Nuytsia floribunda. The understory consists of woody species of Myrtaceae, Fabaceae, Proteaceae, and Ericaceae, and non-woody species in Anthericaceae, Stylidaceae, Cyperaceae, and Haemodoraceae (Trudgen 1977). In October 2018, we sampled six sites at a Hanson Construction Materials sand quarry in Lexia (-31.76°, 115.95°); two reference sites and restoration sites 1, 3, 14, 22 years old. The four restoration sites were established to restore mined areas to surrounding native Banksia woodlands. All restoration was done by Hanson and previous mine owners and included seeding with native species and planting of dominant tree species (Banksia attenuata, B. menziesii, and Eucalyptus todtiana). Plant species richness and density tends to be higher in restoration sites, and percent cover increases with restoration age and is highest in reference sites (Benigno et al. 2013). According to a previous study looking at banksia woodland restoration, birds colonize rapidly and there is little difference in richness or assemblages between restoration and reference (Comer and Wooller 2002). There is less data available of mammalian recovery; some of the more common animals include the Western Grey Kangaroo (Macropus fugilinosus) and the Australian Raven (Corvus coronoides) (www.ala.org.au).

The Jarrah location has a similar Mediterranean climate to the Coastal Plain site. The area has a mean annual minimum temperatures of 8.6°C, a mean annual maximum of 23.7°C, and mean annual rainfall of 668.9 mm (Australian Bureau of Meteorology). The overstory vegetation is dominated by E. marginata, with E. patens, and E. wandoo also common. The understory is sclerophyllous and floristically diverse, dominated by several families (i.e. Fabaceae, Asteraceae, Proteaceae, Dasypogonaceae, Myrtaceae) (Havel 1975). We sampled six sites from the bauxite mine South32 (-32.96°, 116.48°) in October 2018; two reference sites and restoration sites 2, 6, 11, 20 years old (Figure 1). All restoration was undertaken by South32 or the previous mine owners and included seeding with over 100 native species. Reference and restoration sites are dominated by species of Myrtaceae and Fabaceae. Total cover increases with age of restoration and reaches similar cover percentages to reference sites. There are uncontrolled feral animals including Sus scrofa (pig), and Vulpes vulpes (fox) (Williams and Mitchell 2002). Fauna surveys indicate bird diversity increases with age of restoration and a greater abundance of Carnaby’s black-cockatoo (Calyptorhynchus latirostris) and new-holland honeyeater (Phylidonyris novaehollandiae) in restoration sites (Data from South32). Mammalian richness remains similar throughout restoration, although the restoration sites have a higher abundance of house mouse (mus musculus) and Western Grey Kangaroo (M. fugilinosus) (Data from South32).

The open woodlands of the Pilbara occur in a hot, arid location of north-western Australia with most rainfall occurring in the summer, associated with cyclonic activity (McKenzie et al. 2009). Temperatures have a mean annual minimum of 15°C and mean annual maximum of 30.6 °C, with a mean annual rainfall of 263.8 mm (Australian Bureau of Meteorology). The open snappy gum (E. leucophloia) woodland occurs over hummock grasses (Triodia wiseana, T. basedowii, T. lanigera) and low acacia shrubs (McKenzie et al. 2009). This site is located in a region that includes globally significant iron ore resources and accounts for up to 39% of global iron ore production (Government of Western Australia 2019). We sampled 6 sites at a BHP iron ore mine (-22.84°, 118.95°) in September 2018, two reference sites and restoration sites 4, 7, 11, and 15 years old (Figure 1). These sites are within 1 km of several ephemeral creeks. Restoration was conducted by the mine owners and restoration areas tended to have higher coverage of woody shrubs (Acacia), while reference sites and older restoration areas has more hummock grasses (Triodia) and mature gum (E. leucophloia) (Data from BHP). There is limited data in the Pilbara, none of the 31 studies examining vertebrate recovery in Australasia used sites restored post iron-ore extraction (Cross et al. 2019a). Some of the more common animals in the area, according to the Atlas of Living Australia (www.ala.org.au) include the Pebble-mound mouse (Pseudomys chapmani) and the Galah (Eolophus raseicapilla)

2. Sample Collection

Scat samples were collected from 4 randomly chosen points at each restoration/reference site. At each point, scat was collected over an 8-minute period along a 200 m transect, at least 5 m from the edge of the site. A portion of all scats seen during this transect were collected; herbivore, carnivore, and bird, no matter how degraded. Latex gloves were used in collection to minimize contamination and new gloves were used for each sample.  A total of 72 scat samples were collected, 24 from each location. Scat samples were frozen as soon as possible in a mobile freezer and transported to the Trace and Environmental DNA lab where they were stored at -20°C until processed.  

3. Sample processing

A portion of each scat in the sample was put in a 50 mL falcon tube. Scats that could be easily identified was noted for each sample, to provide a morphological comparison to sequencing results. These included identifying Emu scat because it is very distinctive, bird scat (not including Emu), and kangaroo and wallaby scat. Scat was homogenized in 50mL falcon tubes using a TissueLyser (Qiagen) with 4 steel balls (5mm) for 3 minutes in 30 second intervals.  DNA was extracted from 250uL scat homogenate, using the DNeasy PowerLyzer PowerSoil kit (Qiagen) on the QiaCube Connect automated platform (Qiagen). A final elution volume of 100 μL (Tris buffer) was used, and extraction controls (blanks) were included in every set of extractions.  Quantitative PCR (qPCR) was run on neat extracts and a 1/10 dilution to see if samples exhibited inhibition, and to determine optimal DNA input for each sample (Murray et al. 2015). The qPCR assays were run with primers that target the 12S gene 12SV5-F/R (Riaz et al. 2011). This assay is both short enough (98 bp), to pick up degraded DNA and broad enough to amplify mammalian, avian, and reptilian DNA.

The PCR mix for quantitation contained: 2.5 mM MgCl2 (Applied Biosystems, USA), 1× PCR Gold buffer (Applied Biosystems), 0.25 mM dNTPs (Astral Scientific, Australia), 0.4 mg/ml bovine serum albumin (Fisher Biotec, Australia), 0.4 μmol/L forward and reverse primer, 1 U AmpliTaq Gold DNA polymerase (Applied Biosystems) and 0.6 μl of a 1:10,000 solution of SYBR Green dye (Life Technologies, USA). The qPCRs were run on a StepOne Plus (Applied BioSystems) real-time qPCR instrument with the following conditions: 5 min at 95°C, 40 cycles of 95°C for 30s, 30s at the 60°C and 45s at 72°C, a melt curve stage of 15s at 95°C, 1 min at 60°C and 15s at 95°C, ending with 10 min elongation at 72°C. Extraction controls were included in qPCR assays. DNA dilutions that showed uninhibited amplification (Murray et al. 2015) were selected for further analysis.

The DNA extracts that were deemed of sufficient template number and free of inhibition (qPCR described above) were assigned a unique combination of multiplex identifier (MID) tags. These MID tags were incorporated into fusion tagged primers, and, to prevent cross contamination, none of the primer-MID tag combinations had been used previously in the lab. Fusion PCRs for scat samples were done in duplicate and the PCR master mixes were prepared in a dedicated ultra-clean lab before DNA was added to minimize the potential for contamination. The PCR cycling conditions were as stated above for the quantification qPCRs. Samples that amplified successfully were then pooled into approximately equimolar concentrations to produce a final metabarcoding library that was size-selected (150-450bp) using a PippinPrep 2% ethidium bromide cassette (Sage Science, Beverly, MA, U.S.A).  Libraries were cleaned using a QIAquick PCR Purification Kit (Qiagen, Germany) and final library quantification with a Qubit Fluorometric Quantitation (Thermo Fisher Scientific). Sequencing was performed on the Illumina MiSeq platform using the 300 cycle V2 kit following the manufacturer’s protocol.

4. Sequencing analysis

Bioinformatics and taxonomic assignments were performed on the Pawsey Supercomputer. Sequences were demultiplexed using OBITools (Boyer et al. 2016). Further sequence processing carried out in R using the “DADA2” package (Callahan et al. 2016) where sequences were quality filtered, the error rates were estimated, and the sequences were dereplicated. The error rates are then used in the sample inference stage to remove sequences likely to be errors and leave Amplicon Sequence Variants (ASV). These ASVs are equivalent to zero radius operational taxonomic units (ZOTUs) in usearch (Edgar 2016). The sequence table was constructed, chimeras removed, and curated using LULU (Frøslev et al. 2017). Taxonomy was assigned using BLASTN to search against the GenBank reference database (https://www.ncbi.nlm.nih.gov/genba nk/). Minimum percent coverage was set at 97%, minimum percent identification at 95%, and taxonomic identification was dropped to the lowest common ancestor if sequences matched with more than one taxa. Any identifications with less than 97% identity were dropped to family level. Taxonomic identifications were validated against the Atlas of Living Australia database (www.ala.org.au) to verify the organisms occurred in the area. If an identification did not occur in the area, the identification was dropped to the lowest common ancestor that occurred in the area or removed from the dataset (See Supplementary table 1 for details).