Measuring, Reporting, and Verification of Forest Restoration

2.3.1 Rationale
Forest restoration will affect belowground and aboveground biodiversity via physical, chemical and biological ways by creating a range of different habitats and niches for diverse taxonomic groups of animals, plants and fungi. The restoration of trees, forest plant, moss and lichen communities will increase and diversify organic matter and nutrient inputs into the soil via the addition of leaf litter, root exudates and woody debris. Different forest structures will also create various microclimates and affect the physical structure of soils, via root growth, aggregation of soil particle types, aeration and the balance of water in the soil ecosystem .In addition to close ecological links with bacterial communities, the root systems of plants are intricately linked with fungi. Different species of mycorrhizal fungi, that form symbiotic links with plants to help with nutrient uptake will be closely associated with different plant species (Van Der Heijden et al 2007). Moving up the subterranean food chain, microbial eukaryotes (i.e. protists) are incredibly diverse and are known to graze upon bacterial communities (Adl et al. 2012). Microscopic metazoan meiofauna (e.g. Nematoda, Platyhelminthes 45-500mm) and finally macroinvertebrates (e.g. springtails, earthworms >500mm) will comprise different functional trophic groups of predators, detritivores, grazers and omnivores. Aboveground, forest restoration will increase habitat diversification by creating structural complexity via the establishment of tree canopy and understory layers but also the growth of smaller shrubs, herbs and groundcover plants (Mestre et al. 2017). The aboveground structural complexity will create an array of microhabitats, defined by different temperatures and light levels, and chemical drivers such as leaf and bark exudates. Deadwood also boosts biodiversity by creating unique habitats for a variety of saprophytic invertebrate species.

To develop a rigorous understanding of the reference condition trajectory for forest restoration the SUPERB project aimed to monitor a comprehensive and representative portion of above- and belowground biodiversity. This taxonomic diversity is in turn representative of successional ecological processes operating over different spatial and temporal scales to provide a detailed basis for assessment. SUPERB used molecular techniques to generate inventories of fungi, plants, microbial eukaryotes, meiofauna and arthropods from soil samples, as well as inventories of flying arthropods captured in Malaise traps.

2.3.2 Survey methods
Assessing biodiversity using traditional taxonomy approaches is labour intensive, time consuming and requires different specialists to identify different groups from the tree-of-life (Creer et al. 2016). Even within different taxonomic groups, taxonomists specialise on different groups (e.g. botanists, bryologists, nematologists) and in soils, the vast majority of microbial life cannot be identified using visual approaches (Curtis et al 2002). Microbiologists had been using molecular genetic approaches to identify biodiversity and construct phylogenies for several decades (Woese and Fox 1977), but a game changer for biodiversity characterisation was the standardised DNA barcode concept for the identification of animal, plant and fungal species, devised at the turn of the 21st century (Hebert et al. 2003; Ratnasingham and Hebert 2007). DNA barcoding uses standardised genetic markers to help identify species by the construction of curated large reference databases, enabled by voucher specimen deposition and open sharing of data, combined with an interactive web interface
(https://boldsystems.org/) and links to other global DNA databases. DNA barcoding uses highly degenerate DNA primers that will amplify a barcode marker from large numbers of species, using DNA extracted from single individuals. The barcode region has been selected to be variable enough to enable different species to be identified from each other. After 2005, through advances in high-throughput DNA sequencing, the field of ‘metabarcoding’ has evolved. Instead of generating single sequences from single specimens, metabarcoding uses similar degenerate DNA primers to amplify barcode regions, but from whole communities of biota (Creer et al. 2016). The standardisation of the DNA markers in DNA metabarcoding is not as routine as single specimen barcoding (Deiner et al. 2017), but in combination, the two approaches are revolutionising biodiversity discovery and assessment (Handley 2015). The genetic material for metabarcoding can come from environmental samples (e.g. soil, air, water), but also from communities or bulk sample DNA (Creer et al. 2016; Deiner et al. 2017). In SUPERB, we use soil eDNA and Malaise (a type of aboveground flying invertebrate sampler) (Geiger et al 2016) arthropod bulk community DNA to assess forest biodiversity.

To characterise the reference condition expected for future projects, the SUPERB protocol adopted a monitoring design that extracted DNA from five separate bulk arthropod malaise trap samples (Geiger et al 2016) and five topsoil samples, each a composite of five cores (5cm wide x 5cm deep). For the soil and arthropod sampling, we adopted standards generated by the LIFEPLAN project. For the malaise samples, the arthropods were first weighed (wet weight) and homogenised (Buchner et al. 2021). After DNA extraction, PCR amplification and library preparation are performed by 2-step PCR metabarcoding pipeline (Bohmann et al. 2021). The first round of PCR amplifies the DNA from the chosen barcode marker (e.g. ITS for plants and fungi, 18S for microscopic eukaryotes and CO1 for arthropods) and the second round of PCR incorporates known DNA ‘tags’ to each sample library so that each PCR can be tracked back to the original sample after sequencing multiple samples on the same Illumina sequencing run (Deiner et al. 2017; Bohmann et al. 2021). These techniques are increasingly prevalent in research by forestry agencies (e.g. Barsoum et al. 2018), but where possible it is recommended users seek advice from molecular biologists and ecologists to take full advantage of technological developments.

2.3.3 Data collection to data reporting
Metabarcoding generates large volumes of sequencing data in specialised file formats that are processed through a multi-stage bioinformatic pipeline to determine which taxa have been detected in which samples. One bioinformatic pipeline option is called DADA2, and this takes files through a series of quality control steps, filtering and cleaning the data, merging forwards and reverse sequencing reads and removing errors (Callahan et al 2016). The initial bioinformatic steps reduce the complexity of the dataset considerably to just the unique sequences detected from the environmental sample. To identify groups of sequences that are an approximation of species, algorithms are used to group together DNA sequences that are similar (SUPERB adopted a common standard of 97%) (Rojnes et al. 2016), and then compare the resulting clusters to a curated database of known sequences to find a matching taxonomic name. The resulting taxonomic data lists in which samples and stands a species was detected and can be used to calculate a range of common ecological diversity metrics.