The use of analogue sites for native ecosystem mine rehabilitation—a case study incorporating effective approaches for selection, monitoring and analysis

Authors: Spain, CS; Jones, JL; Nuske, SJ; Robinson, D Download Paper Open access courtesy of:

DOI https://doi.org/10.36487/ACG_repo/2315_080

Cite As:
Spain, CS, Jones, JL, Nuske, SJ & Robinson, D 2023, 'The use of analogue sites for native ecosystem mine rehabilitation—a case study incorporating effective approaches for selection, monitoring and analysis ', in B Abbasi, J Parshley, A Fourie & M Tibbett (eds), Mine Closure 2023: Proceedings of the 16th International Conference on Mine Closure, Australian Centre for Geomechanics, Perth, https://doi.org/10.36487/ACG_repo/2315_080



Abstract:
Evaluating the success of any native ecosystem rehabilitation requires its careful comparison to a desired target. In mine rehabilitation, these targets are often native vegetation communities that were present pre-mining or in the surrounding landscape. Field monitoring plots representative of these community targets are known as analogue or reference sites and are widely used in rehabilitation efforts, yet there is a dearth of literature available on their effective use in mining contexts. Using Whitehaven Coal’s rehabilitation operations as a case study, we explore considerations for effective analogue site selection, monitoring and data analysis in the context of native ecosystem mine rehabilitation. To identify the most appropriate targets for the existing or desired mature rehabilitation at Whitehaven operations, an evaluative comparison of landform pattern, location, floristic and structural similarity between the rehabilitation areas and native vegetation communities was conducted. After accounting for operation-specific mine rehabilitation objectives, the best matching candidate communities, based on landscape and rehabilitation context, were selected as final targets. Spatially replicated analogue sites, representing the best-on-offer examples of the targeted ecological community, were established and subjected to an ongoing monitoring program focused on appropriate characterisation of the communities, given temporal and spatial variations in species cover and detectability that occur in the targeted woodland ecosystems. This variability is captured through multiyear resampling of a subset of analogue sites, thereby efficiently providing quantitative data on climatic responses to community composition. The resultant multi-year ‘library’ of analogue site data allows calculation of analogue benchmark values that can be adapted to provide evolving, dynamic benchmarks, for example, by using the most appropriate subset of current and historical data reflective of contemporary climatic conditions. Methods to calculate these benchmarks encompass averaging replicated site data across climatically similar years but may be extended to complex techniques such as ordination or hierarchical Bayesian modelling. Overall, methods to enhance mine rehabilitation success are urgently required, and to help achieve this goal, we provide a process framework for the use of analogue sites in mine rehabilitation.

Keywords: mine rehabilitation, native vegetation, reference ecosystem, ecological monitoring In the first year, all replicate sites for each target were analysed together, to create benchmark data for rehabilitation evaluation against completion criteria. This was done by direct comparison of analogue and rehabilitation mean values for each metric of interest. Bray-Curtis dissimilarity indices provided additional insights into community comparisons and these were tested using permutational multivariate analysis of variance (PERMANOVA) (Figure 4a). We also made heat-map tile plots of species occurrences and relative cover to provide site-level insights into species composition (Figure 4b). To provide a complementary perspective, benchmark values, derived from the results of the NSW government state vegetation survey (NSW DPE 2022c), were also provided for comparison against completion criteria targets. In the following years, we began the creation of annual dynamic benchmarking values—benchmark values created using both the current year and a subset of the earlier data that best matched the contemporary climatic conditions. For example, 2019 was a drought year, while the following years (2020–2022) were wetter than average. For the creation of the 2022 dynamic benchmark values, we excluded the drought year and used only 2020–2022 data. This approach differs from static benchmarking, where a single snapshot is made of the analogue target, and current and future rehabilitation is compared to this value. Similar processes have been used to create large-scale dynamic benchmarks for Australian vegetation communities (Yen et al. 2019). From each replicate site, we obtained mean values across the subset of sampling years, and from this, dynamic benchmark mean values were calculated (so that n = 5 for each target). We also performed more detailed statistical testing for some sites where this was a requirement of mine documentation, for example, using generalised linear models with a beta distribution for comparisons of species cover, avoiding the need to transform data for analysis (Douma & Weedon 2019). To visualise community similarity between analogue and rehabilitation sites, we used non-parametric multi-dimensional scaling (NMDS) of Bray-Curtis dissimilarity indices of species occurrence and abundance data (Bray & Curtis 1957; Oliver et al. 2022).An example output is shown in Figure 4a.

References: