Developing completion criteria for rehabilitation areas on arid and semi-arid mine sites in Western Australia

Abstract

Continued expansion of the gold and nickel mining industry in Western Australia during recent years has led to disturbance of larger areas and the generation of increasing volumes of waste rock. Mine operators are obligated to rehabilitate all disturbed surfaces and reconstructed landforms, and considerable effort and expense is now applied to the achievement of this objective. Associated with increasing rehabilitation effort is the requirement to accurately judge rehabilitation success through the development of completion criteria. Completion criteria are rehabilitation performance objectives set as conditions of approval for each stage of rehabilitation and for the project as a whole. They provide standards against which the success of rehabilitation can be measured, or more broadly the point at which responsibility for rehabilitation is complete. The current research project tackles the development of completion criteria by investigating ecosystem function within a variety of rehabilitation trials at four mine sites located in arid and semi-arid Western Australia, and also within surrounding 'natural' vegetation complexes undisturbed by mining, termed analogue sites. Six specific objectives were identified as part of the study: 1. To establish an appropriate end point land use for each mine site where field trials were established; 2. To examine long-term ecosystem development through the assessment of revegetation at a variety of rehabilitation sites; 3. To examine functional components within analogue communities and make appropriate comparisons with rehabilitation trials; 4. To record the potential reproductive capacity of revegetation progeny, and determine how this relates to ecosystem function;5. To provide a better understanding of ecosystem function by investigating the relationship between state factors, interactive controls, and ecosystem processes at rehabilitation and analogue sites; and 6. To develop a methodology for establishing realistic environmental completion criteria at mine sites situated in arid and semi-arid Western Australia. Field trials were established at four mine sites located within three subtly different bioclimatic zones that extend through the arid / semi-arid shrubland belt of Western Australia; Northeastern Goldfields (Granny Smith Gold Mine, Sunrise Dam Gold Mine), Eastern Goldfields (Black Swan Nickel Mine), and Northeastern Wheatbelt (Westonia Gold Mine). 1 The re-establishment of a self-sustaining vegetation cover integrated with the surrounding ecosystem, was the common end land use objective at the four mine sites selected for this study. For three sites located in the Northeastern Goldfields and Eastern Goldfields of Western Australia, sheep were grazed on surrounding rangeland; the fourth site located in the Northeastern Wheatbelt of Western Australia, and was surrounded by Crown Land. 2 To better understand ecosystem function, the dynamic behaviour and interaction of plant biodiversity parameters was monitored regularly at 19 post-mining rehabilitation sites up to 11 years after direct seeding. For functional ecosystems, plant biodiversity parameters changed rapidly during the initial five years after seeding following predictable trends, after which time they remained within a relatively stable range.The stabilising of parameters over time was identified as a key indicator of rehabilitation success, however the point at which the parameters stabilised was influenced by numerous variables and was difficult to accurately predict. Prolific seed germination resulted in high seedling density during the initial growing season. Plant density then progressively decreased in response to competition, before stabilising within a range approximately five years after seeding. Revegetation cover was typically low during the first growing season, increasing rapidly there after before also stabilising in line with plant density. Maximum species richness was generally achieved during the first and second year when annual Atriplex species were prominent. Perennial Atriplex species established more slowly during the early stages of revegetation development, but eventually replaced the annual component as the dominant taxa. Perennial Maireana species required up to three years before germinating in the field and establishing themselves in the revegetation; in many cases they replaced perennial Atriplex as the prominent taxa. The presence or absence of cyclonic rainfall during the first growing season was a major determinant of the ecosystem trajectory, controlling revegetation structure and composition. The germination and successful establishment of hard seeded species, including Acacia and Senna, was reliant on heavy summer rainfall during the early stages of ecosystem development to break seed dormancy and extend the length of the first growing season. This provided an important competitive advantage against faster growing Atriplex species, which possessed greater drought tolerance.The intensity of summer rainfall was also beneficial in leaching surface salts from the upper profile and hence, reducing salinity within the rooting zone. In the absence of heavy summer rainfall during the first growing season, the establishment of a low chenopod dominated vegetation cover was favoured, total species richness for the rehabilitation tended to be lower, and the variety of plant life forms was restricted to low and mid stratum shrubs. Increasing water stress resulted in progressively higher rates of local species extinction, with fewer taxa possessing the drought tolerance adaptations required to survive. For established revegetation, cyclonic rainfall increased productivity (as measured by % foliage ground cover) and stimulated the establishment of new taxa, which in many cases were brought in from adjacent unmined vegetation complexes (analogue sites). While the benefits of summer cyclonic rainfall were undoubtedly important to ari and semi-arid ecosystems, the occurrence of drought was also important in buffering the ecosystem against large-scale change by acting as a negative feedback to constrain cumulative productivity. Parent waste rock material varied considerably between rehabilitation sites, affecting the soil resource supply and associated functional components. Extreme salinity was a typical limitation of the rehabilitation medium, reducing the variety of salt tolerant species and favouring annual Atriplex during the early stages of ecosystem development. The cover of annual species present during early stages of ecosystem development contributed to decreasing salinity in the plant rooting zone, by reducing surface temperature and hence capillary rise of salts during summer months.Annual Atriplex species were replaced by perennial Atriplex in line with decreasing surface salinity. Fundamental to successful revegetation of the post-mining rehabilitation site was the requirement that reconstruction and contouring focus on maximising water retention and reducing salinity within the upper soil profile. Once the initial vegetation community established and plant parameters became relatively stable, change continued to occur, albeit slowly. One factor contributing to this change was the immigration rate of biota from adjacent revegetation or more commonly from surrounding analogue complexes. Linking rehabilitation areas to surrounding functional ecosystems ensured the movement of plants and animals, and ultimately increased the rate of recovery. The sustainability of post-mining rehabilitation was enhanced where these links were established early, allowing for the provision of additional seed and the migration of displaced species. The life cycle pattern of keystone species in the revegetation was found to be an important determinant in long-term sustainability of the plant cover, particularly for chenopod shrublands where one species was typically dominant. The senescence and death of large numbers of a dominant revegetation species together, had the ability to significantly alter the revegetation structure and composition. The impact for rehabilitation where a number of dominant taxa co-exist was less pronounced. Thus it follows that a minimum level of species richness was important to long-term rehabilitation sustainability, as was the development of an age-class structure in the rehabilitation.The most common disturbances encountered at the rehabilitation trial sites were drought, overgrazing and weed infestation. All three disturbances decreased the plant biodiversity parameters measured. Ecosystem recovery following disturbance was dependent on effective rainfall, but occurred rapidly with plant parameters returning to pre-disturbance levels within one to two growing seasons. The recovery of plant biodiversity parameters followed the same trends identified at functional rehabilitation sites during the initial five years following direct seeding. 3 Assessment of plant biodiversity parameters occurred at 15 analogue sites supporting native vegetation undisturbed by mining. It was anticipated that data from analogue sites could be used as a 'reference' against which to compare developing rehabilitation. However, analogue vegetation complexes were less dynamic in comparison to rehabilitation sites. Minor seasonal changes were recorded for plant biodiversity parameters, but overall annual change was minimal. Significant and sudden changes within analogue communities only occurred following disturbance, such as severe overgrazing, and recovery to pre-disturbance levels was rapid following the removal of the disturbance and return of effective rainfall. A major difference between rehabilitation and analogue sites related to their age. Rehabilitation sites were 'juvenile systems' assessed against a time frame much shorter than had been required for natural processes to achieve the developmental state represented at analogue sites.Hence, it was important not to model one specific analogue site too closely, but instead model the desired revegetation structure and species composition on a variety of local analogue complexes occurring in parent materials 'matched' closely to those of the rehabilitation site. Data from analogue sites should be utilised extensively during rehabilitation planning, but cautiously when interpreting the rehabilitation outcome. For mine sites in arid and semi-arid Western Australia, the application of specific numeric targets for plant biodiversity parameters as a measure of rehabilitation success was not recommended. A number of factors and controls in the developing ecosystem together determined the rehabilitation outcome. These factors were site and time specific; minor changes in any number of variables led to significantly different rehabilitation outcomes, making them difficult to accurately predict. 4 Quality and germination testing confirmed progeny seed from a number of rehabilitation trials was of similar or higher viability than the maternal seed originally sown. This was further confirmed by field responses at trials in the Northeastern Goldfields one year after the 1994 drought, when elevated plant density was recorded following the return of above average rainfall. The ability of rehabilitation to show an immediate response to rainfall following a seven-month drought, and for vegetation parameters to subsequently recover to pre-disturbance levels within one to two years, provided an indication that the revegetation cover was resilient. The relationship between plant production and rainfall was dependent on a 'carryover' effect between seasons or following drought years, and 'pulses' mediated, for instance, by the amount of seed in the soil store.The 'reserve' component in and ecosystems was responsible for both the memory of the system between pulses and for its long-term resilience. 6 The analysis of time series data collected from 19 rehabilitation trials emphasized the importance of planning and implementation of best practice techniques to subsequent rehabilitation success, and reinforced the difficulty associated with accurately predicting the final rehabilitation outcome. The large spatial heterogeneity of undisturbed vegetation complexes across the landscape of arid and semi-arid Western Australia, provided the foundation on which site-specific rehabilitation scenarios could be modelled, albeit with caution. The translation of data into useful completion criteria was dependent on the realisation that successful rehabilitation requires the implementation of best practice rehabilitation techniques, as determined by technically prescriptive (design) based standards, as much as the identification of a successful rehabilitation outcome, as determined by performance (outcome) based standards. With this in mind, completion criteria were developed as part of a robust theoretical framework incorporating the larger mine plan, and were not simply based on numbers generated as stand-alone performance standards. The broad methodology generated could be adopted by any mine site across the mining industry, however the criteria and, more specifically, the standards for each criterion should always remain site specific.The methodology designed for developing completion criteria has been addressed in three stages: 1. Planning, 2. Operational and Monitoring, and 3. Post-Mining Hand-Over. Within each stage three parameters are addressed: 1. Criteria, 2. Process, and 3. Standard. 'Planning' is the most important stage in the development of completion criteria. It is the stage when an appropriate end land use is determined, analogue sites are assessed, a rehabilitation plan developed along with specified design standards ensuring implementation of best practice techniques, and a process of risk assessment implemented. The 'Operational Monitoring Stage' focuses on rehabilitation success during the period of ecosystem development. This stage is concerned largely with rehabilitation monitoring, from which performance standards can be developed to gauge rehabilitation success for specific periods during revegetation development. The initial task in Stage 2 is to ensure all aspects of the rehabilitation plan have been implemented as specified in Stage 1, and meet agreed design standards. The final stage of the completion criteria process, 'Post Mining Hand Over', is to ensure the rehabilitated site is safe, and able to successfully revert to the end land use.While plant biodiversity parameters formed the focus of the current study, a variety of other functional ecosystem components may also make sound assessment criteria for determining rehabilitation success. Increasing the knowledge base for other functional components in arid and semi-arid ecosystems would further increase the ability to accurately determine rehabilitation success.