How clean is clean enough? Current knowledge from clean-up to restoration in Antarctic soils

Fuel-affected soils around operational areas remain a practical management issue in Antarctica. Campaign windows are short, on-site resources are limited, and existing Antarctic guidance recognises the need to identify and repair long-standing contaminated sites [1]. There are roughly eighty research stations in Antarctica, most of them on ice-free coastal ground [2]. This operational footprint requires large volumes of fuel to be moved, stored, and used each year, so national programmes must manage both legacy contamination and occasional new incidents. Using the Council of Managers of National Antarctic Programs (COMNAP) facilities list [3] as a baseline, and publicly reported operational figures as reference points, we estimate that station operations—excluding ships and aircraft—consume on the order of tens of millions of litres of diesel per year (approximately ~30 million litres, depending on the balance of year-round and seasonal stations). A minority of stations report some renewable generation alongside fossil fuels, while many list fossil fuel only; in practice, diesel remains the primary energy source or the guaranteed backup across most of the network, with renewables typically supplementing rather than replacing diesel supply [3]. In this context, managing diesel-affected soils becomes a routine element of station operations rather than an exceptional response. 

The Protocol on Environmental Protection to the Antarctic Treaty, commonly known as the Madrid Protocol, and its Annex III on Waste Disposal and Waste Management require Parties to identify and clean up past and present waste disposal sites and abandoned work sites, and to remove or remediate wastes to minimise adverse environmental impacts, except where such removal would cause greater harm [1]. To support implementation, the Committee for Environmental Protection (CEP) of the Antarctic Treaty developed a Clean-Up Manual that frames the overarching objective as minimising environmental risks and interference with the natural values of Antarctica by repairing legacy sites and locations contaminated by fuel or other hazardous substances [4]. Across programmes, this has been accompanied by increasing use of low-impact remediation options for fuel-affected soils, including: biopiles (engineered ex situ soil piles managed to enhance biodegradation), ecopiles (where pile-based treatment is combined with vegetation or ecological recovery goals) and in situ biostimulation, in which nutrients, moisture, or aeration are adjusted directly in the contaminated soil to stimulate indigenous hydrocarbon-degrading microorganisms. As these approaches become more common, closure decisions—the point when the remediation or clean-up is completed—repeatedly return to the same practical question: how clean is clean enough to meet the Protocol’s obligations and the CEP’s clean-up objective? 

A strict restoration approach would have a definitive answer to this question: “clean enough” would mean that the affected soil has returned to the condition it had before the spill. However, while ideal and the best possible outcome, in many circumstances it is not possible to achieve this high standard. 

In practice, programmes often combine numerical endpoints (commonly bulk total petroleum hydrocarbons, and/or hydrocarbon fraction ranges, and/or Benzene/Toluene/Ethylbenzene/Xylene (BTEX) and Polycyclic Aromatic Hydrocarbons (PAH) [5]) with qualitative assessments of risk, site stability, and anticipated future use in reaching their closure decision. However, harmonised closure criteria are not applied consistently across Parties. Chemical thresholds are frequently borrowed or adapted from domestic guidelines that were developed for temperate systems and different land-use assumptions, despite the distinctive vulnerability and slow recovery of Antarctic terrestrial ecosystems. Furthermore, functional attributes of soil—such as nutrient cycling, decomposition, or basic water regulation—are measured less often and seldom incorporated formally into closure decisions in Antarctica. One consequence is that treated soils may remain in biopiles or stockpiles for extended periods because managers lack a clear, auditable basis for declaring them ready for reuse or reintegration into station landscapes. This uncertainty ties up space and infrastructure and contributes to an accumulating inventory of semi-permanent, “partially remediated” soils whose environmental performance is not consistently documented. 

This information summary provides a review of how closure decisions for diesel-affected Antarctic soils are currently supported, based on existing policy guidance and the published literature. Across Antarctic programmes, evidence used for closure is commonly assembled at station scale around three recurring elements (Figure 1): chemical endpoints, exposure and stability, and indicators of functional recovery, reflecting the types of information most often needed at the station scale. 

First, chemical endpoints for total petroleum hydrocarbons (TPH) and relevant fractions must be interpreted explicitly for Antarctic soils and climate. Temperate-country guideline values are often used as reference points, but their underlying assumptions differ from Antarctic settings (e.g., productivity, exposure scenarios, land use, recovery rates). Recent work on site-specific guideline derivation (e.g., [6]) illustrates an approach that is closer in spirit to what Antarctic programmes may require: thresholds grounded in local hydrology, contaminant behaviour, and ecological context, rather than imported wholesale from vastly different regions across the globe. Within this perspective, measuring bulk TPH remains a useful screening metric, provided that it is paired with information on local hydrology, mobility, plausible exposure routes and toxicity to local species [7].   

 Second, pathway-and-stability lines of evidence address whether remaining hydrocarbons are creating active exposure pathways to water, biota, or people, and whether treated soils are physically stable under local wind, melt, and freeze–thaw conditions. In contaminated-land practice, this type of pathway assessment is typically structured using a conceptual site model (CSM) that links sources, fate and transport processes, and receptors to support risk and closure decisions [ISO 21365:2019; [8]. In practice in Antarctica, this is commonly documented through a combination of field observations and targeted checks focused on: the absence of free product or visible sheen; the absence of discharge to surface waters or meltwater channels under representative wetting events; and the robustness of pads, covers, or landforms under expected seasonal dynamics. Where infrastructure and human receptors are present, additional receptor-specific exposure assessment and engineering controls may be required; for example, vapour intrusion into buildings can be evaluated using contaminant-transport modelling and vapour barrier performance [9].   

Third, a functional-recovery line of evidence is useful for “restoration” outcomes, even when full return to pre-spill conditions is unrealistic, especially in coastal Antarctica where contamination from the soil into the ocean is possible. A practical starting point is a small set of soil-function indicators referenced against nearby pristine soils, focused on processes largely mediated by microbial communities. Three functional themes recur across the Antarctic soil literature and provide a practical starting point for functional recovery benchmarks: (i) nitrogen transformation processes (e.g., nitrification) [10, 11]; (ii) soil stabilisation and basic water regulation, often linked to the development and recovery of phototrophic biological soil crusts (surface soil communities dominated by photosynthetic microorganisms) [12]; and (iii) decomposition and mineralisation of organic matter and associated recycling of carbon (C), nitrogen (N), and phosphorus (P) [13]. While recovery of soil function is the central focus of this component, basic community metrics (e.g., richness and diversity indices) can provide helpful context by indicating whether treated soils remain dominated by a few opportunistic taxa, or whether they are moving towards the more even assemblages typical of nearby undisturbed soils; persistent petroleum pollution can also drive marked shifts in microbial community responses [14]. These metrics are best interpreted as supporting diagnostics rather than as stand-alone closure threshold. 

Taken together, these three recurring evidence streams can be viewed through a weight-of-evidence lens, consistent with triad-style site assessment frameworks (ISO 19204:2017, [15]), which aim to make explicit what is being measured, where evidence converges or diverges, and which uncertainties remain relevant for decision-making. It also highlights a practical research gap: published treated-versus-pristine comparative datasets remain limited and methodologically heterogeneous, constraining realistic benchmark ranges across Antarctic regions. 

From a regulatory perspective, closure decisions sit at the interface between the Antarctic Treaty System obligations and domestic contaminated-land frameworks. The Madrid Protocol and CEP guidance set common objectives but leave numerical criteria largely to national interpretation. As a result, programmes may apply domestic thresholds and methods that differ in risk basis, land-use assumptions, and levels of ecological protection [6]. The variety of benchmarks and approaches explains why closure has often relied on a patchwork of numbers and judgement in Antarctica, and why it remains difficult to demonstrate that comparable levels of protection are being applied across programmes. It also underscores the value of wider uptake of low-impact remediation practice—including bioremediation where appropriate—supported by more consistent documentation of closure evidence. 

A final consideration is that remediation itself can leave a measurable footprint. Nutrient addition -biostimulation- can alter soil conductivity and select for taxa specialised in nutrient cycling [10]; bioaugmentation—the addition of known degrading microorganisms—can increase biomass and shift community structure; and phytoremediation -the use of plants to enhance microbial biodegradation- or revegetation approaches may favour atypical plant cover relative to surrounding ground [16]. Recognising and documenting this footprint—alongside hydrocarbon attenuation and pathway closure—helps frame closure as a balance between reducing contaminant-driven impacts and avoiding persistent treatment-driven departures from nearby reference soils. 

This section focuses on challenges in applying station-scale closure decision criteria in Antarctica (Figure 1), rather than bioremediation efficacy. The recurring issue is that Antarctic variability and logistical constraints limit how chemical, pathway/stability, and functional indicators are measured, compared, and interpreted across sites. 

A first cluster of challenges relates to pronounced cross-site and seasonal variability. Air temperature, wind exposure, snow accumulation, permafrost depth, and the timing and intensity of melt differ sharply between coastal peninsular stations, continental nunataks, and sub-Antarctic islands. Even within a single station footprint, microtopography and snowdrift patterns can create steep moisture and thermal gradients over short distances [17]. This variability affects contaminant mobility and exposure conditions, and it can shift both measured hydrocarbon concentrations and functional indicators, complicating cross-site comparisons and uncertainty bounds. 

A second challenge is water limitation during key parts of the season. Melt inputs can be brief, episodic, and difficult to capture and redistribute consistently at the station scale. This constrains the ability to test pathway control under representative wetting events (e.g., leaching) and to obtain comparable functional measurements. Under very dry conditions, contaminant attenuation and functional recovery may be driven as much by rare high-melt years as by year-to-year management decisions, which stretches the timescales over which closure can be evaluated. 

A third limitation is the winter data gap. Most monitoring of contaminant decline and soil function occurs in summer, even though soils experience long periods outside the main field campaign window. Therefore, relying on data from a single summer season—or a handful of short seasonal snapshots—can misrepresent both hydrocarbon exposure conditions and longer-term trajectories of change. 

Even where measurements exist, comparability of evidence remains a major obstacle. The practical ability to apply standard analytical methods consistently across programmes varies (e.g., limited on-site analytical capacity, off-continent turnaround times, and variable quality assurance/quality control documentation), affecting detection limits, fraction reporting, and data resolution. Nutrient regimes and biopile designs are also adapted to local soils and climates, meaning that “successful treatment” can reflect both site properties and optimisation history [18, 19]. Sampling designs also vary widely, and hotspot-scale heterogeneity means this can affect estimated concentrations, uncertainty, and the probability of detecting residual contamination. In that regard, within the current Antarctic governance framework, the CEP Clean-Up Manual [4] provides the most practical non-prescriptive mechanism to encourage more consistent minimum expectations for sampling design description and reporting of core metadata, improving comparability while respecting programme-specific logistics. 

Finally, emerging contaminants and mixtures introduce a forward-looking complication [19]. Diesel-affected soils may co-occur with microplastics from infrastructure and clothing, additives in fuels and lubricants, firefighting foams, and other organic or inorganic co-contaminants. These mixtures can alter sorption, transport, and toxicity in ways not captured by bulk TPH, and their ecological effects in cold, low-productivity soils remain poorly constrained [20]. 

Together, these constraints explain why closure remains inconsistently defined across Antarctic stations and why cross-site synthesis is difficult. They also point to practical priorities—better seasonal coverage, a small set of more comparable core indicators, and improved characterisation of contaminant mixtures—while recognising that benchmarks must accommodate Antarctic diversity rather than rely on a single uniform endpoint.