**1. Introduction**

Metals are persistent, bioaccumulate and are well-known for their toxicity. The presence of elevated concentrations of metals in soil due to natural or anthropogenic activity represents a potential risk to human and ecological health; consequently, the assessment of soil metal contamination risks is of great interest to governments and regulators [1]. The measurement and understanding of the bioavailability of elevated metals in soil are an important part of any risk assessment. The bioavailability of metals in soil is a complex dynamic process and believed to be driven by speciation, sorption and biological processes in soil, but still little is known about their interplay in soil systems [2]. Previous studies have shown the importance of soil metal bioavailability in determining metal toxicity; however, there is still no general agreement on how best to measure metal bioavailability in soils [2–6]. Over the years, researchers have attempted to measure bioavailable metals in soil using many techniques such as ion exchange extraction, chelate extractions, rhizosphere extractions, isotopic solution extractions and the diffusive gradients in thin films technique (DGT). Effective elemental concentrations available for uptake from the solution-phase and solid-phase [7] can be simulated using the 2D DIFS model used in DGT. The DGT technique has also been used to estimate metal release kinetics from the solid to solution phase in soils [8,9]. The response time of soil to the resupply of metals in to the solution phase represents the response to metal depletion, and is directly related to the rate constant of the metal resupply process from the solution phase to the DGT film and therefore to plant roots [7]. To predict the concentration of metals in wheat plants as a result of the uptake of metals from uncontaminated soil, mathematical models including multiple regression analysis have also been validated and applied [10,11]. Generally, these techniques may not account for metals that have been accumulated over time in soil i.e., the aging effect of metals in soil. Thus, there is still a need to find and validate techniques for measuring bioavailable metals in aged contaminated soil.

Metal aging in soil is the process by which the bioavailability, mobility and/or the exchangeability of metals declines over time. The aging process is also referred to as natural attenuation or the fixation effect of metals in soil [12]. Aging of metals in soil is a long-term process and an important aspect affecting metal bioavailability [13,14]. However, our current understanding of the kinetics of metal aging in soil is limited. Experiments have confirmed that metal bioavailability in soil decreases as a result of the aging process [15–17]. Temperature, pH, moisture content and climate affect the rate of metal aging in soil, with pH considered as the most important parameter [18]. However, the majority of experiments on metal aging have been conducted using soils that have been spiked with metals in a laboratory [19], and thus have only assessed short-term aging effects.

The aim of this study was to systematically investigate metal bioavailability in roadside soil at varying road ages and metal concentrations. Wheat plants were used as a bioassay to accumulate bioavailable metals. The ability of DGT, soil water extraction, CaCl2 extraction, total metal concentrations in soil and optimised linear models to measure metal bioavailability, together with the release response time of metals and its impact on metal bioavailability in field-aged soils were studied. We used roadside soil to investigate soil properties and metal aging effects on bioavailability for a wide range of metals. Cadmium, Cu, Ni, Mn and Zn have previously been found to be elevated in roadside soils from Melbourne, Australia [20]. More information on the vehicular sources of Cd, Cu, Mn, Ni and Zn in roadside soils can be found in De Silva et al. [21]. Soil samples were collected from sites representing three different road ages, i.e., new, medium and old-aged roads.
