As described in the present study and elsewhere 3, 8, there are many processes and interactions related to GCC with the potential to influence the fate and transport of chemicals in the environment at different spatial and temporal scales. The behavior of chemicals in the environment, however, is strongly influenced by their physical–chemical properties and, consequently, so is the sensitivity to various perturbations associated with GCC.
In an effort to quantify the physical–chemical properties most sensitive to changes due to GCC, we utilized a chemical space approach, whereby a large number of model simulations, based on a set of hypothetical persistent chemicals covering a wide range of possible combinations of partitioning behavior (see Supplemental Data), are performed. Model output is then summarized in the form of chemical space plots 75. Such an assessment provides information on the directionality and magnitude of the influence of these alterations, which can then serve as the background context against which other potential changes related to GCC can be compared. A key consideration of the model parameterization relates to the temperature dependence of degradation reactions in the environment as well as the physical–chemical properties of the chemicals, especially when comparing model output between different climates. The activation energies for degradation in different environmental media assumed for all hypothetical chemicals and the approach adopted for their partitioning properties are summarized in the Supplemental Data. Briefly, the internal energy of the phase change for octanol–water partitioning (dUOW) is defined as −20 kJ/mol, which is consistent with values reported for a number of known POPs 76. The enthalpies of phase change for octanol–air partitioning (dUOA) are described in Supplemental Data and that for air–water partitioning (dUAW) has been defined in a manner to ensure internally consistent property data (i.e., dUOW – dUOA 76).
Global-scale fate and transport simulations
The main purpose of the global-scale fate and transport simulations conducted here is to identify broad patterns in the response of different environmental compartments (e.g., air, water, soil) to changes in thermodynamic forcing (e.g., temperature dependence of environmental partitioning and degradation reactions) and precipitation scavenging (an important factor in establishing long-range transport potential for some compounds). Based on how the model has been parameterized, these simulations are most relevant to neutral organic compounds, but the approach described here could also be adopted to include ionogenic organic chemicals (e.g., 51–55).
A novel version of the multimedia environmental fate and transport BETR-World model 77 was used for these simulations (Supplemental Data). A fugacity-based fate and transport model, BETR-World represents the globe as a series of interconnected regions, each subdivided into bulk compartments representing the different environmental media present (e.g., atmosphere, freshwater, sediments, soil, vegetation, surface ocean water). The model has been revised to match the geographical units for which regional climate projections (temperature and precipitation change) were generated by the IPCC 22 and is therefore referred to here as BETR-IPCC. Additional details regarding model parameterization and application are provided in the Supplemental Data.
Briefly, projections for temperature and precipitation changes for the A1B scenario for the period 2080 to 2099 were summarized for 30 regions in the IPCC report 22 on a seasonal and annual basis and then compared to the baseline period, 1980–1999. The regional identifiers are presented in Supplemental Data and the projections for each region for the A1B scenario based on 21 models are presented in Supplemental Data (annual basis). Figure 2 summarizes the chemical space plots produced from the BETR-IPCC model, comparing output for three different geographical regions. Concentration ratios (GCC to baseline) are presented for the lower atmosphere and soil solids, as are freely dissolved concentrations for surface ocean water. Freely dissolved concentrations in the aqueous environments were selected because they are the most relevant for exposure in pelagic food webs. As a reference point, the majority of known POPs occupy the region defined approximately by log KAW = −1 to −5 and log KOA = 7 to 13.
Figure 2. Comparison of steady-state surface air (A), soil (S), freely dissolved freshwater, sediment pore-water, and freely dissolved surface ocean concentrations in the Arctic (ARC), northern Europe (NEU), and South Africa (SAF) model regions of BETR- Intergovernmental Panel on Climate Change under the global climate change (GCC) scenario compared to the baseline scenario (presented as the ratio of GCC model output to baseline model output). Partitioning properties for polychlorinated biphenyl (PCB)-153 and α- and β-hexachlorocyclohexanes (25°C) are also shown.
Download figure to PowerPoint
All model outputs under the GCC scenario are well within a factor of 2 of the output from the baseline scenario across the entire chemical space considered here (and typically ± 20%) (Fig. 2). One broad response across the chemical space is that concentrations are projected to increase in the surface air compartment and projected to be reduced in the bulk soil compartment. Concentrations of dissolved compounds are projected to exhibit both negative and positive responses to the GCC scenario, with more positive responses exhibited by the more hydrophobic chemicals (lower right side of panels shown in Fig. 2). The Arctic environment is projected to be somewhat more sensitive to changes introduced in the GCC scenario, as can be seen in the results for the lower atmosphere and concentrations dissolved in marine waters. Given that the model simulation assumes that emissions to this region are very low, responses to the GCC scenario are thus sensitive to alterations in long-range transport to the region in addition to changes in processing/redistribution within the region.
Interestingly, the area of the chemical space showing an elevated positive response to the GCC scenario for surface air concentrations (i.e., log KAW = −4 to −5, log KOA = 6 to 9, Arctic [ARC] region) is similar to the partitioning properties of the hexachlorocyclohexanes (HCHs), which were among the group of chemicals recently suggested to be exhibiting climate change–induced alterations in environmental fate in the Arctic (enhanced revolatilization from surface reservoirs, attenuation of decline in atmospheric levels over recent decades) 18. Other POPs suggested to be affected 18 (e.g., chlordane, DDT) have partitioning properties corresponding to elevated surface air concentrations in the Arctic as well. It is important to note, however, that the pattern of responses in Arctic surface air in this region of the chemical space is not identical to other model output, demonstrating that behavior in one environmental compartment may not be representative of changes elsewhere.
As an additional point of reference, we also generated output for a chemical with the partitioning properties shown for PCB-153 in Figure 2 and with the degradation half-lives and temperature dependencies specific to this chemical. In general, the results are consistent with the main conclusions reported by Lamon et al. 15, using a similar modeling approach, namely, that the concentrations of PCB-153 in the lower atmosphere are elevated globally under warmer conditions but that overall persistence in the environment (POV) is reduced (Table 2). Assuming no change to primary emission rates under the GCC scenario (e.g., due to temperature dependence of passive volatilization from stockpiles), the BETR-IPCC model projects that while concentrations in the atmosphere may be elevated (≈5–30%), the total global mass inventory is reduced (≈20–25%). In other words, as in Figure 2, changes in the lower atmosphere due to GCC do not reflect patterns and trends in other environmental media. This output also suggests that the magnitude of change in temperature (favoring the gas phase) is more influential than increased precipitation rates (favoring air-to-surface deposition) in determining the overall environmental fate of such compounds.
The key message of the output illustrated in Figure 2 implies that the long-term forcing on contaminant fate, transport, and bioavailability related to temperature and precipitation changes expected under GCC appears to be limited in magnitude and that this conclusion is valid across a wide range of physical–chemical properties. Other factors, such as changes in emission rate over time and proximity of sources to receptor sites of interest are likely to be far more influential on the evolution of contaminant burdens in the global context. Finally, surface air concentrations for neutral organics may not be a suitable proxy for characterizing the potential implications of GCC on ecological and human exposure.
Global climate change affects the bioaccumulation of contaminants through three primary mechanisms: environmental exposure (driven by concentrations encountered in surrounding media, Fig. 3), dietary exposure (driven by concentrations in food and the predator–prey connections that constitute the food web), and uptake and loss rates in organisms (driven by bioenergetic processes that control consumption, respiration, and elimination rates; Fig. 3 inset).
As discussed earlier, direct impacts of GCC may increase emissions of volatile chemicals and increase temperature-dependent degradation rates. However, indirect effects (such as regulation) may lead to decreases in primary emissions, and changing environmental conditions, such as to oxic/anoxic zones in aquatic systems, may lead to either net increases or decreases in degradation. Potential changes in food-web structure, which affect bioaccumulation via exposures encountered in the diet, can at times be difficult to predict, resulting in challenges with respect to integrating biological information across many organisms at different trophic levels 78. The within-organism impacts of GCC due to changes in bioenergetics rates have been previously explored by Ng and Gray 17 for several species and a single chemical. Their methods are expanded here (see Supplemental Data for details) to explore the potential impacts of GCC on the bioaccumulation in an organism of a wide range of chemicals.
The impact of GCC on bioaccumulation through bioenergetics depends largely on how annual temperatures relate to an organism's temperature preference and critical thermal limits. These can be broadly classified into three regions, illustrated in Figure 4A for the round goby (Apollonia melanostomus) 17: (1) the lower critical thermal limit (CTmin), (2) the optimal temperature(s) for growth (Topt), (3) and the upper critical thermal limit (CTmax). In terms of maximizing consumption and growth, the longer time spent near Topt, the better. Crossing either CTmin or CTmax, on the other hand, causes a steep decline in these rates. Here, differences in round goby growth are illustrated based on a typical Lake Erie annual temperature cycle compared to a 1, 2, or 3°C increase in the annual average temperature (Fig. 4A; see Supplemental Data for details). Different annual temperature profiles can result in substantial differences in growth over the lifetime of an organism (Fig. 4B), driven largely by the acceleration and deceleration of growth at winter lows and summer highs (Fig. 4B, year three growth comparison). Thus, increasing annual average temperatures only increases growth when temperatures remain below CTmax (compared to the Lake Erie baseline scenario, an increase of 1 to 2°C results in substantial increased growth); but above a certain threshold, the benefits decline (at a 3°C increase, growth drops closer to the baseline case). Of course, the shape of these curves will vary according to the thermal limits and sensitivities of an organism, and impacts from GCC will depend in large part on whether organisms are at the colder (below Topt) or warmer (close to CTmax) limits of their thermal range. Thus, the determining factor for the influence of GCC on an organism will not be the absolute change in temperature alone but rather how the change in temperature interacts with the species' thermal niche, that is, how close to its critical upper or lower thermal limits it is in the environment undergoing climate change.
Figure 4. Impacts of global climate change on bioaccumulation within an organism. Species thermal range interacts with temperature scenarios (A) and with bioenergetics to affect growth (B), leading to highly seasonal bioaccumulation patterns that can be above or below baseline values (C, D).
Download figure to PowerPoint
We investigated the impacts of these temperature changes on bioaccumulation potential by considering the ratio of concentrations in the baseline and warmer scenarios, predicted in a round goby over the course of one year (year three in a 5-year simulation). We focus on the 2 and 3°C increases, which represent the fastest growth and highest temperature scenarios, respectively (Fig. 4C and D). Concentration ratios >1 indicate more bioaccumulation relative to the baseline case, whereas ratios <1 indicates less bioaccumulation. We do this for a set of hypothetical chemicals with biotransformation half-lives ranging from 0.1 to 1,000 d and log KOW values ranging from 0 to 8.
We observe that seasonal effects are pronounced for both scenarios, with concentration ratios both above and below 1 in a single year. Increases in bioaccumulation relative to the baseline case occur during the spring and autumn, while decreases occur during the summer. This is an intuitive observation: when temperatures approach Topt, consumption rates increase. This causes an increase in uptake of the chemical with food. On the other hand, when summer temperatures exceed CTmax, consumption rates, and thus uptake, fall.
The magnitude of the deviation from the baseline case depends most strongly on the biotransformation half-life and then on KOW. The strongest impact on bioaccumulation potential is on chemicals with biotransformation half-lives between 0.1 and 1 d (Fig. 4C and D) and log KOW >6. Spring and autumn increases in bioaccumulation of the T3 scenario can be as high as 50% in concentrations relative to the baseline for these chemicals. More substantially, the summer decreases for this scenario can be dramatic, as large as a factor of 10. Reaching or exceeding CTmax can thus strongly influence chemical uptake and loss but only for those chemicals that undergo rapid biotransformation and thus are not bioaccumulative.
Interestingly, for what we might consider the classical POP-like chemicals (with log KOW between 6 and 8 and half-lives >100 d), changes are much less pronounced seasonally, with maximum increases of only 10% and decreases of 20 to 40% (Fig. 4C and D). Thus, for chemicals that are considered to be problematic from a bioaccumulation perspective (i.e., persistent chemicals with low biotransformation rates) the direct influence of GCC through temperature-dependent uptake and loss rates is likely to be minimal. It would be of interest, however, to expand this assessment to chemicals of emerging concern.
It should be noted that projected changes in these simulations consider constant concentrations in the environment. We must also consider the overall landscape of uncertainty concerning chemical fate under GCC. First, there is the magnitude of uncertainty in projected temperature increases (Table 1) relative to the sensitivity of bioaccumulation to temperature changes (e.g., Fig. 4C vs D). Second, other GCC impacts will influence contaminant transport and thus the levels encountered by organisms in the environment, including changes in primary emissions, precipitation, and runoff. Resulting increases or decreases in environmental concentrations could be confounded by the seasonal variability in uptake and loss rates illustrated here (Figs. 4C and 5D), particularly for less persistent chemicals. Third, it is likely that species coexisting in a food web will have different thermal ranges and sensitivities, particularly as they adapt to climate warming. Thus, modeling the changes in bioaccumulation due to GCC in a way that can integrate physicochemical, ecological, and physiological impacts will be particularly challenging.