Key Words

1 The Challenges

Decontamination of radionuclide-contaminated soils presents a particular set of challenges. These challenges, and the knowledge that radioecologists possess that might be helpful in surmounting them, are worth articulating not only because of their potential utility for cleaning radionuclide-contaminated soils but also because of the insights they provide to phytoextraction of inorganic contaminants in general.

Romney et al. (1) and Nishita et al. (2) reported, using radionuclides in the 1950s, some of the first calculations of the potential of plants to extract in-organic contaminants from soils. A single crop of Trifolium repens (clover) was able to remove 4.42% of added 90Sr, and nine crops over 520 d removed 24%. From: Methods in Biotechnology, vol. 23: Phytoremediation: Methods and Reviews Edited by: N. Willey ŗ Humana Press Inc., Totowa, NJ They suggested that this might make phytoextraction of 90Sr possible and contrasted this with the limited potential for 137Cs. Since then there have been a number of phytoextraction trials for radionuclides including those at Brookhaven National Laboratory, Upton, NY, United States (3-5), the Chernobyl exclusion zone (6,7), Argonne National Laboratory, Argonne, IL, United States (8), and Bradwell Nuclear Power Station, United Kingdom (9). Recent retrospectives (10-12) have concluded that phytoremediation for radionuclides could become useful in the near future. However, this is primarily because of the potential for rhizofiltration from effluent or phytostabilization of contaminated soils, rather than phytoextraction from soil. Reviews of potential rehabilitation schemes around Chernobyl have concluded that phytoextraction is not currently useful (www.strategy-ec.org,uk) (13). In fact, effluent filtration or soil stabilization is more efficient for radionuclides than is extraction from soil, with or without plants. The challenge is, therefore, to develop a viable method for extracting radionuclides from soil. Here, I outline a context and suggest research that might help increase the utility of phytoextraction for radionuclides, and use this to make comments on phytoextraction of inorganic contaminants in general.

2 Understanding the Context of Phytoextraction

2.1 The Problem of Radionuclide Contaminated Soil

There is a significant volume of soil contaminated with radionuclides world-wide. It has arisen primarily from nuclear weapons-related activities plus accidents at Chelyabinsk and Chernobyl (14). The concentration of radioactivity in contaminated soils varies widely, and is greatly skewed toward low activity concentrations, but is high enough in some locations to prevent agricultural production or be a potent hazard to human health. This potential to be a potent hazard to humans means that there is great pressure, and in many instances legilation actually in place, to regard soils with any activity concentrations detectably above background as contaminated and legally requiring decontamination. It seems unlikely that legislation on radioactively contaminated soils will become less stringent anywhere in the world in the near future, ensuring that there is very significant pressure to decontaminate large volumes of soil of radionuclides.

The total volume of soils contaminated with nonradioactive inorganics is much larger than that contaminated with radioactive but there is, at present, often much less pressure to decontaminate them. Thus, in addition to the very immediate necessity to deal with those radioactively contaminated soils that are hazardous, the pressure to decontaminate soils of radioactivity provides an interesting case study for those interested in phytoremediation. For example, it is probably sensible to imagine what the soil decontamination challenge mightdd be if public and legislative pressure on heavy-metal-contaminated soils became, as is at least possible, more like that on radionuclide-contaminated soils. Much phytoextraction research is prompted by the perceived problems arising from soil contamination. Such perception is a sociopolitical construction and is variable and changeable. I suggest that the perception of contaminated soils is likely, overall, to change toward the current perception of radioactively contaminated soils and that decontamination efforts for such soils are, therefore, particularly instructive for the phytoremediation community.

2.2 Methods of Soil Decontamination for Radionuclides

Despite there being very significant pressure to decontaminate soils of radioactivity there are relatively few instances of it having been carried out. Perhaps the most thorough decontamination operations for radionuclide-contaminated soils were those at Chelyabinsk and the Bikini Atoll (14,15). For terrestrial ecosystems they primarily involved physical removal of large volumes of contaminated soil plus some regolith reconstruction. Regolith conditions now only approximate, at best, what they were before contamination occurred (15). The method of removing contaminated soil, mostly to waste repositories, is sometimes referred to as an “established” decontamination option but it has a limited track record of successful implementation, primarily for small volumes of highly contaminated soil. It is very doubtful that it is a strategy that could be applied to anything like the majority of radioactively contaminated soils, not least because of restricted volumes in currently approved waste repositories (9), and certain that if it could be it would produce a large regolith reconstruction problem. Rehabilitation of soils contaminated with lower levels of radioactivity is currently achieved almost entirely through other means (16).

With the exception of small volumes of highly contaminated soil, it is at least debatable whether there is an established method that can deal with the volumes of radioactively contaminated soils that currently exist. When faced with the challenge of dealing with soil contaminated with inorganics, it is tempting to think that, if necessary, ultimately it can be capped or dug up and washed or just buried. Radioactively contaminated soils have been in a position of great pressure for clean up for years, perhaps revealing that the problem of contaminated soil is actually less tractable than is often believed. Radioactively contaminated soils may show that it might be sensible to consider that phytoextraction should be viewed within a context in which there are not really any “established” decontamination options for inorganics, rather than a context in which it has to be cheaper and more efficient than “established” decontamination options. If there are established, economically viable decontamination options it is difficult to account for the persistence of so much contaminated soil.

2.3 Radioecological Transfer Factors and Phytoextraction

The rate of phytoextraction of inorganic contaminants depends on the net soil-to-plant transfer rate. Radioecologists have long measured concentration ratios and transfer factors (TFs), and long argued about their potential utility. They have infrequently noted, however, how useful the concepts underpinning TFs are for phytoextraction. The discussions and experience of radioecologists in using TFs might be of quite wide utility in phytoextraction of inorganics.

TFs were developed primarily as part of “empirical” environmental models at the dawn of the nuclear age more than 50 yr ago. These models were among the very first for the behavior of contaminants in the environment and there is much experience of using them over long time periods in the radioecological community. TFs were primarily used to model transfers between ecosystem compartments. Some of these models helped to reveal the limitations of “empirical” environmental models, i.e., their undynamic nature and the difficulty of transferring them between environments. Although physicochemical models that are theoretically more dynamic and more widely applicable than “empirical” models have been constructed for many years for radionuclides, TFs still prove to be extremely useful, especially when urgent responses are necessary. Particularly in the early stages of developing phytoextraction systems, the lessons learned about TFs by radioecologists might be very useful for phytoextraction of inorganics.

Soil-to-plant TFs for radionuclides taken either for a single species on many soils, many species on a single soil, or many species on many soils, have been shown empirically by radioecologists, to be very variable (17), lognormally distributed (18), time dependent (19), and concentration dependent (20). In addition, soil availability can clearly be no more of a guarantee of significant soil-to-plant transfer than high plant uptake because some plants have extremely low uptake of radionuclides just as some soils have very low availability. As it is the net soil-to-plant transfer that determines the utility of phytoextraction, it is the overall behavior of this transfer that needs to be the ultimate aim of phytoextraction research. The TF concept and the experience of radioecologists in applying it is an essential reminder that overall soil-to-plant transfer can be complex on wide spatial and temporal scales and, as the ultimate empirical expression of relevant system behavior, needs analysis in its own right. I am not aware of any such analysis for phytoextraction of nonradioactive inorganic contaminants. The extensive experience of radioecologists in application of TFs might provide a foundation for such analysis.

2.4 The Economics and Environmental Impact of Phytoremediation

Phyoremediation is frequently touted as a cheap and environmentally benign decontamination method (21). This might turn out to be true but very few rigorous economic or environmental assessments of the technology have been reported. Because of the unique challenges of radionuclide-contaminated soils, radioecologists have highlighted some aspects of the technology that merit specific attention in economic and environmental assessments.

Plants used in phytoextraction trials for radionuclides become radioactive waste (22). Rigorous assessment of their potential disposal highlights problems seldom mentioned in discussions of phytoextraction of inorganic contaminants (23). For example, for most waste disposal streams, radioactive or not, fresh plant material, and especially that from fast-growing herbaceous plants, can be problematic and dried material is preferable. Drying of large amounts of plant material to requisite water content can be costly. It is tempting, therefore, to drip feed fresh phytoextraction waste into conventional waste streams. However, many operators are not only loathe to deal with contaminated waste but also unable to because of emission or leakage restrictions. Experience with radioactive waste is not directly transferable to other inorganic contaminants but it seems likely that disposing of large volumes of contaminated fresh plant material is likely to be more costly than is often envisaged. There are also some genuine difficulties in demonstrating how much has been gained environmentally if contamination is extracted from the soil in one place and buried in the ground or dispersed into the air or sea at another.

With phytoextraction of radioactively contaminated soil there are also significant impacts of operator protection (23). Although clearly a potentially low-maintenance option, phytoextraction, and particularly harvesting and waste disposal, does require some operator time. With radioactivity the potential for contamination can necessitate quite expensive protective measures. For some other inorganics, such as Cd, at high concentrations similar protective measures are already necessary in some countries and such legislation is tightening rapidly worldwide. Operating costs seem mostly to be ignored in assessments of the potential for phytoremediation but might be of some significance over the course of years.

3 Increasing Soil Availability of Radionuclides

For those involved in phytoextraction research, not only might the general context for radionuclides previously outlined be useful, but also advances in understanding radionuclide behavior in the soil-plant system. Specific aspects of radionuclide behavior that might be of interest to phytoextraction research are outlined next.

3.1 Radionuclides With High Soil-to-Plant Transfer

There are a number of reports of a single cropping almost completely removing 99Tc from contaminated soils (24). 99Tc is not radiologically significant in terrestrial ecosystems, although there is increasing interest as its disposal in permanent terrestrial waste repositories becomes more necessary (24), but it demonstrates phytoextraction’s potential for ions that are available in soil. 36Cl is another radionuclide that is highly available and targeted for terrestrial waste repositories. A number of assessments have noted the potential for 36Cl to be moved up into vegetation even from depths in the soil significantly below the rooting zone (25). There seems little doubt that increases of soil availability of 99Tc or 36Cl are not necessary and that contamination could be attacked using phytoextraction. Therefore, although they have not thus far been significant in terrestrial systems, 99Tc or 36Cl provide useful examples of the potential of phytoextraction.

Sr isotopes behave very similarly in the soil-plant system to the nutrient Ca, often having high soil-to-plant transfer (26). 90Sr is among the most radio-ecological significant isotopes and is a major contributor to doses at the most radioactively contaminated places on Earth, such as the environs of Chelyabinsk and Chernobyl (14). It has long been noted that 90Sr transfer from soil to plant is close to being high enough for significant phytoextraction to be a reality. 35S, a radioisotope of the plant nutrient S, can also have high soil-to-plant transfer. For 90Sr and 35S it seems very likely that the detailed understanding of Ca and S in the soil-plant system could enable the design of useful phytoextraction systems, probably without specific emphasis on increasing soil availability. It might, however, be salutary to analyze the case of 90Sr more closely. 90Sr is of great radiological significance, established methods involve great environmental disruption and there is potential for improving soil-to-plant transfer to levels that would be considered suitable for phytoextraction, yet there is little sign of phytoextraction being useful for 90Sr in the near future.

3.2 Radionuclides With Low Soil-to-Plant Transfer

There are radionuclides of great radioecological concern, e.g., 137Cs, 238U, and 239Pu, that are generally considered to have the lowest soil-to-plant transfer rates. They represent one of the greatest challenges for phytoextraction of inorganic contaminants. In general, 137Cs is so tightly bound by illitic clays, both on frayed-edge and interlayer sites, that it is very useful for tracing patterns of erosion in soils with such minerals (27). Soil-to-plant transfer from soils with just traces of illite can be very low (28).

However, even for 137Cs there are some angles of attack that give hope. Over the course of years 137Cs concentrations in waters show that 137Cs adsorption in illitic soils is not irreversible but that there is slow leakage (29). This is a long way from sufficient to give soil-to-plant transfer necessary for significant phytoextraction but shows that there are equilibria that might be manipulated. NH4+ has long been known to desorb 137Cs from clays (30), primarily nonillitic, and some authors have noted that the presence of NH4+ and/or nitrification inhibitors increases 137Cs uptake (31). NH4+ from illitic clay interlayers is accessed by plants in significant quantities (32) and although the Cs adsorption properties in clay interlayers lead to collapse, interlayers can be opened by molecules such as oxalate (33). Competitor-binding agents such as Na tetraethly-borate can draw K out of clay interlayers (34) and other binding agents such as Norbidine A (35) might also be able to do so. Recent research has also reported that changes in octahedral Fe in the rhizosphere can increase the availability of NH4+ from interlayer sites (36). Thus it seems possible that 137Cs might be removable from illitic clay interlayers, although this is some way from realization at present and achieving it in the field is quite another challenge. There has been much radioecological focus on illitc clay binding of 137Cs because many nuclear facilities are located in areas with such soils. This is, however, not true of all nuclear facilities and there are very significant areas of the planet in which 137Cs is quite available in the soil. For example, organic soils including histosols and spodosols, lateritic soils including oxisols, and allophanitic soils including andosols can all produce high soil-to-plant concentration factors (37-39). Thus, even for a recalcitrant contaminant such as 137Cs, phytoextraction should not perhaps be dismissed as a technology without potential. Radio-ecologists have a very detailed knowledge of its binding to soils and this is potentially very useful in manipulating its transfer from soil to plant.

U and Pu isotopes are generally very unavailable to plants (40). Further, many of the areas of most attention for U are mining spoils in which a significant proportion of the regolith is composed of U. In many instances there seems little hope of phytoextracting U or Pu in useful quantities. However, both U and Pu have complex soil chemistries and are very available to plants under certain circumstances (41). Organic acids have been demonstrated to make U highly available to plants (42). Similar effects might be achievable for Pu because of its solubility under certain conditions. Thus, phytoextraction is not without potential for both U and Pu. Some of the most formidable phytoextraction challenges, 137Cs, 238U, and 239Pu, are therefore not without hope. It is possible that the chinks of light might be the basis of the development of phytoextraction systems that might be useful in some instances at least. In the absence of other truly effective methods of extracting Cs, U, and Pu from soils this is significant.

4 Methods for Increasing Plant Uptake of Radionuclides

4.1 Ion Availability and Root Exudates

The physicochemical availability of an ion in the soil solution is frequently not what a plant experiences-many plants actively manage the availability of ions in the rhizosphere through symbioses or root exudates. There have been recent advances in engineering root exudates for the breakdown of organic contaminants but they also have great potential for manipulating the availability of inorganic ions.

Phosphate is poorly available in many soils and was probably a major limitation to the colonization of the land by plants. Mycorrhizal associations and proteoid roots have long been known to increase phosphate uptake by plants (43). Soils in which Fe2(PO4)3 predominates have extremely low concentrations of phosphate but plants such as Cajanus cajun (pigeon pea) have evolved exudates based on piscidic acid that can dissolve the highly insoluble Fe2(PO4)3 for uptake (44). Lack of soluble Fe limits plant growth on perhaps 35% of the world’s soils (45) but many plants exude a variety of phytosiderophores based on mugineic acid to mobilize it (45). Rice plants modified to exude phytosiderophores, which mobilize Fe from soils in which pedological processes render it unavailable, have been produced (46). Root exudates also play a key role in controlling As availability to plants in anaerobic soils. Thus, unavailable ions are mobilized by plants (47) and manipulating this process is widely considered to be part of the solution to nutrient limitations in agriculture (48). As yet, there has been very little consideration of this phenomenon for mobilizing radionuclides.

4.2 Manipulating Ion Uptake by Plants

The importance of the concentration of nutrient and toxic ions to agricultural production and to food and forage quality ensures that ion uptake by plants is a vibrant research topic. Plant breeders have succeeded in altering uptake of ions by crop plants and the genetic engineering of ion uptake is now a reality. Plant breeding and genetic engineering strategies to manipulate Cs uptake by plants are now within the realm of the possible. There has been less progress in research focused on manipulating plant uptake of other radionuclides but there is every reason to think that it might be possible.

The molecular biology of K uptake by plants is now advanced (49) and has had a great impact on our understanding of Cs uptake by plants. Although it has long been known that Cs and K probably have at least some common modes of entry into plants (50) it was not until relatively recently that electrochemical models were used to implicate voltage-independent cation channels in Cs transport (51). Recent research has focused on other types of channel (Corinna Hampton, personal communication), at least partly because knockout mutants have eliminated some types of K transporter as possible Cs transporters (52). It seems unlikely that there will be a single, or even a few, transporters that might be engineered to manipulate Cs uptake by plants but, as is already the case for other nutrients such as S (53), manipulating K uptake by plants is likely to be possible soon.

It now seems likely that crop breeding might play a significant role in manipulating ion uptake by plants. Quantitative trait loci, which identify areas of a genome associated with a phenotype, have been determined for Cs concentration in plants (54), the biodiversity available for breeding assessed (17), and field tests with crop lines carried out (M. Broadley personal communication). All these studies suggest that crops with significantly decreased or increased uptake of Cs might be possible, and suggest considerable potential for manipulating plant uptake of other radionuclides.

5 Conclusions

For radionuclide-contaminated soils, and perhaps for soils with other inorganic contaminants, a first phase of trials revealed phytoremediation systems to have limited utility, especially for phytoextraction. Certainly, phytoremediation is successfully being used in certain circumstances, but there seems little prospect of it being widely used to make significant inroads into the problem of contaminated soils in the near future. It is tempting, therefore, to be pessimistic about the prospects for phytoextraction in particular but this is precipitate. Experience gained from attempts at phytoextraction of radionuclide-contaminated soils provide insights that might help spur further phytoextraction research.

137Cs and 90Sr, neither of which are routinely phytoextracted from soils, might serve as examplars for the challenges facing the development of phytoextra ction research. There is very significant pressure to cleanse soils of them and much long-term knowledge of their behavior in the soil-plant system, in particular their TFs. In many soils, 137Cs is bound in collapsed illitic interlayer sites and is probably as difficult to phytoextract as any inorganic contaminant. In contrast, 90Sr is certainly available enough to plants, and there is enough knowledge of manipulating Ca transfer from soils to plants, for phytoextraction systems to have been perceived to be possible for nearly half a century (2). I think that identifying the barriers preventing the utilization of phytoextraction for these radionuclides might provide useful insights for phytoextraction of inorganic contaminants.

Phytoextraction of 137Cs and 90Sr is not prevented by the existence of other cheap, environmentally benign decontamination methods. Certainly, there are soils contaminated with these radionuclides that are dealt with but only with great environmental disruption and expense. The largest soil decontamination projects for radionuclides, perhaps for any inorganic contaminants, are for 137Cs at Bikini and 90Sr around Lake Karachay near Chelyabinsk. In both cases urgent clean up was necessary and soil removal the only option but neither case identified a good cheap, environmentally friendly method of soil decontamination. It seems likely that similar removal of large soil volumes for other inorganic contaminants will become less acceptable as the importance of sustainable environmentally benign decontamination increases. Thus, there is good news for phytoextraction-the competition is not as stiff as it is often made out to be and it will probably get even less stiff. Perhaps there are currently no methods for extracting inorganic contaminants from soils-just damaging ways to remove soil or to wash it. In this light, the challenge is to come up with any method for extracting contaminants from soils-it does not necessarily have to be cheap and it just has to make more environmental sense than soil washing. The management of waste might play a key role in determining whether or not these criteria can be met.

For 137Cs in particular, increases in soil-to-plant transfer are necessary and present a major challenge. I suggest that it might be possible to affect such increases in soil-to-plant transfer-certainly the soil decontamination challenge is big enough, and scientific knowledge great enough, to at last seriously attempt to effect it. At present there are probably enough possible avenues for research from both soil and plant science to make research worthwhile. However, it is also very relevant that ion transfer from soil to plant is a discipline that is undergoing a period of very rapid advancement and that many of these advances will be very useful for phytoextraction research. Clearly, sustainable human existence on Earth depends crucially on sustainable food production systems (55) but even before this is achieved there is a global micronutrient crisis in food to be solved (56). For these reasons, managing nutrient transfer from soils to plants underpins at least 2 of the top 10 challenges that environmentalists (57) now identify for the planet. Much research focus and investment can therefore be expected into management of the soil-plant system in the next 50 yr. Phytoextraction research needs to be sustained not least so that it can benefit from, and contribute to, global research into management of the soil-plant system.