No Significant Contribution of Arbuscular Mycorrhizal Fungi to Transfer of Radiocesium from Soil to Plants

Abstract

The diffuse pollution by fission and activation products following nuclear accidents and weapons testing is of major public concern. Among the nuclides that pose a serious risk if they enter the human food chain are the cesium isotopes ¹³⁷Cs and ¹³⁴Cs (with half-lives of 30 and 2 years, respectively). The biogeochemical cycling of these isotopes in forest ecosystems is strongly affected by their preferential absorption in a range of ectomycorrhiza-forming basidiomycetes. An even more widely distributed group of symbiotic fungi are the arbuscular mycorrhizal fungi, which colonize most herbaceous plants, including many agricultural crops. These fungi are known to be more efficient than ectomycorrhizas in transporting mineral elements from soil to plants. Their role in the biogeochemical cycling of Cs is poorly known, in spite of the consequences that fungal Cs transport may have for transfer of Cs into the human food chain. This report presents the first data on transport of Cs by these fungi by use of radiotracers and compartmented growth systems where uptake by roots and mycorrhizal hyphae is distinguished. Independent experiments in three laboratories that used different combinations of fungi and host plants all demonstrated that these fungi do not contribute significantly to plant uptake of Cs. The implications of these findings for the bioavailability of radiocesium in different terrestrial ecosystems are discussed.

A principal function of mycorrhizal fungi is transporting mineral elements from soil to plants, and in this process, they act as highly efficient scavengers for several mineral elements (33). Quantitatively, mycorrhizal transport is most important for plant nutrients like NH₄ and PO₄, but both essential and nonessential trace elements may take the same route (23). It is well known that many ectomycorrhizal fungi accumulate radiocesium, and during the years following the Chernobyl nuclear accident in April 1986, fruiting bodies of many ectomycorrhizal fungi were found to contain high levels of radiocesium (1, 10, 20). In certain soils, radiocesium held in belowground fungal biomass was found to account for a large part of the soil's radiocesium content (9, 26, 27). This phenomenon is, however, limited to forest ecosystems, where ectomycorrhizas occur in high densities. Agricultural land, ecosystems dominated by herbaceous plants, and tropical and subtropical plant ecosystems are characterized by an extensive occurrence of arbuscular mycorrhizas (AM), formed by a group of fungi for which the capacity of Cs transport is largely unknown.

Some conflicting results on the role of AM in plant uptake of Cs obtained in pot experiments have been reported. For example, Rogers and Williams (28) found that AM increased the ¹³⁷Cs content of clover twofold, while in Sudan grass, the content of ¹³⁷Cs was similar to that of nonmycorrhizal controls. In this experiment, AM inoculation also resulted in a twofold increase in growth of clover, which may have affected the uptake of Cs. Similar results were obtained by Entry et al. (7) with three grass species. Here, a 20 to 40% increase in Cs concentration was observed upon inoculation with either of two fungi, even though controls were highly mycorrhizal (55% of control roots were colonized versus 70 to 87% in inoculated plants), and inoculation led to increased growth in some cases. Finally, Berreck and Haselwandter (2) found that mycorrhizal colonization caused up to a 30% reduction in Cs uptake by the grass Agrostis tenuis at three growth stages, but in this experiment, inoculation led to reduced plant growth. The experimental setup used in these studies does not assess fungal transport but rather the combined uptake by roots and mycorrhizal fungi. Thus, the mycorrhizal effects on Cs uptake may be confounded by the effects of plant nutrient status, plant size, and root density. To distinguish the contribution of fungal transport, one may employ compartmented growth systems in which Cs is made available only to mycorrhizal hyphae (17). This was recently attempted with soil-free in vitro cultures (6) in a study in which a mycorrhizal fungus was grown in a jellified medium in petri plates containing a symbiotic root organ culture. In this experiment, both fungal uptake of radiocesium and its translocation towards the host root could be demonstrated. However, the third and final step in the fungal transport process, which is transfer from fungus to host plant, could not be shown. Thus, even this report was inconclusive with respect to the role that AM play in plant uptake of radiocesium.

The objective of the present investigation was to demonstrate whether or not AM fungi (AMF) can transport radiocesium from soil to plants. We approached this objective by using compartmented growth systems containing uninoculated soil or soil inoculated with different AMF. Furthermore, we used both positive and negative control treatments. The former would verify the presence of the AMF in the labeling compartments and mycorrhizal functioning through transport of a second radiotracer for which AM are known to be efficient. The negative control would account for any radiocesium uptake by roots resulting from diffusion and mass flow through the buffer zone of the labeling compartment. In one experiment, we included the isotopes ⁸⁵Sr and ¹⁵²Eu, which also represent nuclides that pose a risk to human health, to obtain complementary data on their mobilities and transport as affected by AM.

MATERIALS AND METHODS

Experiment 1.

Individual plants of subterranean clover (Trifolium subterraneum L.) or eucalyptus (Eucalyptus globulus L.) were grown from pregerminated seeds in compartmented polyvinyl chloride (PVC) tubes (50-mm diameter, 20-cm height) (Fig. 1) containing a mixed substrate (steam-pasteurized sand-expanded clay-peat-clay loam soil, 2:1:1:1, by volume) inoculated with Glomus mosseae BEG 69 (200 spores extracted from pot cultures with leek) and an inoculum filtrate (10 ml of a filtrate obtained by sifting 50 g of crude inoculum plus 1,000 ml of distilled H₂O through a 20-μm-pore-size nylon mesh) or with only an inoculum filtrate (nonmycorrhizal control treatments). Plants were kept in a growth chamber (300 μmol of photosynthetically active radiation [PAR] m⁻² s⁻¹, 16 h of light day⁻¹, 60 to 70% rH) and watered daily with a high-K (1 mM), low-P (0.1 mM) nutrient solution (18) to 60% of the substrate’s water-holding capacity (WHC) for 70 days. Compartmentation was achieved by introducing a nylon mesh sandwich (three layers with respective pore sizes of 20, 700, and 20 μm, creating a 0.5-mm air gap) 3 cm from the bottom of the tube. Each bottom compartment, or hyphal compartment (HC), was dually labeled with 1 ml of a 10 mM Na₂HPO₄ solution labeled with 37 kBq of ³²PO₄ (NEN) ml⁻¹ and with 1 ml of a 10 mM CsCl solution labeled with 37 kBq of ¹³⁷Cs (Isotope Products Europe) ml⁻¹ applied to the substrate in the distal end of the HC. Each of the four treatments was replicated three times. Plants were harvested 20 days after labeling by cutting aerial parts at the soil surface and recovering roots by gentle shaking and subsequent washing. Activities of ³²P and ¹³⁷Cs were measured on dried (60°C) plant material with a Cobra Autogamma scintillation counter (Packard). We corrected for counting efficiency by using standards with known activity and separating the activity of the two isotopes by comparing counts of the window of 590 to 690 keV at harvest and 180 days after harvest when ³²P had died out. Root colonization by AMF was estimated by a line intersect method after clearing roots and staining AM fungi with trypan blue (21).

FIG. 1.

Schematic diagram of the three growth systems used in experiments 1 to 3. Black arrows indicate isotope placement, and white arrows indicate mesh barriers restricting root growth towards labeled soil.

Experiment 2.

Single maize plants (Zea mays L.) inoculated with Glomus intraradices BEG 157 (25 ml of crude inoculum containing >2,200 spores obtained from pot cultures with Plantago lanceolata L.) or autoclaved pot culture medium and a filtrate of the mycorrhizal inoculum (obtained by filtering 100 g of inoculum plus 1,000 ml of H₂O_dest. through a Whatman no. 1 filter) were grown in branched, compartmented PVC star pots (six radial root compartments [RCs] with 80-mm diameters, 14-cm heights, and branching at a height of 7 cm, connected by buffer compartments of different lengths [4, 6, 8, 10, 12, and 14 cm] and with diameters of 2 cm [see Fig. 1 and reference 13] to a central labeling compartment [80-mm diameter, 14-cm height]) in a growth chamber (500 μmol of PAR m⁻² s⁻¹, 16 h of light day⁻¹, 70% rH) and watered automatically with a tensiometer-controlled watering facility to 50 to 60% of WHC. All compartments contained a mixed substrate (autoclaved sand-expanded clay-soil, 2:2:1, by volume), and compartmentation was achieved by placing a 20-μm-pore-size nylon mesh between the compartments. Each treatment was replicated three times, and each labeling compartment received 401 kBq of ¹³⁴Cs (Amersham) and 141 kBq of ⁶⁵ZnCl₂ (NEN) as carrier-free solutions 3 weeks after the plants were sowed. Plants were fertilized weekly with a Hoagland nutrient solution without P, supplying 82.1 mg of K per plant, and harvested after 46 days. AM colonization of roots was assessed as described above, and the activity of radioisotopes on dried roots and shoots was measured with a high-purity Ge detector (¹³⁴Cs, 605 and 795 keV; and ⁶⁵Zn, 511 and 1,116 keV) featuring borehole geometry and an anticoincidence shield, with counting for up to 200,000 s (55.5 h).

Experiment 3.

Barrel medic (Medicago truncatula L.) plants inoculated with G. intraradices BEG 87 (50 g of crude inoculum obtained from pot cultures with T. subterraneum L.) or 50 g of growth substrate were grown in compartmented PVC tubes (50-mm diameters, 30-cm heights) (Fig. 1) with one side arm at a 17-cm height. The pots contained a nutrient-amended (34 mg of K kg⁻¹ and no P) substrate (electron beam-pasteurized sand-sandy loam soil, 1:1, by weight) and were maintained in a growth chamber for 27, 37, or 47 days (500 μmol of PAR m⁻² s⁻¹, 16 h of light day⁻¹, 70% rH), being watered daily with deionized water to 55% of the WHC. Compartmentation was achieved by placing either a 20- or 700-μm (for the last harvest only)-pore-size nylon mesh between the vertical RC and the horizontal HC, allowing passage of only AM hyphae or both hyphae and roots, respectively. Each treatment was replicated three times. HCs were filled with two layers of soil and sand, one 0 to 25 mm from the mesh without radioisotopes, and one 25 to 40 mm from the mesh containing either ³³PO₄ (Amersham) at 4 kBq g⁻¹ (for the last harvest only) or a mixture of ¹³⁴Cs, ⁸⁵Sr, and ¹⁵²Eu (Amersham) each at 1 kBq g⁻¹ added as carrier-free solutions and mixed into the soil before sowing. The plants were harvested and the roots were assessed for AM colonization as described above. Activity of ³³P was measured on dried roots and shoots by liquid scintillation (Packard 1900 TR) analysis after mineralization in HClO₄-HNO₃ (1:4), and ¹³⁴Cs, ⁸⁵Sr, and ¹⁵²Eu were measured on dried roots, shoots, and soil from HCs (sliced in sections of 6 to 8 mm perpendicularly to the compartment interface) by use of a Ge detector (¹³⁴Cs, 605 and 795 keV; ⁸⁵Sr, 514 keV; and ¹⁵²Eu, 121 and 344 keV), with counting for up to 250,600 s (ca. 70 h).

RESULTS

Experiment 1.

Thirty-nine percent of the total root length of clover became colonized by mycorrhiza following inoculation with G. mosseae, while the corresponding value for eucalyptus was 35%. Uninoculated plants remained nonmycorrhizal. Mycorrhizal colonization led to a 30% growth increase in eucalyptus, while mycorrhizal clover was 20% smaller than nonmycorrhizal controls (results not shown). Both symbioses transported a substantial amount of ³²P that could be detected in both roots and shoots at harvest (Table 1). Nonmycorrhizal clover contained no ³²P in either roots or shoots, while roots and shoots of nonmycorrhizal eucalyptus contained 10 and 5% of the ³²P in the respective parts of mycorrhizal plants. The quantities of ¹³⁷Cs recovered in roots and shoots of clover were <0.02 Bq g⁻¹ and <0.01 Bq per plant of the 37 kBq initially added per pot, and there were no differences between mycorrhizal and nonmycorrhizal plants. In eucalyptus, roots of nonmycorrhizal plants contained slightly increased amounts of ¹³⁷Cs (11.5 Bq g⁻¹), which was not significantly different from the activity in mycorrhizal roots (0.68 Bq g⁻¹). In shoots of eucalyptus, the activity was again very low (<0.02 to 0.14 Bq g⁻¹) and similar in mycorrhizal and nonmycorrhizal plants.

TABLE 1.

Isotope activity in shoots and roots of three plant species

Plant part	Activity (Bq g−1) ina:
	Clover		Eucalyptus		Maizeb
	137Cs	32P	137Cs	32P	137Cs			65Zn
					4 cm	6 cm	8 cm	4 cm	6 cm	8 cm
Shoots
Nonmycorrhizal	<0.02	1	0.07	125	<0.01	<0.01	<0.01	<0.10	<0.13	<0.02
Mycorrhizal	<0.02	359*	<0.02	2,620*	<0.38	<0.29	<0.08	895*	1,024*	116*
Roots
Nonmycorrhizal	<0.02	1	11.5	1,730	ND	ND	ND	ND	ND	ND
Mycorrhizal	<0.02	4,413*	0.68	17,079*	<0.39	<0.27	<0.10	581	736	178

^aActivity was measured following labeling in compartments accessible only to mycorrhizal hyphae. All values preceded by a < symbol represent counts below the detection limit. *, significantly different between nonmycorrhizal and mycorrhizal treatments (P < 0.05, n = 3).

^bData for uptake in maize plants include three treatments with labeled soil placed at 4, 6, or 8 cm from the plant roots. ND, not determined.

Experiment 2.

Ninety-six percent of the total root length of inoculated maize plants was colonized by G. intraradices, and this colonization caused a dry weight reduction in mycorrhizal plants of approximately 8.5%. As much as 6 to 8 m of extraradical hyphae g⁻¹ was detected in corresponding buffer compartments up to 10 cm from the host roots in parallel pots prepared with the same growth substrate. ⁶⁵Zn was transported by AM hyphae over all tested distances (4 to 14 cm), though quantities decreased with increasing distance between plant roots and labeled soil (Table 1) and very low values were recorded for transport over >8 cm (data not shown). No ⁶⁵Zn was detected in nonmycorrhizal plants. The amount of ¹³⁴Cs recovered in plants was below the detection limit and thus >3 orders of magnitude below the activity of ⁶⁵Zn. The transported ⁶⁵Zn represented up to 3.5% of the activity in the labeling compartment.

Experiment 3.

Barrel medic plants inoculated with G. intraradices had 56, 60, or 65% of their root lengths colonized at the day 27, 37, and 47 harvests, respectively, while noninoculated plants remained nonmycorrhizal. Plant growth was enhanced by mycorrhiza at all harvests, with an increase in dry weight between 100 and 200% (results not shown). Positive control treatments labeled with ³³P demonstrated a highly efficient and functional symbiosis that transported 46% of the initially added 160 kBq from the HC to the plant, while corresponding nonmycorrhizal plants contained either nil (two plants) or very little (one plant) ³³P. Diffusion of Cs from the site of labeling towards roots was limited, with less than 1% of added ¹³⁴Cs being recovered only 5 mm from the labeled soil. At the soil section closest to the roots and farthest away from the labeled soil, the ¹³⁴Cs activity was approximately 6 orders of magnitude lower than in the labeled soil (Fig. 2). This diffusion was not affected by the presence of AM hyphae, while the presence of roots slightly increased the diffusion; this increase could also have resulted from the presence of ¹³⁴Cs-containing root pieces in the soil sections analyzed. ¹³⁴Cs diffusion did not result in significantly higher ¹³⁴Cs activity in individual layers of the buffer soil between days 27 and 47 (results not shown). Mycorrhizal and nonmycorrhizal plants having root access to the labeled soil (700-μm-pore-size mesh separating the RC and the HC) contained 130 to 190 times as much ¹³⁴Cs as did plants which could access the labeled soil by AM hyphae only (37-μm-pore-size mesh separating the RC and the HC) (Table 2). For the latter group of plants, ¹³⁴Cs concentrations decreased with time. No significant differences in uptake of ¹³⁴Cs were observed between mycorrhizal and nonmycorrhizal plants at any harvest time. Diffusion of ⁸⁵Sr through the buffer compartment was very high and precluded detection of a possible low transport capacity. ¹⁵²Eu diffusion was very similar to that of ¹³⁴Cs, and thus the very low activities in plants, being similar in mycorrhizal and nonmycorrhizal treatments, showed that no measurable fungal transport of ¹⁵²Eu took place.

FIG. 2.

Activity of ¹³⁴Cs, ⁸⁵Sr, and ¹⁵²Eu in labeled soil and buffer soil of the horizontal labeling compartment of pots containing nonmycorrhizal (NM) or mycorrhizal (Myc) barrel medic plants as a function of distance from the compartment interface made up by 20- or 700-μm-pore-size mesh (experiment 2). The vertical dotted line indicates the interface between labeled soil and buffer soil. Error bars indicate standard deviations (n = 3).

TABLE 2.

Isotope activity in shoots and roots of barrel medic

Plant part	Activitya
	134Cs				33P	85Sr				152Eu
	Root access	Hyphal access			Hyphal access	Root access	Hyphal access			Root access	Hyphal access
	27b	27	37	47	47	27	27	37	47	27	27	37	47
Shoots
Nonmycorrhizal	6.4	0.12	0.03	0.03	119	485	0.32	0.28	0.84	2.52	0.16	0.04	0.03
Mycorrhizal	9.5	0.06	0.04	0.02	15,371*	332	0.56	2.50	1.10	0.82	0.08	0.02	0.01
Roots
Nonmycorrhizal	2.6	0.28	0.07	0.05	26	29	0.39	0.17	0.50	2.64	0.28	0.07	0.05
Mycorrhizal	3.0	0.14	0.04	0.05	17,711*	12	0.34	1.16	0.55	2.95	0.14	0.04	0.04

^aActivity (in becquerels per gram) measured following labeling in compartments accessible only to mycorrhizal hyphae (Hyphal access) or to hyphae and roots (Root access). *, significant difference between nonmycorrhizal and mycorrhizal treatments.

^bLabeling period given in number of days.

DISCUSSION

Mycorrhizal transport is well documented for a number of plant nutrients like PO₄ (11), NH₄ (16), NO₃ (15), Zn (3), Cu (24), and Fe (4, 24), while it is generally accepted that it is insignificant or does not occur for others nutrients, like K, Ca, S, Mo, B, etc. (8, 22). Yet hyphae take up K from the soil, as the concentration of K is high within extraradical hyphae (30). Whether or not mycorrhizal transport of an element occurs seems to be related to the different mobilities of these elements in soil. It appears that fungal transport of the most immobile nutrients was a prerequisite for the colonization of land by terrestrial plants some 420 million years ago (32, 34) and that it has been maintained during the coevolution of plants and mycorrhizal fungi because it constitutes an important competitive advantage for the plants that form these symbioses (34). In some cases with nonessential elements like Cd, Cr, Ni, Pb, and U, hyphal translocation has been demonstrated, whereas transfer from fungus to host does not seem to occur (19, 29). The latter fact may be due to a lack of specific transporters at the interface of exchange between the two organisms or more-or-less specific mechanisms of sequestration within the fungus that render nonessential elements immobilized within the fungus. Cs, Sr, and Eu, are nonessential elements, so no evolutionary benefit would be obtained by symbioses that can transport these elements. Furthermore, no fungal transport has been observed for the Cs analog K (8, 22), and the Sr analog Ca is not transported by AMF (8, 25). The concentration of Ca in extraradical hyphae is significantly lower than that of K (30), and thus, Cs may not enter AM plants through the fungus, even when it is mixed with a K pool available to plants.

Mycorrhizal effects on plant growth are difficult to avoid in pot experiments, which hampers any straightforward interpretation of mycorrhizal effects on any parameter other than the growth-limiting factor (which is most commonly P availability). Tracer isotopes and compartmented pots, in which nutrient transport may be studied without such confounding effects, have thus been used extensively since the method was first published (12).

Cs uptake and translocation by an AMF have been demonstrated in vitro with root organ cultures in which carrier-free ¹³⁴Cs was added to a liquid medium with a very low concentration of K (6). This system precludes the verification of transfer from fungus to host plant due to the lack of shoots and the difficulties associated with precise localization of Cs, which may be contained exclusively within intraradical fungal structures. Radiocesium from the Chernobyl accident was commonly found in soil as particulate fallout (hot particles), even as far away from the reactor site as Norway (31). In soil, K/Cs ratios are on the order of >100,000. This fact, together with the limited bioavailability of radiocesium in hot particles, makes the effective K/¹³⁴Cs and K/¹³⁷Cs ratios orders of magnitude higher. Finally, Cs gradually becomes irreversibly fixed when it comes in contact with clay minerals in soil (5). These are certainly important discrepancies that disqualify the in vitro system from being an ecologically relevant measurement of mycorrhizal Cs transport to plants. When K levels are manipulated experimentally in pot experiments with soil, plant uptake of Cs decreases with increasing amounts of available K; this also occurs when plants are mycorrhizal (2, 28).

Negative results rarely constitute conclusive evidence for the nonexistence of a process. Nevertheless, we believe that the present results are generally valid and that the AMF have no ecologically or toxicologically relevant capacity to transport Cs to their host plants under natural conditions. In our experiments, we used four common mycotrophic plants that are representative of both agricultural and natural vegetation and two fungal species that commonly colonize these and a wide range of other plants (14). Furthermore, radiocesium was added in high amounts, both with and without nonradioactive carrier Cs and at different but realistic concentrations of available K. The Cs in our experiments was thus highly bioavailable, more so than radiocesium in hot particles or contained in soil after prolonged contact with clay minerals. The extremely low values for Cs uptake we recorded therefore actually overestimate the transport capacity of these fungi under natural conditions.

Agricultural ecosystems are likely to behave quite differently from forests, where radioactive Cs isotopes from fallout can remain bioavailable for decades, due to extensive recycling in ectomycorrhizal fungi that limits leaching from an organic surface horizon into deeper, clay-containing soil. In agricultural ecosystems, even the uppermost soil may be rich in clay that fixes Cs more or less irreversibly, and the routine application of K fertilizers will further reduce the risk of transfer to the human food chain through a displacement of radiocesium for K. We thus conclude that AM are highly unlikely to play any significant role in plant uptake of radiocesium through the process of fungal transport. Indirect effects of AM may still affect Cs accumulation in plants, as the symbiosis may alter plant growth and/or root development and thus possibly dilute absorbed Cs.