The causes and consequences of biodiversity are central themes in ecology. Perhaps one reason for much of the current interest in biodiversity is the belief that the loss of species (by extinction) or their gain (by invasion) will significantly influence ecosystem function. Arbuscular mycorrhizal (AM) fungi are components of most terrestrial ecosystems and, while many research programs have shown that variability among species or isolates of AM fungi does occur (Giovannetti & Gianinazzi-Pearson, 1994), the basis for this variability and its consequences to the function of communities and ecosystems remains largely unexplored. Smith et al. (pp. 357–366 in this issue) now show clearly that ecologically significant functional diversity exists among AM fungal species in the regions of the soil from which they absorb phosphate, and their results suggest that such diversity may have significant ecological consequences.

Many human activities, such as ore mining and smeltering, sewage sludge treatment and fossil fuel consumption, result in toxic soil concentrations of ‘heavy metals’ (Al, Cd, Co, Cr, Cu, Hg, Mn, Ni, Pb, Ti, Zn and others) (Gadd, 1993). There are also natural soils, such as serpentine, with levels of heavy metals that inhibit or preclude the growth of many plants and soil micro-organisms. However, certain plants and microorganisms do grow in these metalliferous sites. Understanding the physiology, ecology and evolution of tolerance to elevated soil metal concentrations is important in an applied setting, and is also of interest in theoretical biology. Applied importance relates to the improvement of forest health in areas subject to increasing pollution, rehabilitation of severely polluted sites by phytostabilization of metals, and metal removal using hyperaccumulating plants (Krämer, 2000; Ernst, 2000). Areas of theoretical interest include the evolution of local adaptation (Sork et al., 1993) and how it is shaped by the combined influences of natural selection, gene flow and genetic architecture, as well as metal influences on various species interactions (Pollard, 2000). A paper appears on pages 367–379 in this issue by Jan Colpaert and coworkers which adroitly combines the disparate fields of physiology, genetics and ecology to answer several outstanding questions concerning heavy metal tolerance in mycorrhizal fungi.

Mycorrhizal fungi, which interact mutualistically with the majority of plant species, are well known for improving the P status of their hosts (Smith & Read, 1997). Some mycorrhizal fungi are also able to mobilize N and P from organic substrates and to provide plants with improved micronutrient and water acquisition, pathogen resistance, and a variety of other benefits (Smith & Read, 1997). One of these additional benefits is the amelioration of toxicity in metalliferous soils.

Summary 241

I. INTRODUCTION: CAESIUM IN THE ENVIRONMENT 242

II. UPTAKE OF CAESIUM BY PLANT ROOTS 243

1. Evidence for multiple mechanisms of Cs+uptake by plant roots 243

2. Caesium uptake is affected by the presence of other cations 244

3. Caesium inhibits the uptake of other cations 244

III. MOLECULAR MECHANISMS CATALYSING CAESIUM UPTAKE 245

1. ‘High-affinity’ transport mechanisms 245

2. Inward-rectifying potassium (KIR) channels 245

3. Outward-rectifying potassium (KOR) channels 248

4. Voltage-insensitive cation (VIC) channels 249

5. Ca2+-permeable channels 249

IV. MODELLING CAESIUM INFLUX TO ROOT CELLS 249

1. Predicted Cs+influx through high-affinity mechanisms 250

2. Predicted Cs+influx through cation channels 250

3. Predicted dependence of Cs+influx on [Cs+]ext 252

V. PERSPECTIVE 253

Acknowledgements 254

References 254

Caesium (Cs) is a Group I alkali metal with chemical properties similar to potassium (K). It is present in solution as the monovalent cation Cs+. Concentrations of the stable caesium isotope 133Cs in soils occur up to 25 μg g−1 dry soil. This corresponds to low micromolar Cs+ concentrations in soil solutions. There is no known role for Cs in plant nutrition, but excessive Cs can be toxic to plants. Studies of the mechanism of Cs+ uptake are important for understanding the implications arising from releases of radioisotopes of Cs, which are produced in nuclear reactors and thermonuclear explosions. Two radioisotopes of Cs (134Cs and 137Cs) are of environmental concern owing to their relatively long half-lives, emissions of β and γ radiation during decay and rapid incorporation into biological systems. The soil concentrations of these radioisotopes are six orders of magnitude lower than those of 133Cs. Early physiological studies demonstrated that K+ and Cs+ competed for influx to excised roots, suggesting that the influx of these cations to root cells is mediated by the same molecular mechanism(s). The molecular identity and/or electrophysiological signature of many K+ transporters expressed in the plasma membrane of root cells have been described. The inward-rectifying K+ (KIR), outward-rectifying K+ (KOR) and voltage-insensitive cation (VIC) channels are all permeable to Cs+ and, by analogy with their bacterial counterparts, it is likely that ‘high-affinity’ K+/H+ symporters (tentatively ascribed here to KUP genes) also transport Cs+. By modelling cation fluxes through these transporters into a stereotypical root cell, it can be predicted that VIC channels mediate most (30–90%) of the Cs+ influx under physiological conditions and that the KUP transporters mediate the bulk of the remainder. Cation influx through KIR channels is likely to be blocked by extracellular Cs+ under typical ionic conditions in the soil. Further simulations suggest that the combined Cs+ influxes through VIC channels and KUP transporters can produce the characteristic ‘dual isotherm’ relationship between Cs+ influx to excised roots and external Cs+ concentrations below 200 μM. Thus, molecular targets for modulating Cs+ influx to root cells have been identified. This information can be used to direct future genetic modification of plants, allowing them to accumulate more, or less, Cs and thereby to remediate contaminated sites.

Summary 257

I. DEFINING AND QUANTIFYING OCEANICITY 257

II. ECOLOGICAL CONSEQUENCES OF OCEANICITY 261

III. ECOLOGICAL HISTORY OF OCEANICITY IN WESTERN EUROPE 263

IV. POSITIVE AND NEGATIVE INFLUENCES OF OCEANICITY 267

1. Species versus communities 267

2. Case study – Primula scotica – climatic effects on reproduction 268

3. Case study – cranberry production and anoxia tolerance 268

V. MODIFYING EFFECTS OF OCEANICITY ON PLANT DISTRIBUTION 269

VI. PHYSIOLOGICAL IMPACT OF WARM WINTERS 271

1. Phenology 271

2. Metabolic consequences of warm winters 272

3. Warm winters and mountain-top vegetation 274

VII. TREELINES AND OCEANICITY 276

VIII. CONCLUSIONS 278

Acknowledgements 279

References 279

A cyclic behaviour in the intensity of maritime conditions which varies with the periodic behaviour of the North Atlantic oscillation has recently become apparent in the climatic record of northern Europe. Periodic increases in oceanicity are usually viewed as having a positive effect on plant survival, as milder winters, reduction of temperature extremes, low risk of exposure to frost, and freedom from drought reduce many aspects of environmental stress. However, warmer winters in maritime environments may also have a powerful influence in creating habitats that are unfavourable for many species. The dangers of long periods of soil saturation for overwintering plants, soaking injury to germinating seeds, premature bud burst in spring, depression of treelines by increased cloud cover and high lapse rates, as well as the constant leaching of soils, are all negative aspects of the maritime environment. Woody species in which root dormancy is delayed by mild winters are particularly vulnerable to the consequences of winter flooding. Subsequent re-exposure to oxygen as water tables fall in spring can aggravate flooding damage through post-anoxic injury, leading to severe dieback of anchoring roots. Soil leaching, particularly after human disturbance, can induce nutrient deficiencies and the establishment of oligotrophic communities with reduced productivity. Podzolization and iron pan formation have in the past facilitated the processes of paludification in oceanic regions. The resulting spread of bogs and acid moorlands can further reduce the potential for productivity in both agriculture and natural plant communities. Given the probability that current climatic trends may increase the degree of oceanicity in western and northern regions of Europe, the potentially negative consequences of such a climatic change need to be considered in relation to future ecological changes and their consequences for conservation and land use.

N2 fixation, measured as acetylene reduction, was studied in laboratory cultures and in natural assemblages (both as a mixed population and as individually picked colonies) of the heterocystous cyanobacteria Aphanizomenon sp. and Nodularia spp. from the Baltic Sea. During a diurnal cycle of alternating light and darkness, these organisms reduced acetylene predominantly during the period of illumination, although considerable activity was also observed during the dark period. In both laboratory cultures and natural populations N2 fixation was saturated below a photon flux density of 600 μm−2 s−1. In cyanobacterial blooms in the Baltic Sea, nitrogenase activity was mostly confined to the surface layers. Samples collected from greater depths did not possess the same capacity for acetylene reduction as samples from the surface itself, even when incubated at the photon flux density prevailing in surface waters. This suggests that, with respect to N2 fixation, Baltic cyanobacteria are adapted to the intensity of illumination that they are currently experiencing.

Axial and radial transport and the accumulation of photoassimilates in carrot taproot were studied using 14C labelling and autoradiography. Axial transport of the 14C labelled assimilates inside the taproot was rapid and occurred mainly in the young phloem found in rows radiating from the cambium. The radial transport of the assimilate inward (to cambium, xylem zone and pith) and outward (to phloem zone and periderm) from the conducting phloem was an order of magnitude slower than the longitudinal transport and was probably mainly diffusive. The cambial zone of the taproot presented a partial barrier in the inward path of the assimilate to the xylem zone. We suggest that this is due to the cambium comprising a strong sink for the assimilate on the basis that our previous work has shown that it contains very low concentrations of free sucrose. By contrast, a high accumulation of nonsoluble 14C was found in the cambium region in good agreement with the active growth of this zone. Autoradiography following the feeding of 14C labelled sugars to excised sections of taproot indicated that only a ring of cells at and/or just within the cambium take up sugars from the apoplast. This indicates that radial movement in the phloem and pith must be symplastic. An apoplastic step between phloem and xylem is possible. The rapid uptake of sugars from the apoplast at this point might represent a mechanism for keeping photoassimilates away from the transpiration stream and re-location back to the leaves.

Uptake, translocation and metabolism of 14C-labelled formaldehyde in the leaves of Epipremnum aureum (Golden Potho) and Ficus benjamina (Weeping Fig) were investigated. Plants were exposed in light and dark to 14C-formaldehyde (500 μg m−3) in gas exposure chambers. The amount of 14C-incorporation into the soluble (water-extractable) and insoluble fractions of leaves, stem sections and roots was determined. The soluble 14C-activity was fractionated by ion exchange chromatography followed by thin-layer chromatography/autoradiography. Approximately 60–70% of the applied 14C-formaldehyde was recovered from the plants. In the light about five times more 14C-formaldehyde was assimilated than in the dark. The amount of 14C-label derived from 14C-formaldehyde, which was incorporated into acid-stable metabolites, was enhanced to an even larger extent in the light. The 14C-activity pattern closely resembled the general labelling spectrum of photosynthates, obtained after a 14CO2 exposure. A substantial amount of labelled material, mostly sucrose, was translocated into the stems and roots. Our results suggest that in the light 14C enters the Calvin cycle after an enzymatic two-step oxidation process of 14C-formaldehyde to 14CO2. The activities of the respective enzymes, formaldehyde dehydrogenase and formate dehydrogenase, were determined. Among 27 ‘leafy’ indoor decorative plants, a screening experiment revealed no outstanding species with regard to its capacity for metabolism of formaldehyde, and rate of uptake through stomata was too low to justify claims that plants contribute usefully to indoor air purification.

Mosses have been used as biomonitors of atmospheric pollution for some years, but few studies have been carried out on the effect of NOx emissions from traffic on moss tissue N. Eight species of moss (102 samples) growing on walls or roofs next to roads exposed to different traffic densities were collected from urban and rural sites in the UK. The shoots were sampled for total N, their stable isotope 15N/14N content (δ15N) and heavy metal content (Pb, Zn). There was a lack of correlation between tissue total N and traffic exposure, but a very good correlation between traffic exposure and tissue δ15N. Plants collected near motorways or busy urban roads had δ15N values ranging between +6 and −1‰, while in rural areas with hardly any traffic these ranged from −2 to −12‰. In a separate survey of mosses, the average δ15N of shoots from busy roadsides in London was +3.66‰, whereas from samples collected from farm buildings near poultry or cattle pens it was −7.8‰. This indicates that the two main atmospheric N sources, NOx and NHx, have different δ15N signatures, the former tending to be positive and the latter negative. Tissue concentrations of both Pb and Zn show a strong positive correlation with traffic exposure, with Zn in particular being greater than Pb. The results are discussed with regard to the use of moss tissue Zn as a means for monitoring or mapping pollution from vehicles, and of δ15N as an aid to distinguish between urban (NOx) and rural (NHx) forms of N pollution.

Ozone effects on plant species mixtures could depend on the characteristics of the species involved, their mixing ratio, or on environmental conditions. Predicting long-term effects on the dynamics of plant communities requires an understanding of the interactions involved. The present experiment was designed to determine the effects of ozone on grassland species in relation to mixing ratio and soil water content (irrigation) using binary mixtures. The grass Trisetum flavescens was grown in potted replacement-series mixtures with Centaurea jacea (Experiment A) or Trifolium pratense (Experiment B). The plants were exposed to three concentrations of ozone in open-top chambers in two irrigation treatments. Total above-ground dry weight over three growth periods was measured. The competitive ability of T. flavescens was expressed as the competitive ratio (CRT). In Experiment B, total above-ground dry weight was reduced by elevated ozone and by reduced soil moisture, and significant interactions were found for ozone × irrigation and ozone × ratio. In Experiment A these effects were not significant. Under well watered conditions, CRT tended to be reduced by elevated ozone in Experiment A, but increased significantly in Experiment B, indicating the importance of the competing species in modifying the ozone effect on T. flavescens. In both experiments reduced irrigation decreased the magnitude of ozone effects on biomass production, which could be related to observed reductions in specific leaf conductance. The results suggest that under well watered conditions the effect of elevated ozone on the competitive balance between species depends on the species mixture, but that the mixing ratio is less important.

Availability of nitrogen (N) and phosphorus (P) might significantly affect N2 fixation in legumes. The interaction of N and P was studied in common bean (Phaseolus vulgaris), considering their effects on nodulation and N2 fixation, nitrate reductase activity, and the composition of N compounds in xylem sap. The effect of N on the uptake of P by plants was estimated by analysing rhizospheric pH and P concentration in xylem sap and in plant shoots. Inoculated bean plants were grown in pots containing perlite/vermiculite in two experiments with different amounts of P and N. In a third experiment, bean plants were grown on two soil types or on river sand supplied with different concentrations of N. At harvest, shoot growth, number of nodules and mass, and nitrogenase activity were determined. Xylem sap was collected for the determination of ureides, amino acids, nitrate and phosphate concentration. At low nitrate concentration (1 mM), increasing amounts of P promoted both nodule formation and N2 fixation, measured as ureide content in the xylem sap. However, at high nitrate concentration (10 mM), nodulation and N2 fixation did not improve with increased P supply. Glutamine and aspartate were the main organic N compounds transported in the xylem sap of plants grown in low nitrate, whereas asparagine was the dominant N compound in xylem sap from plants grown in high nitrate. Nitrate reductase activity in roots was higher than in shoots of plants grown with low P and high N. In both soils and in the sand experiment, increased application of N decreased nodule mass and number, nitrogenase activity and xylem ureides but increased the concentration of asparagine in xylem sap. Increasing P nutrition improved symbiotic N2 fixation in bean only at low N concentrations. It did not alleviate the inhibitory effect of high nitrate concentration on N2 fixation. A decrease in plant P uptake was observed, as indicated by a lower concentration of P in the xylem sap and shoots, correlating with the amount of N supplied. Simultaneously with the specific inhibition of N2 fixation, high nitrate concentrations might decrease P availability, thus inhibiting even further the symbiotic association because of the high P requirement for nodulation and N2 fixation.

The gelatinous layer (G-layer) of tension-wood fibres in reaction wood of beech showed alterations as a result of the physiological processes involved in the conversion of sapwood into false heartwood or reaction-zone tissue. Using transmitted-light, fluorescence and UV microscopy, polyphenolic compounds were found to infiltrate and encrust the cellulose microfibrils within the G-layer. Experiments with naturally infected and artificially inoculated wood showed that these processes affect the rate and mode of degradation by wood-decaying fungi. Thus, although the ascomycete Ustulina deusta was able to degrade the G-layer from within the lumina of tension-wood fibres in unaltered sapwood, it failed to do so for a prolonged period within false heartwood and reaction zones. In both situations, however, there was some degradation of the underlying secondary wall in the form of erosion troughs which can be attributed to soft rot ‘type II’, and internal cavity formation typical for ‘type I’ attack. The present study indicates that not only cell type, but also alterations in the cell wall structure, affect the activity and degradation mode of decay fungi in beech.

Responses of Medicago truncatula to colonization by two arbuscular mycorrhizal fungi, Scutellospora calospora isolate WUM 12(2) and Glomus caledonium isolate RIS 42, were compared in the light of previous findings that the former fungus can be ineffective as a beneficial microsymbiont with some host plants. The plants were grown individually in two-compartment systems in which a lateral side arm containing soil labelled with 33P was separated from the main soil compartment by a nylon mesh that prevented penetration by roots but not fungal hyphae. Fungal inoculum was applied as a root–soil mixture in a band opposite the side arm. Nonmycorrhizal controls were set up similarly, without inoculum. There were harvests at 28, 35, 42 and 49 d. Both sets of mycorrhizal plants grew better than nonmycorrhizal plants and initially had higher concentrations of P in shoots and roots. Plants grown with S. calospora grew better than plants grown with G. caledonium, and this was associated with somewhat greater fungal colonization in terms of intraradical hyphae and numbers of arbuscules. Scutellospora calospora formed denser hyphae at root surfaces than G. caledonium. By 28 d there were extensive hyphae of both fungi in the side arms, and after 35 d S. calospora produced denser hyphae there than G. caledonium. Nevertheless, there was very little transfer of 33P via S. calospora to the plant at 28 d, and thereafter its transfer increased at a rate only c. 33% of that via G. caledonium. The results showed that plants colonized by S. calospora preferentially obtained P from sites in the main soil chamber relatively close to the roots, compared with plants colonized by G. caledonium. Hence formation of a highly beneficial arbuscular mycorrhizal symbiosis does not necessarily depend on development of hyphae at a distance from the roots or on large-scale translocation of P from distant sites. The results are discussed in relation to previous studies with compartmented systems that have involved the same fungi. Possible causes of the variable effects of S. calospora in symbiosis with different host plants are briefly assessed. Differences in spatial abilities of individual arbuscular mycorrhizal fungi to acquire P might have strong ecological implications for plant growth in soils low in P.

Twenty-one isolates of the ectomycorrhizal fungus Suillus luteus were screened for their tolerance to the heavy metals Zn, Cd, Cu and Ni, measured as inhibition of radial growth and biomass production. Two populations from even-aged pine stands were investigated: 10 isolates were obtained from an area polluted with high levels of Zn, Cd and Cu, and 11 isolates were obtained from a control population located in a nearby unpolluted area. RFLP patterns of the internal transcribed spacer region of the isolates confirmed the morphological identification of the carpophores. All isolates were maintained on basic medium without elevated metals to avoid phenotypically acquired metal tolerance. The in vitro Zn and Cd tolerance of the S. luteus isolates from the polluted habitat were significantly higher than the tolerances measured in the isolates from the nonpolluted site. This observation suggests that the elevated soil metal concentrations might be responsible for the evolution of adaptive Zn and Cd tolerance. Tolerance was maintained in an isolate not exposed to elevated metals for 3 yr. The two S. luteus populations did not differ in tolerance to Cu and Ni. The mechanisms for the adaptive Zn and Cd tolerance are not identical as there was no correlation between response to the two metals; the most Zn-tolerant isolate was the most sensitive for Cd in the metal-tolerant population. Zinc did not accumulate in basidiocarp tissue, whereas Cd levels in basidiocarps were significantly higher in the population on the polluted site. Inter-simple sequence-repeat fingerprints showed that 90% of the isolates were from different individuals. The genetic variation in the population from the unpolluted site was considerably larger than that observed at the polluted site.

There has been conflicting debate over many years regarding the trophic status of Matsutake. Here we examined the morphology, structure and ultrastructure of Pinus densiflora roots collected from a Tricholoma matsutake Shiro within a pure Japanese red pine stand. Molecular investigations (PCR–RFLP analyses) indicated that T. matsutake was the highly dominant fungus within both the Shiro and the colonized root tips, suggesting that reported root morphology modifications can be attributed to T. matsutake infection. The external morphology of Matsutake-colonized roots is consistent with previous descriptions. The presence of extraradical mycelium, mantle, and intracortical Hartig net hyphae indicates clearly that T. matsutake forms an ectomycorrhizal association with P. densiflora in naturally occurring Shiros. The elucidation, for the first time, of the Hartig net ultrastructure at the host–fungus interface provides further and convincing evidence of a conventional ectomycorrhizal association. The progressive blackening, observed from base to tip in dominant mycorrhizal types, due to increased deposition of polyphenol and subsequent necrosis, appears to be a result of infection. However, the presence of highly nucleated vascular tissue indicates the viability of the vascular cylinder in these roots bearing necrotic cortices. Such a preponderance of black necrotic cortical tissues among colonized roots may reflect some atypical behaviour of T. matsutake.

The morphology and anatomy of ectomycorrhizas of Rhizopogon parksii, Rhizopogon vinicolor and Rhizopogon subcaerulescens, and a hypogeous relative, Truncocolumella citrina, synthesized on Douglas fir in glasshouse conditions using spore slurries as inoculum, are described and compared. Mycorrhizas formed with R. parksii and R. vinicolor did not exhibit their characteristic subtuberculate morphology in these tests, but rather had a pinnate form. All species had diagnostic features of ectomycorrhizas: a well-developed Hartig net and a fungal mantle. In addition, several species exhibited crystal inclusions in the outer mantle, usually at the interface between the mantle and soil. Truncocolumella citrina had crystal-like inclusions within the mantle but external to fungal hyphae, a feature rarely described in ectomycorrhizas.

Three ectomycorrhizal legume trees, Microberlinia bisulcata, Tetraberlinia bifoliolata and T. moreliana, form discrete groves in the southern part of Korup National Park, in southwest Cameroon and contribute c. 45–70% of stand basal area locally in a matrix of otherwise species-rich arbuscular mycorrhizal forest. A transplant experiment was performed to assess the importance of ectomycorrhizal infection associated with proximity to parents in seedling establishment of the grove-forming species. Nonectomycorrhizal seedlings of the three species were transplanted into plots of two forest types, one of high (HEM, within-grove) and one of very low (LEM, outside the grove) abundance of all three species as adult trees. For two species (T. moreliana and M. bisulcata) there was no difference in survival over 16 months, but for the third (T. bifoliolata) survival was best in HEM forest, and correlated with the basal area of adult trees of ectomycorrhizal species. Only one species (T. moreliana) increased in biomass over the experimental period; the others declined. There was no effect of forest type on overall growth of any species, but the survivors of two (T. moreliana and M. bisulcata) had heavier stems in the HEM forest. Differences in survival and growth of transplants between the three species were in accord with the ecology of the species as inferred from the frequency distributions of adult tree size in the forest. Seedlings became infected with ectomycorrhizas in both forest types; where there was a difference in extent of infection (T. moreliana) this was not related to survival or growth; and where there was a difference in survival (T. bifoliolata) this was not related to extent of infection. These results confirm that mycorrhizal inoculum associated with conspecific adults is neither a prerequisite nor a guarantee of seedling establishment, but indicates that in some circumstances there might be benefits of being close to parents. Further research is required to unravel the complexities of ectomycorrhizal community structure in this spatially and temporally heterogeneous forest, and to clarify the extent to which the various hosts share ectomycorrhizal partners.

New Phytologist147 (2000), 73–85

In the July 2000 issue of New Phytologist, we published the research paper entitled ‘Dynamics of root systems in native grasslands: effects of elevated atmospheric CO2’ by J. A. Arnone III, J. G. Zaller, E. M. Spehn, P. A. Niklaus, C. E. Wells and C. Körner. Since its publication, it has come to our attention that there are two important errors. In Table 1, for alpine grassland (Switzerland, 2480 m) in year 2 (cool), the observed soil depth should be given as 0–10 cm (as originally printed there is no soil depth given). In Table 2, the period covered should be 28 April 1994 to 6 April 1995, not 28 April 1994 to 6 April 1996, as erroneously stated.

We apologize to the authors and to our readers for these mistakes.