summaryLittle information currently exists on species diversity in communities of arbuscular mycorrhizal fungi (AMF), mainly owing to difficulties in identification of field extracted spores on the basis of morphology. The possibility was explored to identify individual AMF spores from the field on the basis of a molecular marker, namely the nuclear ribosomal DNA encoding the highly conserved 5.8S rRNA with the two flanking internal transcribed spacers (ITS region), known to vary between species. A technique involving polymerase chain reaction followed by restriction fragment length polymorphism analysis (PCR–RFLP) was developed to amplify and characterize the ITS region from single AMF spores. PCR reactions with extracts from single spores of three AMF species, raised under glasshouse conditions, yielded reproducibly a single amplification product of the ITS region in sufficient amounts to allow cleavage with several restriction enzymes. The size of the ITS region, c. 600 base pairs, varied only slightly between species. Digestion of the PCR products with the restriction enzymes Hinfl and Taq I resulted in banding patterns that were reproducible for different individual spores of a given species, but showed clear differences between the three species tested. The sum of the fragment sizes was sometimes greater than the size of the original PCR product, e.g. in Glomus mosseae. Clones of the amplification product from a single spore of this fungus were obtained and sequenced. This yielded two closely related but different sequences, indicating that two different ITS regions co‐existed in the spore. The RLFP pattern of the amplification product of the spore was a result of an amalgamation of these two sequences. The technique was applied to AMF spores collected from a species‐rich grassland. Spores were sorted into morphological groups on the basis of their colour, size, and shape, and then subjected to PCR–RFLP analysis. In some morphological groups, a large percentage of spores failed to yield an amplification product, probably because they had lost their contents. A group of Glomus spores yielding amplification products in the majority of cases was further investigated: PCR RFLP analysis on 10 individual spores from the field produced 10 different patterns. Similar results were obtained with other groups of spores. The results suggest that the diversity in natural AMF communities and the genetic diversity within individual spores might he much greater than previously thought.

Ethylene production and activities of chitinase and ß-1,3-glucanase, two parameters often induced in the plant's response to pathogenic fungi, were measured in the roots of various plants after inoculation with the vesicular-arbuscular (VA) mycorrhizal fungus Glomus mosseae and after mock-inoculation in the absence of VA mycorrhizal fungi. Tomato, a host plant for VA mycorrhizal fungi, was compared in this respect with three species of non-host plants, rape, spinach and lupin. In the case of rape, two cultivars were examined, cv. Jet Neuf (rape 0) and Arabella (rape 00) with normal and low levels of glucosinolates, respectively. Only roots of rape 00, spinach and lupin showed a slightly enhanced ethylene production after inoculation. In roots of tomato, spinach and lupin, chitinase activity was not affected during the first days after inoculation but was depressed afterwards in comparison to mock-inoculated controls. In roots of rape 0 and rape 00, chitinase was weakly induced upon inoculation. Levels of ß-1,3-glucanase were not altered in tomato roots during VA mycorrhiza establishment, but they were slightly enhanced in all non-host plants in an initial stage after inoculation. In a later stage, most of the inoculated non-host plants showed less activity than the mock-inoculated controls. We conclude that both host and non-host plants perceive the VA mycorrhizal fungus and respond to its presence. Since the different non-host plants showed different response patterns, none of the reactions studied can be taken as a general indicator for the inability of plants to entertain the VA mycorrhizal symbiosis.

Various container systems have been described in which soil regions available to hyphae only are separated from the mycorrhizal root region by 30–60 μm mesh screens to study nutrient exchange between plants and fungi in the mycorrhizal symbiosis. The screens designed up to now prevent penetration by roots but allow easy passage of fungal hyphae as well as diffusion or mass flow of water and nutrient solutions. We tested hydrophobic microporous polytetrafluorethylene (PTFE) membranes (GORE-TEXR) with 5 to 15 μm diameter pores in an attempt to obtain a better seal between compartments and to prevent uncontrolled nutrient transport by diffusion or mass flow. We found that these membranes completely prevented diffusion or mass flow of ions between two soil compartments but could be penetrated easily by the vesicular-arbuscular mycorrhizal fungus Glomus mosseae, as demonstrated by the rapid colonization of soybean roots (Glycine max L.) from an inoculum across the membranes.