Quantification of mycelial development of arbuscular mycorrhizal fungi using image analysis

Quantitative and reproducible information concerning the development of the extraradical mycelium of arbuscular mycorrhizal fungi (AMF) is lacking due to the difficulties in extracting, identifying and estimating hyphal lengths. In this study, using a rhizobox growth system, the lengths of hyphae of AMF estimated using an image analysis system were not significantly different from data obtained by a trained observer using a modified grid-line intersect method. The assessment of lengths of hyphae on membrane filters or slides was, however, much quicker using image analysis, and allowed the complete sample to be quantified, unlike the grid-line method where a limited number of fields of view are assessed. The image analysis procedure is objective, observer-independent and less laborious than the manual method of assessment. Of the four different methods of sample preparation compared, membrane filter methods were found to be the most appropriate for quantitative sampling from three non-soil substrates. Glomus monosporum (UKC M3) produced twice as much extraradical mycelium and hyphal length per centimetre of colonised root than G. geosporum (BEG 11) on both leek and linseed in a durite sand at final harvest (63 days). Both AMF also produced more hyphal length per centimetre of colonised root on linseed than on leek. The spatial distribution of both AMF, however, was similar in durite sand and no correlation with levels of NaHCO₃-extractable phosphorus was noted. In a third experiment, with G. manihotis (UKC INDO-1) colonising a tropical forage legume, Pueraria phaseoloides, in two other growth substrates, a different pattern of development of the extra-radical mycelium was observed. Because of a higher content of particulate matter, which collected on the membrane filters, the extraction technique had to be modified to give optimal performance of the image analysis system.