In the present work we compare responses to hypoxia of the carbohydrate content of related plant species, which grow naturally on sites prone to flooding (Senecio aquaticus Hill., Myosotis palustris (L.) Lehm. Rchb.) with plants from habitats with only a low risk of oxygen shortage (Senecio jacobaea L., Myosotis arvensis (L.) Hill.). Whilst the amounts of glucose, fructose and sucrose changed only slightly following hypoxia and peaked at a maximum approximately double that of the aerated control specimens, the fructan content increased fivefold to tenfold. In nearly all the plants studied, fructan became the main polysaccharide reserve. For example, the flooding-tolerant Senecio aquaticus accumulated the highest amount, in particular fructans with a degree of polymerization up to 35 compared with 10 under control conditions. Nearly 70% of the soluble carbohydrates were fructans, compared with 30% under aerated conditions.

In all species tested the starch fraction marginally increased or remained constant.

Fructans were found to accumulate as a response to oxygen deficiency in both flooding-tolerant and intolerant species but with higher absolute values and ratios between fructan and starch in the flooding-tolerant species. At 24 h after the onset of hypoxic treatment, the sugar content rose, in spite of the diminished photosynthetic rates. The ability to accumulate fructans seems to vary in plants coping with oxygen shortage but could be interpreted as advantageous in comparison to starch synthesis in terms of (i) the location of the fructan metabolism in the vacuoles, with no negative feedback in the photosynthetic apparatus and (ii) the possibility of storage of sucrose in the form of fructose polymers without the intermediate stages and energy-using processes which are indispensable for starch synthesis. These results might indicate that fructan metabolism could play a role in the tolerance of oxygen deficiency.

Errors occurred in eqn (10) and in Figures 1, 3, 4, 7, 8 and 10. We apologize to the authors for these errors.

For scientific consistency, the corrected paper is printed here in full to precede the related paper by the same authors.

Wheat stems serve as a store for fructans to buffer the plant against nutritional and environmental influences. It has been suggested that fructan storage influences yield stability and tolerance of environmental factors. Near infra-red spectroscopy (NIR) analysis provides a rapid and accurate assessment of the fructan content of the wheat stem, as well as allowing detection of growth-limiting nutrient stresses, and so is proving to be a useful technique for making crop management decisions. Commercial laboratories using NIR analysis have been tissue-testing crops in the eastern Australian wheat belt since 1993. In healthy, normally developing crops not under stress there is a predictable relationship between nitrogen and fructan. Investigation of the nitrogen and fructan concentrations in commercial crops has confirmed an inverse relationship between these two constituents. The function: Fructan (%)=a+bN%+cN%2 accounted for up to 81% of variation in tissue fructan concentration. In commercial tissue-testing this relationship is used to detect crops under stresses other than nitrogen deficiency. If the fructan concentration deviates by more than 4%, cereal growers are advised that their crop might be subject to other stresses which might reduce its response to applied nitrogen.

The growth and fructan contents of plants of Vernonia herbacea (Vell.) Rusby regenerated from fragments of the underground organ (rhizophore) were analysed during two developmental cycles. When cultivated away from their natural ecosystem, these plants went through the same phenological phases as those naturally grown in the cerrado. The first aerial shoots started developing from initial fragments before the new underground system, the former acting as the main metabolic sink for the fructan present in the initial fragments. The new rhizophores started to develop by the sixth month, as photosynthates were transformed into fructans. Growth parameters measured reached higher values in the second growth cycle, exhibiting growth curves synchronized with those obtained for fructan contents in the new rhizophores. Fructo-oligosaccharides were available mostly during the period of fast growth of aerial organs whilst the fructo-polysaccharides were reduced. Fructan variation indicates that these compounds are metabolized during development. Additionally, the rapid growth, intense production of rhizophores and high fructan content make Vernonia herbacea a promising source of fructan for commercial use.

Pathozone behaviour of soil-borne plant pathogenic fungi is characterized by curves for the change in probability of infection with distance of inoculum from a host. We identify three major components that give rise to the pathozone profile for infection efficiency: the germinability of inoculum, given that it occurs in the pathozone (P1), the growth of the resulting fungal colony outwards to make contact with the host (P2), and the infectivity of mycelium once contact is made (P3). The probability of infection (P) is then given by the product, P1 × P2 × P3. Using Raphanus sativus and two contrasting types of inoculum of Rhizoctonia solani as a model experimental system, we measured each of the three components by quantifying germinability, colony architecture (to determine the chance of contact by one or more hyphae) and the change in susceptibility over time as a measure of infectivity as mycelium arrives from different distances away. The germinability of inoculum is uniform across the pathozone whereas the probability of contact declines with distance, and the net effect of host susceptibility and infectivity increases. The characteristics of these components result in pathozone profiles that vary in shape. Infection efficiency can decline in an exponential or sigmoidal fashion with increasing distance from the host or follow a curve that rises close to the host and then falls asymptotically to zero with increasing distance. Colony architecture was summarized by the distribution of the furthest extent of hyphal growth amongst colonies growing out from single inoculum units (described, because of its asymmetrical pattern, by a gamma distribution) and by the radial density of hyphae on concentric circles at different distances from the centre of the colony (described by a negative binomial distribution with parameters changing with distance).

The mechanistic model for the components of pathozone behaviour is tested by comparing the magnitude and shape of pathozone profiles predicted by the model with independent estimates for P obtained by placing inoculum at fixed distances from the host, and measuring the proportion of successful infections. The shape depends on the relative magnitudes and change with distance of the three components, P1, P2 and P3, which differed between the two types of inoculum. The profiles could be reproduced accurately when the probability of contact was based on the distribution and density of hyphae reaching the host rather than on the furthest extent of hyphal growth amongst colonies. The model is used to analyse the effects of replication and stochastic variability in the components of pathozone behaviour on the comparison of treatments for disease control.

This paper presents and tests a method to scale up from the dynamics of infection and disease on single plants in order to predict the behaviour of epidemics in populations of plants, in the presence of a biological control agent. Specifically, we quantify the effect of the antagonistic fungus, Trichoderma viride, Pers ex. Gray, on the pathozone dynamics of the damping-off fungus, Rhizoctonia solani Kühn on radish. The results from these individual-based experiments are used to predict the progress of an epidemic and the results are compared with experimental epidemics in microcosms. The addition of T. viride close to a germinating radish plant reduced the extent of the pathozone influence from 22·6 to 13·8 mm. Trichoderma viride also inhibited the evolution of infection efficiency of R. solani. The evolution of infection efficiency over time is described by a simple non-linear model for the probability of infection with distance, in which certain of the parameters vary with time. By combining this with a model based on conditional probabilities for the location and infectivity of randomly dispersed propagules within the pathozone, we were able to scale up and predict the disease progress of R. solani in a population of radish plants, and the extent of control effected by T. viride. Disease progress rose progressively to a maximum of 42% diseased plants in the unprotected crop compared with a sigmoidal approach to 13% in the protected crop. We show that these properties are consistent with a monomolecular function for primary infection with a temporally-varying rate parameter. We also show that the central region of the pathozone, where the joint probabilities of occurrence of a randomly located propagule and its ability to infect are maximal, has a large influence on the sensitivity of epidemics to pathozone dynamics. This is an important example of interpreting population behaviour of an epidemic from the infection and disease of single plants.