Modelling fungal sink competitiveness with grains for assimilates in wheat infected by a biotrophic pathogen

Abstract

Background and aims: Experiments have shown that biotrophic fungi divert assimilates for their growth. However, no attempt has been made either to account for this additional sink or to predict to what extent it competes with both grain filling and plant reserve metabolism for carbon. Fungal sink competitiveness with grains was quantified by a mixed experimental-modelling approach based on winter wheat infected by Puccinia triticina.

Methods: One week after anthesis, plants grown under controlled conditions were inoculated with varying loads. Sporulation was recorded while plants underwent varying degrees of shading, ensuring a range of both fungal sink and host source levels. Inoculation load significantly increased both sporulating area and rate. Shading significantly affected net assimilation, reserve mobilization and sporulating area, but not grain filling or sporulation rates. An existing carbon partitioning (source-sink) model for wheat during the grain filling period was then enhanced, in which two parameters characterize every sink: carriage capacity and substrate affinity. Fungal sink competitiveness with host sources and sinks was modelled by representing spore production as another sink in diseased wheat during grain filling.

Key results: Data from the experiment were fitted to the model to provide the fungal sink parameters. Fungal carriage capacity was 0·56 ± 0·01 µg dry matter °Cd(-1) per lesion, much less than grain filling capacity, even in highly infected plants; however, fungal sporulation had a competitive priority for assimilates over grain filling. Simulation with virtual crops accounted for the importance of the relative contribution of photosynthesis loss, anticipated reserve depletion and spore production when light level and disease severity vary. The grain filling rate was less reduced than photosynthesis; however, over the long term, yield loss could double because the earlier reserve depletion observed here would shorten the duration of grain filling.

Conclusions: Source-sink modelling holds the promise of accounting for plant-pathogen interactions over time under fluctuating climatic/lighting conditions in a robust way.

Fig. 1.

Following Bancal and Soltani (2002), the enhanced carbon repartition model that now accounts for fungal sink competitiveness. Sources (metabolic activities that release free carbon) are drawn as triangles and arrows, while sinks (metabolic activities that consume free carbon) are drawn as ellipses and arrows. The carbon compartments are drawn as rectangles with plain lines, and the plant/fungus organs with dashed lines.

Fig. 2.

The effects of inoculation and shading levels on rates of net assimilation U_a (A), grain filling U_g (B) and rust sporulation U_s (C). Inoculation levels (I-1 to I-4) indicate either no inoculation (I-1) or inoculation of one-third, two-thirds or all of the flag leaf area (treatments I-2, I-3, I-4 respectively). Shading treatments refers to 0, 35, 50 and 65 % shading of flag leaf (S-1, S-2, S-3, S-4, respectively). Box plots represent the range between the first and third quartile. Medians are shown as bold lines and minimum and maximum recorded values as circles.

Fig. 3.

Relationship between mean lesion size and lesion number in flag leaves (A) or lesion density in the inoculated area (B). Each point represents one single plant, and symbols differentiate shading treatments, as indicated in the key.

Fig. 4.

Following hypothesis H1, evaluation of the model represented in Fig. 1 based on observed versus simulated values for rates of grain filling U_g (A), reserve balance U_p–U_h (B) and spore production U_s (C). Each point represents a single plant.

Fig. 5.

Simulated carbon fluxes in a wheat crop infected or not by sporulating leaf rust, according to incoming PAR. The crop is built with 550 culms m⁻², each bearing 40 grains. Infected plants also bear 1000 rust lesions equally distributed on the different leaves. Their net assimilation rate (U_a) is reduced to 88 % of the control plant (A). The respective rates of (B) sporulation (U_s), (C) grain filling (U_g,) and (D) reserve balance (U_p–U_h) are then calculated in control (bold line) and in infected plants according to hypotheses H1 (thin line) and H2 (dotted line). The H1 and H2 lines generally overlap, indicating that the results do not depend on the hypothesis.