Summary

This is the first Australian study investigating genetic variation in wheat spike architectural traits and their relationship to grain yield using X-ray Computed Tomography. The project explored both historical varieties and novel genetic populations. Lines from a Mace/MW#293 recombinant inbred lines (RIL) population with high grain number and high average grain weight were identified that may be suitable for future breeding. X-ray CT methods need further investigation as they did not accurately predict grain weight.

Background

Australian wheat breeding has led to major improvements in disease resistance, harvest index and adaptation, but genetic improvements in yield are slowing. If genes can be identified that are linked to certain features of spike architecture, like number of grains or spikelets per head, grain size, spike length or compactness, then breeders can use DNA markers to speed up development of new varieties with higher yield potential. X-ray Computed Tomography (X-ray CT) is a fast and efficient way to measure traits that affect crop performance without having to cut plants open. This project focused on grains per spike and average grain weight.

Research Aims

The core objective of the project was to determine if wheat yield could be improved by manipulating spike architectural traits in Mace/MW#293 RIL population, specifically increased spikelet number per spike and grain number per spikelet.

In The Field

A controlled environment ‘drought spotter’ experiment was conducted at the Australian Plant Phenomics Network (APPN) facility to examine the effects of drought on spike architectural traits in a historical wheat set spanning six decades of breeding (Heron-1958 to Kingston-2022). Drought stress was imposed at Zadoks growth stage 31.

A field trial run at Roseworthy in 2023 had some emergence issues, potentially due to herbicide residues. However, with support from Analytics for the Australian Grains Industry (AAGI), the team was able to analyse physical characteristics and genetic makeup of 189 of the original 396 lines within the Mace/MW#293 RIL population.

Results

In the controlled environment experiment, a moderate effect of drought stress on spikelet number was recorded with up to a 20 per cent reduction compared to well-watered treatment. This was less than expected and in future studies drought stress should be imposed at Zadoks 30 to see a greater impact of drought on spikelet number.

In the historical wheat varieties, plant grain weights in some varieties were two times higher than others whether the crops were drought stressed or well-watered. Vixen and Warigal produced the lowest and highest grain weights under both drought and well-watered conditions respectively. Drought tolerance as measured by relative grain weight ranged from 25 per cent in Calibre to 44 per cent in Spear.

In the field trial, unseasonal late rains prior to harvest in December and significant pre-harvest sprouting meant grain yield could not be recorded. Therefore, links between spike architectural traits and grain yield could not be determined.

Significant differences were recorded for all traits analysed in the Mace/MW#293 RIL population and the historical wheat set. Lines were identified in the Mace/MW#293 RIL population that had high grain number and high average grain weight. The team were able to identify at least one part of the wheat genome linked to each spike architectural trait.

X-ray CT was unable to count spikelet number per spike or grain number per spikelet or consistently predict grain weight across lines. However, X-ray CT predicted grain number accurately which was then used in conjunction with manual grain weights to predict average grain weight per spike. Spikelet number per spike and grain number per spike were counted manually with additional funding from SARDI.