Genetic, edaphic and procedural factors determining the inherent pasting properties of wheat starch

Abstract

A Minipaster apparatus, based on a Haake rotary viscometer but with a small sensor head resembling that of the Brabender Amylograph, was developed to investigate thermal pasting of starch water suspensions with starch weights as low as 1.5 g. A channelysing computer interfaced to a Coulter Counter was developed to examine size distributions of starch granules. Effects of starch preparation, granule damage, drying, ageing, acidity, disaggregation, and rate of heating were investigated and standardised for the major investigation.

Starch from 59 wheat cultivars, chosen as likely to be of importance in future wheat breeding in New Zealand, exhibited peak viscosities from 82-269 units. Each starch was prepared without damage and without drying, pasted in the absence of amylase, at 10% concentration with pH adjusted to 5.5 before pasting.

For a separate set of wheat samples, analyses of variance to determine the relative contributions of cultivar, year and location to variation in wheat starch pasting properties showed that genetic control was most important by an order of significance and that year and location affected variation about equally. Differences due to year and location could be ascribed to edaphic stresses on the growing wheat crop. Neither amylose content, lipid content, nor granule size -distribution of the starch determined the pasting differences found.

Two cultivars were selected whose starches had contrasted pasting properties. When fractionated on the basis of granule size, the original pasting differences were observed mainly in the large A-granule fractions, weight-average diameter above 23 µm, while few differences were seen among the B-granules of each starch, diameter <10 µm. Examination of immature granules confirmed that differences between starches were established during the period of A-granule growth, that occurs within about 15 days of fertilisation, rather than during the later period of grain filling and desiccation.

The traditional pasting viscosity parameters, peak viscosity, breakdown -viscosity after 20 min holding, and setback viscosity on cooling to 60°C, were found to be serially dependent as can be demonstrated for the Arnylograph. Together with peak viscosity temperature these four variables were found to be adequate descriptors of starch pasting variability. Their ability to discriminate between starches depended markedly on starch concentration. At 6% concentration granule swelling phenomena were distinguished, but at 14% properties due to composition were more obvious. Neither the coefficients for the ln-ln regression of viscosity on starch concentration, nor the ratios of the pasting viscosity parameters, were useful in discriminating between starches. The adjustment of pasting parameters to a fixed value for breakdown viscosity was also useful, as were the pasting factors, defined as the multiplying factors for viscosity increase caused by doubling concentration.

Thermal granule swelling proceeded without discontinuity until all available suspension volume was occupied with swollen granules, and was not noticeably affected by either gelatinisation or exudation up to a concentration-dependent temperature. Pregelatinisation thermal swelling was not reversible from 40°C or 50°C. At 10% concentration, swelling and exudation had by 80°C proceeded to differing degrees for different starches to an extent sufficient to differentiate starches in a similar order to that determined by later pasting processes. An interpretation based on the distribution of order within the starch granule is proposed.

Sufficient variation in inherent pasting characteristics was available through choice of cultivar to urge the inclusion of this variable in breeding selection for feedstock to industrial starch manufacture. Of the procedural factors examined, only the decrease in pasting strength with higher temperature drying appears to be of industrial significance and requires optimisation.