A matter of size

by Emily Wilson
Share This:

Within any climatic region in which cereals are grown, there are varieties of the same species that have larger grains than others. Sometimes the genetic differences are expressed through identifiable characteristics, such as endosperm texture, with soft varieties generally having larger grains than hard. In other instances, the reasons for differences in grain size among varieties are not as easy to define, and grain size itself seems to be directly genetically determined.

In addition to varietal differences, variations occur as a result of growing conditions. Other factors influence grain size by affecting the relationships among grains in the same crop; these are physiological factors and they apply irrespective of growing conditions. Although they are applicable to all cereals to some degree, we are considering them here in relation to wheat, and most experience lies in the common wheat Triticum aestivum species.


Grain size varies along the length of both the spike and spikelet in a systematic fashion so that the largest grains occur just below the center on the spike and, within a spikelet, the largest grain is found on the next to the lowest position. A diagram published by Peter Bremner and Howard Rawson of CSIRO Australia (Figure 2), shows the relationship among grain weights on a typical spike.

In some cases the ears on the last-formed tillers do not fully develop before ripening begins. There is thus variation in grain size on ears of the same plant. In general, the earlier the ear is formed, the larger are the grains it bears. Another adaptive mechanism is the redirection of nutrients to the remaining grains if some are damaged before they mature.


In an experiment at Campden and Chorleywood Food Research Association (CCFRA) my colleagues and I found that the amount of flour that can be milled from wheat does not seem to depend upon the size of the grains.

It would perhaps be expected that large grains would yield more flour. The justification for this notion is that for bodies of a consistent shape, the surface to volume ratio declines as the body size increases, and tissues occurring only at the surface would thus be expected to make a reduced contribution in larger grains than in smaller ones.

CCFRA investigated this by carrying out dissections of different sized grains, and we did find that shriveled grains had a greater proportion of fruit coats and seed coats at the expense of endosperm. However, when we compared well-filled grains of different sizes, the endosperm content was very consistent.

On the basis of visual observations, we concluded that this might be because larger grains have proportionally larger endosperm cavities. This proved significant in another aspect of grain physiology.

When grains begin to germinate, a number of hydrolytic enzymes are generated. They serve to solubilize nutrients stored in the endosperm, making them available to the growing embryo. The most important of these, from a processor’s point of view, is alpha-amylase, an enzyme involved in breaking down starch.

A variant of this enzyme is also produced towards the end of the maturation period of all wheat grains, but there is a difference between the location of the aleurone tissue in which the two amylases arise. The germination enzyme is synthesized in the aleurone tissue surrounding the grain; the type produced in late maturity type arises in the aleurone tissue in the crease region, around the endosperm cavity.

We found that varieties characterized by high levels of the late maturity amylase were those that also had large grains, possibly reflecting the fact that larger cavities are present, with more aleurone tissue in that region producing more enzyme.


The aleurone layer surrounding most of the grain is pressed hard against the seed coats and it becomes stuck to these layers so that, in milling, it is more likely to end up in the bran than in the flour. The aleurone in the crease however is attached only to the endosperm within it and, in consequence, it is likely to be milled into the flour fraction rather than the bran.

It is therefore not surprising that Dr. Samuel Millar and Dr. Martin Whit-worth of CCFRA found that flours milled from larger grained varieties contained a higher proportion of pentosans in the flour, as pentosans are found in cell walls, particularly the thick cell walls of the aleurone layer. Pentosans absorb large amounts of water, so again it was not surprising that these authors found that flours milled from large grained varieties had a higher water absorbing capacity than those from smaller grained varieties.

In spite of this they found that this factor was not sufficient alone to account for all the difference in water absorption. They learned from Dr. Peter Gras of CSIRO in Australia that the starches from some large-grained wheat also absorbed more water than those from small-grained types.

Dr. Gras proposed that the difference might be due to the sizes of the starch granules present. He reasoned that if more small granules were present, the absorption would be higher because the surface to volume ratio is greater in small granules. It had already been shown in an earlier study that large grains do indeed have relatively more small starch granules than small grains but Millar saw the need to confirm this in the case of the varieties he was studying. He used both image analysis and laser diffraction techniques to measure the particle size characteristics of carefully prepared starches, and he confirmed significant differences related to grain size.

From the above, we can conclude that there are many compositional differences related to grain size, and more strongly to the grain size associated with different wheat varieties. But there are also processing differences. At the U.K. Satake Centre for Cereals Processing, Dr. Grant Campbell has been examining the effect of grain size on roller milling performance, finding some interesting results.

As discussed at the beginning of this article, grain size in a sample of wheat is quite variable, but the process parameters have to be set without reference to this variation. It would of course be very difficult to suit, for example, the roll gap on first break to grain size because wheats are not graded in size fractions before milling. Thus the roll gap does not relate to individual grain sizes.

It may be thought that with such an energetic process as roller milling, the characteristics of the milled stocks would be determined only by the roll gap, with grain size being insignificant. Not so. Campbell found that when size-graded wheat samples were ground through first break rolls using a consistent roll gap, the influence of grain size was clearly detectable.

As the family of curves in Figure 4 shows, it was the largest grains that were reduced to fragments of smallest size, with a consistent increase in particle size of ground material as smaller grains were ground. When it comes to cereal grains, size does matter.