Editor's Note:This article is part of a semi-regular series of research briefs sponsored by the Department of Viticulture and Enology at UC Davis. For more summaries of research sponsored by UC Davis Department of Viticulture and Enology, see http://wineserver.ucdavis.edu/trellissummary_categories.php.

Today I promise I will get to the seed of the problem. This is thanks to the thorough review by R. Ristic and P. Iland titled "Relationships between seed and berry development of Vitis vinifera L. cv. Shiraz: Developmental changes in seed morphology and phenolic composition" in theAustralian Journal of Grape and Wine Research, 11:43-58. 2005.

We know that during ripening seeds change from green to brown. But when and why does this change happen? Could we use seed color to tell us the changes happening in the main phenolic compounds and, in turn, when the berry is ready to be picked? This is what the authors want to know. Their approach is to perform successive sampling so they can meticulously track berry development, seed development and seed phenolic development. Then, they try to link the three. They do this in two steps. In the first part of the review, they study seed morphology changes as the berry itself changes. In the second part, they study how phenolic compounds change as the seed changes. The idea is to use the easily observable changes in seed browning to inform us of changes in phenolic accumulation.

But before we start, we need to familiarize ourselves with the terminology that would allow us to describe such a small entity as a seed. This is provided by the authors in a very nice introduction. Considering the contribution of seed components to the final red wine composition and the potential ability to use seed color as an indicator of ripeness, this review is one of the missing pieces of the puzzle.

Seed Terminology

Even though seeded grape varieties may contain four functional seeds, most often they contain fewer than four. This is because chances are one or more of the seeds won't reach full development due to endosperm degeneration somewhere along the process. If this failure happens early, the result is a seed trace; if it happens later, the result is an empty seed or a "floater." The finer morphology (shape) and anatomy (tissue types) of a grape seed are characteristics considered constant for each variety, and therefore, of important value for variety classification. Let's review some seed characteristics common to all varieties.

Morphology: A mature seed has a pear-like shape with a beak at the narrow end and a notch at the base. We also talk about a dorsal and a ventral side (back and belly). On the back there is a depression called the chalaza. The belly is divided in two by a ridge that runs from the beak to the notch--the raphe. The two resulting depressions flanking the raphe are called the fosettes. But a seed's shape is, of course, easier to see than to put into words, and I strongly encourage you to check the beautiful seed drawings in the original paper.

Anatomy: The mature seed is formed by three types of tissue: 1) a seed coat, or testa, which comprises an inner integument and an outer integument; 2) an endosperm, of fatty nature; and 3) an embryo.

The role of the endosperm is to provide food for the early embryo. The role of the seed coat is to provide mechanical protection for the embryo and to maintain seed dormancy. The well-known process of seed "hardening" is due to the lignification of the outer integument, which is accompanied by changes in color from green to brown.

Trial Design

Now that we know the different parts of the seed, we are ready to see how the seed changes during ripening. To be able to track seed changes, the authors collected Syrah berries from a Barossa Valley vineyard in Australia at regular intervals throughout berry development, starting when berries were peppercorn sized (3 to 4 mm in diameter) and finishing when berries were at 26° to 27° Brix.

The next step was to divide each berry sample into three batches: 1) One sub-sample was used to determine physical seed characteristics, such as weight, length and coat color; 2) Another sub-sample was used for chemical determination of monomer flavanols and seed tannin; and finally, 3) the remaining berries were peeled, and their skins turned into a homogenate from which the authors determined the skin anthocyanin and total phenol content by absorbance at 520 nm and 280 nm, respectively. Some of the intact berries were also used to determine juice Brix. The authors repeated this routine every week throughout the entire 1999/2000 and 2000/2001 seasons.

Changes in Seed Morphology Throughout Ripening

In an effort to use dates that would be transferable to other locations, the authors use "days after flowering," or DAF, as their time unit. As a reference, veraison normally took place around 60 DAF and harvest on 115-120 DAF. Let's now see the changes in seed shape, color and weight that the authors were able to document.

Shape: Early in seed development (20-60 DAF), seeds were very small, bright green, soft and with smooth surfaces. By veraison (around 60 DAF), they had reached full size and showed a fully developed beak. They also started to harden and develop a rough surface. One important morphological change taking place at veraison was the displacement of the chalaza from the center of the seed toward the base--the notch.

Color: The above changes are accompanied by a change in seed color, with the seed getting browner and browner. To be able to objectively refer to seed color changes, the authors developed a seed color chart comprising 12 colors--including seven shades of brown--matched to 12 numbers. The chart is reproduced here with permission from the authors. As a reference, here are the seed color changes the authors observed on Syrah (this could differ on other varieties): from bright green (color 1 of the scale) before veraison, to yellow (color 5) at veraison, then to light brown (colors 6-9) and finally to dark brown (colors 10, 11) at harvest.

Weight: The authors tracked fresh and dry seed weight separately. Both fresh weight and dry weight increased very rapidly until veraison. (At this point, if the seed had had a period of no growth, followed by another period of rapid growth, its growth pattern would have resembled that of a berry. But that's not what happened.) After the initial fast growth, the seeds started declining in size (up to a 20 percent loss in fresh weight) until they reached a plateau. Thus, maximum seed fresh weight was reached around veraison (64 DAF) and maximum dry weight a bit later (90 DAF) since dry weight did continue to increase after veraison.

In brief, the authors were able to identify three phases of seed development: 1) a phase of seed growth; 2) a transition phase, around veraison; and 3) a phase of seed drying and maturation. At this final stage of seed maturity, the embryo stopped growing while the seed coat became completely dehydrated due to the death of its cells. This programmed death is the final stage of seed maturation--a necessary "collective suicide" to provide the seed with impermeability and the ability to go dormant.

Relationship Between Seed Changes and Berry Changes

But what is the berry doing in the meantime? As we know, berry growth follows a double sigmoidal pattern (meaning two fast periods of growth separated by a lag--or slow growth--phase). The authors noticed that berry lag phase was happening around 60 DAF--that is, when the seeds were reaching maximum size. And the berries themselves were reaching maximum size around 90 DAF--that is, when the seeds were reaching maximum dry weight. After 90 DAF, many Syrah berries started to shrink. In summary, seed growth stopped much earlier than berry growth did.

In fact, when the authors studied in more detail the relationship between berry fresh weight and seed fresh weight, they got an inverse-U curve--that is, until veraison, when the relationship between berry and seed growth was linear and directly proportional: the larger the berry, the larger the seed. But post-veraison, there was an inverse relationship--the ascending part of the "U": that is, the heavier the berry, the less heavy the seed. Not only did the seed stop growing while the berry continued to grow, but the seed started to lose water as part of its natural process of ripening.

The authors found an interesting and significant difference in berry weight across seasons, with weight in season 1999/2000 20 percent greater than in season 2000/2001. Equally--if not more interesting--is the fact that the number of seeds was smaller in season 1 (average of 1.75 seeds per berry) than in season 2 (average of 2.50 seeds). Since the authors do not offer an explanation, one wonders about any differences in the rainfall or irrigation regime during both seasons. Some studies have found that the larger the seed number, the larger the berry size, something that the current authors, obviously, did not find. However, even though there were fewer seeds per berry in the first season, the individual weight of each seed--both fresh and dry--was 30 percent higher than in season 2. So, large berries did have more "seed mass" after all.

Changes in Seed Phenolics Throughout Ripening

Not to complicate things but rather as a reminder, there are four ways in which one could express the amount of tannins present in the seed of a berry. One could use amount of tannins in a given amount of seed (seed concentration); amount of tannins in one whole seed (seed level or seed content); amount of seed tannin in a given amount of berry tissue (berry seed tannin concentration); and amount of seed tannin in a whole berry (berry seed tannin level or content). To get a good understanding of seed development, the authors expressed phenolic changes in all of these ways. But for the purpose of this review, we will concentrate mostly on the third one, or berry concentrations--that is, milligrams per gram of berry weight. (By the way, berry content showed a similar pattern.)

The authors tracked the changes of two types of seed phenolic compounds: tannins and flavanol monomers.

Seed tannins: Seed tannins were at high concentration at the early stages of berry development, reaching maximum values at the beginning of veraison (6.1 mg/g at 56 DAF). After that, they declined gradually (down to 1.5 mg/g at harvest or 110-120 DAF).

Seed flavanol monomers: Flavanol monomers in seeds include catechin, epicatechin and epicatechin-gallate. Flavanol monomer accumulation showed a different pattern than tannins did. They increased sharply all the way to veraison (to a maximum of 1.6 mg/g at 64 DAF) and then they decreased sharply until they reached very low levels at harvest (0.14 and 0.27 mg/g in seasons 1 and 2, respectively).

To recapitulate, both tannins and flavanol monomers reached maximum levels at verasion, but thereafter, monomers decreased sharply whereas tannins decreased only slightly.

Relationship Between Seed Phenolics and Skin Phenolics

In case you were wondering what type of phenolic changes were taking place at the skin level, the authors looked at that, too. More specifically, they measured skin total anthocyanins and skin total phenols.

Skin total anthocyanins (mg/berry) increased from veraison on, reached a maximum around 110 DAF, and then declined. As for skin total phenol, they started at minimal levels, increased gradually from veraison on (64-73 DAF) and reached maximum levels around harvest time (120 DAF).

When the authors looked in more detail at the relationship between seed phenolics versus skin phenolics, they found a highly significant reverse relationship. That is, as seed tannins decreased, skin anthocyanins and total phenols increased. Similarly, when the authors plotted seed tannin against seed coat color values (based on their scale and averaging ventral and dorsal colors whenever they differed), they also found a strong inverse relationship. This meant that as extractable seed tannins decreased, seeds were darker. In other words, all of these changes--color increase, seed tannin decrease, skin tannin increase--occurred simultaneously and were part of the process we call "berry ripening."

Conclusion

This study describes the changes in seed size, seed external appearance and seed coat color involved in the complex process of seed development. The authors believe that this is the first time these developmental changes have been fully described. This study also showed us which stages of seed development were synchronized with which stages of berry development as well as with which changes in phenolic composition. For instance, one of the main findings was the strong correlation between seed coat color and the concentration of extractable tannins. This important result confirms that, yes, seed color can be used as an indicator of when berries are ready--physiologically ripe--to be harvested. wbm