The Use of Seed Morphology to Classify and Identify Subspecies

The shape and texture of seed may provide one way to classify the many variations evident in Portulaca species in Australia.

A botanist friend has sent me a link to an interesting paper that was published in the Israel Journal of Botany, Volume 27, 1978, pp. 177-211. The paper is called "Cytogeography and Taxonomy, of the Portulaca oleracea L. Polyploid Complex" by Avinoam Danin, Irene Baker, and Herbert G. Baker. It is a bit on the technical side but the findings and photographs are very useful. Their methodology could be applied to the Australian situation.

The authors studied populations of P. oleracea throughout the world and documented the variations with detailed electron microscope scans. By grouping consistent characters, they described nine subspecies of P. oleracea in the world. Seven of these they described for the first time.

However the authors unfortunately excluded Australia from their study. They perhaps realized that the P. oleracea complex in Australia is a special case that demands an independent study. For instance, the relationship between P. intraterranea and P. oleracea needs clarification, so that the full range of characters between the two species can be fully understood and documented. It will be interesting if future taxonomists seek to classify all of the variants in P. oleracea in Australia or simply lump them together under one species name.

If any of the subspecies that are described in the Israeli paper do occur in Australia, it should be possible to match them using the key and descriptions provided.

On page 181, the paper also has an illustration that shows the technical names for the various parts of a Portulaca seed.

For those who wish to download this paper it is available here as a free download from Flora of Israel Online. It is a fairly long paper so may take that extra minute to download. But it is definitely a good read for those who wish to understand the complexities associated with variation in a species that has a very wide-ranging distribution.

My only reservation with the paper is that the results would need to be duplicated over a period of some years to test if the characters remain fixed. I am not sure if this process of duplication has occurred. It is apparent that since 1978 very few (if any) herbaria have adopted the circumscription proposed by Danin et. al.. I am not aware of any herbaria that recognize the subspecies names today. The lack of interest may have been due to conservatism relating to convention, or to difficulties faced whenever seed is absent from herbarium specimens. The need for cumbersome electron microscopes is also of limited use to those who require an immediate identification in the field. Also, when we consider that P. oleracea is generally considered a "weed", many people may have felt that it did not merit any attention beyond the level of species classification.

Whatever the case, I wonder if the seed characters that originally defined the subspecies are reliably fixed. Can the results be duplicated today? Are the characters predictable within an entire population of a subspecies, even when it occurs in different habitats, soils, or climates? Can the results be replicated in cultivation?

Consistency and the ability to reliably duplicate results is what science is all about. Any key that depends on the characteristics of a single plant part, in this case seed, to delimit the variations found in a morphologically variable species, may lose credibility if ever the distinctions become blurred. For instance, the key may be confounded if people find seed that shows mixed characters, or if the characters do not remain stable in a range of habitats year after year. Other problems may occur if identical or similar seed is found on two different plants that have other morphologically distinct characters. In this case, are the two plants really the same subspecies?

I would be interested to hear from anyone who has used the 1978 keys and descriptions and found them to be reliable.