Variation in seed production and germinability in common reed (Phragmites australis) in Britain and France with respect to climate
Article first published online: 28 APR 2006
Volume 133, Issue 2, pages 233–243, June 1996
How to Cite
McKEE, J. and RICHARDS, A. J. (1996), Variation in seed production and germinability in common reed (Phragmites australis) in Britain and France with respect to climate. New Phytologist, 133: 233–243. doi: 10.1111/j.1469-8137.1996.tb01890.x
- Issue published online: 28 APR 2006
- Article first published online: 28 APR 2006
- (Received 6 November 1995; accepted 9 January 1996)
- Phragmites australis;
- seed set;
- seed weight;
- seed germination;
Northern populations of Phragmites australis (Cav.) Trin. ex Steud. (common reed) often flower so late that the ability to produce viable seeds before winter dieback can be limited. If climate changes as predicted, the reed may become able to set seed further north in its range. This would have implications for the ability of reed to colonize new sites as well as for the distribution of animals which eat reed seeds in winter.
A survey of 34 British populations in 1993 and 1994 showed that seed production is highly variable between sites, ranging from 0 to 100%. Southern and western populations tended to show the highest seed sets, although some southern populations were infertile and some northern sites set seed. Average seed weight tended to be higher in sites which showed good seed set. Seeds from sites with high average seed weight tended to germinate more readily.
The seed-setting capability of populations was successfully associated with climatic conditions at the time of flowering. However, seed weight was not found to relate to climate, but depended on seed production and overall plant height. Logistic regression modelling indicated that the best seed sets should occur in P. australis when: (i) August rainfall is low; (ii) the combined rainfalls of September and October are high; and (iii) the combined temperatures of these months are high. The model accounts for approximately one third of the variation in observed seed sets between sites, indicating that other, unmeasured, factors might also play a part in determining seed set in P. australis.
At three northern populations which were studied in more detail, a number of genotypes coexisted, and legitimate pollen germination and pollen-tube growth was observed in each case. As no seed set occurred in two of these, self-incompatibility cannot explain poor seed setting.