Changes in density, biomass, seed production and soil seed banks of the non-native invasive plant, Chromolaena odorata, along a 15 year chronosequence

The non-native invasive plant Chromolaena odorata (Asteraceae) was studied at 6 sites, with a chronosequence of ages from <1 to greater than or equal to 15 years, at St Lucia, South Africa. C. odorata density, biomass, seed p roduction and soil seed banks were quantified in three microsites: sun, sem i-shade and shade. C. odorata density decreased with invasion age, apparent ly as a self-thinning process. Biomass per unit area and seed production/pl ant increased over the first 10 years, but declined greatly at greater than or equal to 15 years. C. odorata plants grew larger and had much greater s eed production in the sun relative to semi-shade, with small plants produci ng few if any seeds in the shade. Seed production in the sun varied from 20 00 (<1-year old site) to 260 000 (10 year) seeds m(-2) annum(-1). About 20- 46% of seeds produced were germinable and showed the same trend with age of invasion, but was particularly low after greater than or equal to 15 years . Assessment of soil seed banks immediately prior to seed production (seed greater than or equal to 10 months old), indicates that about 5-10% of seed s in the sun and 11-22% in the shade were still germinable, resulting in ge rminable seed densities of 12-385 and 158-511 m(-2), respectively (excludin g the greater than or equal to 15-year old site). A greenhouse trial showed that burial of seeds, relative to those at the surface, and provision of l ess water, significantly improved seed persistence in the soil, while light intensity had no effect. Control of C. odorata is difficult due to rapid a ttainment of reproductive maturity, large production of wind-dispersed seed s and a short-term persistent seed bank. An integrated control strategy eit her excluding fire (coastal forest sites) or using fire prior to seed relea se in July/August to kill plants and soil-stored seeds immediately prior to seed production, together with biological, chemical and/or physical contro l, should be explored.