Influence of population structure on self-thinning of plant populations

SUMMARY biomass and density but with different population structures (frequency distributions of plant weight), and to ask whether they followed different paths. (2) In experiment 1 we created three population structures by using different combinations of seeds of different sizes. The differences created were small. In the remaining experiments mixed-aged stands were created by sowing a second cohort between 4 and 23 days after a first cohort. (3) In experiments 3 and 4 we found that more light penetrated under the canopy at the centre of the pot in pots at the edge of experimental blocks than in pots at the centre, but this only produced better survivorship of small plants in mixed-age stands. In subsequent experiments no edge pots were used. (4) The paths followed by control even-aged stands in the different experiments varied substantially, for unexplained reasons. It may be best to think of the log biomass-log density boundary along which self-thinning stands travel as a thinning band rather than a thinning line. (5) While in some experiments mixed-aged stands appeared to follow different paths from the control even-aged stand, the nature of the difference was not consistent. Further, the difference within any experiment was small compared with the differences among controls in the different experiments. We conclude that population structure is not an important influence on self-thinning paths relative to unexplained variation among stands. (6) The mean slopes of the paths followed by the controls were less than the expected -1/2. The overall pattern of the results was consistent with the idea that paths are bounded by a line of slope -1/2 only up to some maximum biomass, after which density may continue to decline but biomass per unit area does not increase. (7) During thinning, stands with different initial structures converged towards a frequency distribution of log weights which was roughly symmetrical and about 2-3 orders of magnitude broad.