Assessing populations of tree seedlings in multi-species natural forests

Compared with sampling studies aimed at trees of commercial size, regeneration sampling has received little attention. In recent years, regeneration sampling has become relevant as extensive forest regeneration areas are developing in the wake of the devastations caused by bark beetle attacks in the Northern Hemisphere. Estimating a population of interest is especially challenging in plant communities where the measurement of individual plants is technically complicated, as in the case of seedlings and saplings in multi-species communities. Two methods of regeneration sampling are presented in this study. One involves a 10 m(2) circular plot, the other is based on the distance between neighbouring seedlings. Completely enumerated, mapped and sufficiently large populations of seedlings are rarely, if ever, available. It was necessary therefore, to simulate a range of populations with different combinations of seedling density, species richness and height distribution, for different spatial patterns. The accuracy of each sampling method is evaluated for increasing sampling intensity in terms of species richness, seedling density and seedling height distribution. Our results show that Staupendahl's circular plot method provides unbiased density estimation and is therefore preferred. Seven samples per ha provide estimates of seedling density, mean height and species richness with a relative deviation of less than 20 percent. One seedling may be sufficient in boreal forests while two or more seedlings should be assessed at each sample point when the number of seedling species is high. Our results indicate significant differences between the methods in assessing species richness, and the regeneration neighborhood unit method is recommended. The study presents a systematic approach for evaluating different sampling methods when field data are not available.