Characteristics of soil seed bank after 32 years of different harvesting methods in evergreen broadleaf forests postprint

Abstract: Soil seed bank is both the genetic memory and the future in vegetation succession, and plays an important role in revealing the mechanism in the recovery process of disturbed forests and in predicting its successional direction. In order to reveal the role of soil seed bank in the process of forest restoration, and to provide scientific basis and theoretical guidance for the management of subtropical secondary forests and the development of artificial restoration techniques, the soil seed banks from secondary forests which undertook different modes of harvesting and have recovered ever since, were analyzed. The comparisons were conducted between forests naturally restored for 32 years after clear cutting (CC) or selective cutting (SC) in Tianjingshan Forest Farm, Guangdong, with noncutting (NC) evergreen broad-leaved forests in the same area as controls. The distribution, density, species composition and diversity of the soil seed bank in these three types of forest were investigated through vegetation surveys and soil seed germination experiments, and the relationship between the soil seed bank and the above-ground vegetation was analyzed. The results were as follows: (1) The seed banks of understory soils of different harvesting methods were dominated by herbs, followed by shrubs and least by trees. The seed densities of trees and shrubs were in the order of NC>CC>SC, with NC significantly higher than CC and SC (P<0.05), and there was no significant difference between CC and SC; the seed densities of herbs were not significantly different among different harvesting methods. (2) A total of 43 plant species were found in the soil seed banks, most of which were light-damanded species, including 7 tree species (which were NC=CC>SC), 15 shrubs (which were NC>CC>SC), and 21 herbs (which were CC>NC>SC). (3) Shannon’s diversity index, Simpson’s diversity index and Pielou’s evenness index of trees in the seed banks were in the order of NC>CC>SC among different harvesting methods, the diversity indices and evenness indices of shrubs and herbs were mostly not significantly different among different harvesting methods. (4) The similarity between soil seed banks and vegetation was low in all investigated forests, with 4 shared tree species including 3 in NC and 1 in SC, 2 shared shrub species, one each in SC and CC, and 1 shared herbaceous species in NC. In conclusion, the seed density, number of species and diversity indices in the soil seed bank after 32 years of natural recovery from logging in broadleaved evergreen forests are still smaller than those in unharvested forests, the soil seed bank cannot support rapid natural recovery of disturbed forests.