Topsoil seed banks in grazed and ungrazed eucalypt woodlands at Newholme, Armidale, New South Wales, Australia

Abstract

Germinable soil seed stores were estimated in grazed and ungrazed woodlands at Newholme (Armidale, New South Wales) to establish whether there is a store in the topsoil of shrub species that are no longer present in the extant vegetation. Soils were subjected to combinations of smoke and heat treatments before the germination procedure was commenced. The mean germinable topsoil seed reserve to a depth of 10 cm was 10 855 seeds m‐2 Of this, 72% of seed was found in the upper 5 cm of soil. Two rush species (Juncus filicaulis and J. homalocaulis, 3185 and 1197 seeds m‐2, respectively) and the perennial herb Gnaphalium gymnocepahalum (1728 seeds m‐2) were the most abundant species in the topsoil. Almost half (46%) of the topsoil seed reserve was composed of rushes or sedges, while perennial herbs composed a further 30%. Shrub species were almost absent from the topsoil seed store. Soils collected from grazed plots (12 360 ± 2673 seeds m‐2) did not have significantly different densities of topsoil seed to ungrazed plots (9351 + 2588 seeds m‐2). However, four species showed a significantly greater seed store in the grazed than the ungrazed plots (Asperula conferta, Eragrostis trachycarpa, Plantago lanceolata, and Sporobolus creber), while Elymus scaber (a native grass) exhibited a significantly higher topsoil seed store in the ungrazed plots. Of the 69 species identified, 4 species showed significantly higher germination in smoked treatments (Centaurium tenuiflorum, Crassula sieberiana, Hy‐pericum gramineum, and Plantago lanceolata), while 10 species exhibited significantly figher germination in unsmoked treatments. Five species exhibited significantly higher germination in heated trays (Desmodium varians, Dichondra repens, Geranium potentilloides, Trifolium campestre, and Trifolium repens), while 49 species showed significantly higher levels of germination in unheated trays. Species responding to the smoke treatment were native and non‐native annuals, while those responding to the heat treatment were mainly perennial legume species. Those species that showed increased germination with smoking were killed by the heat treatment, while those species that required heat to germinate showed no response to smoke. Comparison of the topsoil seed reserve and the vegetation present in these areas using Detrended Correspondence Analysis (DCA) showed that there was a large difference in both composition and abundance between the topsoil and the extant vegetation. Only two shrub species were relatively abundant (>10% frequency) in the extant vegetation (Bursaria spinosa and Lissanthe strigosa). The implications of these findings to the management of woodlands at Newholme are discussed.