Testing the potential seed availability in dung samples: comparison of two seedling emergence methods

To optimize the estimation of species composition and viable seed content of herbivore facces and to make different approaches comparable, two seedling emergence methods are evaluated. The Ter Heerdt method (TH) employs concentrated samples, potentially increasing and accelerating seedling emergence, as shown for soil samples (95 % of all seedlings emerged within 6 weeks). Samples are kept under controlled conditions (glasshouse or climate room). Secondly, a common garden method (CG) using unconcentrated samples so that seasonal changes could fulfill the germination requirements of a broad species spectrum (experiment duration approximately 15 months) was applied. The methods were tested by the use of sheep faeces samples, collected during a six-day grazing period in a threatened dry grassland (Allio-Stipetum capillatac). Both methods proved largely similar in species composition (QS 0.81) and viable seed content (QS = 0.69). More species (e.g. monocotyls) and a higher seedling emergence of hard-seeded species (Fabaceae and Cistaceae) were found in the CG method. Besides a higher emergence of some small-seeded winter annuals, few other species emerged exclusively by use of the TH method. Nevertheless, all species detected by only one method were found in low individual numbers (<= 4). Depending on research interest and availability of space and time, the most appropriate method can be chosen. If the main focus is on the species composition, unconcentrated faeces samples can be studied by CG. In case the overall viable seed content is more important and/or a shorter time period is available, TH serves as a suitable alternative. (c) 2008 Elsevier GmbH. All rights reserved.