, 2013). To answer the question of whether the germination rate is influenced negatively CT99021 mw by flooding, the germination rate of samaras after storage in water was tested. For this experiment, samaras of different F. pennsylvanica trees from a stand situated in a floodplain forest along the River Elbe near Dessau in Sachsen-Anhalt were again used. The samaras were collected in autumn 2006 and stored dry at 5–8 °C until spring 2008. Only full seeds were used in the test. The average dimensions of the samaras (mean ± standard deviation) were 45.1 ± 5.5 mm

(length) and 5.7 ± 0.9 mm (width). The weights varied between 17 and 92 mg, with a mean of 49.3 ± 11.7 mg (N = 600). The germination rate of F. pennsylvanica was tested after 0 (control), 2, 10 and 15 days of storage in water. Every variation of the treatment was tested using three replications with

50 samaras ( Baskin and Baskin, 2001). The duration of storage in water was similar to the mean flood times (depending on the altitude) during the vegetation period in the floodplain forest investigated ( Klausnitzer and Schmidt, 2002). For the purposes of storing the samaras in water a basin was used for Selleck SB431542 each treatment (diameter = 29 cm; depth = 9 cm), filled with distilled water and kept at room temperature. Distilled water was used to allow for comparability with other studies. The samaras were placed on the surface of the water, sinking over the course of the experiment. The subsequent germination test followed the ISTA (International Seed Testing Association) guidelines (ISTA, 2005) for ash, namely in a germination box on moist paper and in a germination cabinet with alternating temperatures, 16 h at 20 °C in darkness and 8 h at 30 °C in light. The same temperatures were used successfully for germination tests on F. pennsylvanica by Steinbauer (1937) and Bonner (1974). The germination test was terminated

after 20 days following the recommendation made by Baskin and Baskin (2001, p. 19). The data were analysed using Origin 8G and SPSS 11.5. Given the sample size N = 12, the critical value D0.05 was used for verification Rolziracetam (with N = 12: 0.375; Sachs, 1997). Significant differences were considered at the P < 0.05 level. A non-linear regression analysis was performed in order to predict the number of germinated seeds as a function of the duration of storage in water. In order to address data obtained for the different variants over time, different fitting models were compared using the χ2 minimisation fitting routine in Origin 8G. The fitting was based on 200 iterations. The Boltzmann fit was selected as the best fitting model on the basis of an evaluation of the goodness-of-fit criteria (R2 and χ2/df values).