Abstract
Vertical seed dispersal, which plays a key role in plant escape and/or expansion under climate change, was recently evaluated for the first time using negative correlation between altitudes and oxygen isotope ratio of seeds. Although this method is innovative, its applicability to other plants is unknown. To explore the applicability of the method, we regressed altitudes on δ(18)O of seeds of five woody species constituting three families in temperate forests in central Japan. Because climatic factors, including temperature and precipitation that influence δ(18)O of plant materials, demonstrate intensive seasonal fluctuation in the temperate zone, we also evaluated the effect of fruiting season of each species on δ(18)O of seeds using generalized linear mixed models (GLMM). Negative correlation between altitudes and δ(18)O of seeds was found in four of five species tested. The slope of regression lines tended to be lower in late-fruiting species. The GLMM analysis revealed that altitudes and date of fruiting peak negatively affected δ(18)O of seeds. These results indicate that the estimation of vertical seed dispersal using δ(18)O of seeds can be applicable for various species, not just confined to specific taxa, by identifying the altitudes of plants that produced seeds. The results also suggest that the regression line between altitudes and δ(18)O of seeds is rather species specific and that vertical seed dispersal in late-fruiting species is estimated at a low resolution due to their small regression slopes. A future study on the identification of environmental factors and plant traits that cause a difference in δ(18)O of seeds, combined with an improvement of analysis, will lead to effective evaluation of vertical seed dispersal in various species and thereby promote our understanding about the mechanism and ecological functions of vertical seed dispersal.