SYMP 12-2
Decadal changes in masting behavior of oak trees: A resource allocation shift caused by climate change as a possible mechanism

Wednesday, August 12, 2015: 8:30 AM
309, Baltimore Convention Center
Mitsue Shibata, Tohoku Branch, Forestry & Forest Products Research Institute, Morioka, Japan
Background/Question/Methods

It has been shown that there is considerable intraspecific variation in seed masting patterns, not just between regions but also over time. Environmental variations and/or climate differences (e.g., global warming) have been focused on as possible causes of spatio-temporal variations in masting. However, the mechanism controlling masting and variations in masting have not been studied extensively to date, and there are few examples of quantitative evaluation of masting patterns. Therefore, we performed a time-series analysis of seed production of Japanese oak (Quercus crispula) in Japan over 34 years (1980–2013) using a log-linear autoregressive (AR) model in order to detect temporal changes in masting patterns. We then generated a time series using a resource budget model and compared AR features between the empirical and simulated time series in order to determine whether a resource allocation shift caused temporal changes in masting patterns.

Results/Conclusions

Two coefficients in the AR model of the time-series analysis, a1 (the effect of seed count with a 1-year lag) and a2 (the effect of seed count with a 2-year lag), were used to describe the features of masting patterns. The observed oak trees showed a 3-year cycle of masting during the earlier period in this study, whereas the length of the cycle shortened to 2 years in the later period. This temporal change in masting patterns was mainly attributed to a change in a2, which showed a concave curve for the study period and regressed in accordance with a quadratic model with atmospheric temperature. The simulated time series showed a similar shift in masting pattern with productivity rate; a2 showed a concave curve against productivity rate. Because the reproductive rate may depend on temperature, we believe that the observed shift in masting cycle of Japanese oak may be caused by medium- and long-term increases in atmospheric temperature through the mediation of the dynamics of net production per year. In this study, the combined analyses of empirical data and a theoretical model provided clear insight into the mechanism of the masting cycle.