Epsilon Archive for Student Projects

Abstract

Masting is a synchronous and highly variable production of seeds in tree species. Masting may
reflect a reproduction strategy to cope with seed predation. In beech, temperature is the cue
controlling annual variability in seeding.
Masting behaviour in beech typically begins around the age of 40 years. This is however based on
observations and therefore lacking a formal statistical approach. It is unknown if beech trees younger
than 40 years exhibit masting signals in their ring width chronologies. By considering potential
masting signals in ring width chronologies, possible underused data can be included to better
understand the relationship between environmental variability and seed production. Ring widths can
furthermore act as a proxy for the dominance of a tree. More dominant trees may express less
pronounced and later occuring masting signals due to their greater accessibility to resources.
This study aimed to 1. Assess the age of appearance of masting signals in tree-ring chronologies of
beech in southern Sweden 2. Evaluate the effect of social position on (the onset of) masting signals.
Ring widths were used as a proxy for masting signal and dominance in beech trees. A masting record
from a previous study in the same area was used to measure the effect of identified masting years
on ring widths, with smaller ring widths indicating a stronger effect of masting. The effect of age
and dominance on the expression of masting signal in non-seed producing trees was evaluated by
considering the effect of environmental conditions during identified masting years on radial growth.
Mixed effect models were developed in Rstudio with a nonlinear model being most suited for this
research to capture complexity between variables.
The nonlinear mixed effect model ultimately showed that the onset of masting signal in beech tree
ring chronologies in southern Sweden occurs when the cambial age of tree at the breast height (the
height at which the growth data were collected) is above 20 years. Samples in this study were
obtained at 1.3 m height and thus lack information on the tree growth starting at ground level.