Resource-related variables drive individual variation in flowering phenology and mediate population-level flowering responses to climate in an asynchronously reproducing palm
Data files
Mar 26, 2020 version files 317.26 KB
Abstract
Many tropical plant species show wide intra-population variation
in reproductive timing, resulting in the protracted presence of flowering
and fruiting individuals. Various eco-evolutionary drivers have been
proposed as ultimate causes for asynchronous phenology, yet little is known
about the proximate factors that control reproductive onset among
individuals, or that influence the proportion of trees producing new
inflorescences within a population. We employed a nine-year phenological
record from 178 individuals of the hyperdominant, asynchronously flowering
canopy palm, Oenocarpus bataua (Arecaceae)¸ to assess whether
resource-related variables influence individual- and population-level
flowering phenology. Among individuals, access to sunlight increased rates
of inflorescence production, while the presence of resource sinks related
to current investment in reproduction –developing infructescences– reduced
the probability of producing new inflorescences. At the population level,
climate anomalies induced by El Niño Southern Oscillation (ENSO) affected
the proportion of the population producing inflorescences through time.
Moreover, the effects of ENSO anomalies on flowering patterns depended on
the prevalence of developing infructescences in the population, with
stronger effects in periods of low developing-infructescence frequency.
Taken together, these results suggest that resource-related variables can
drive phenological differences among individuals and mediate
population-level responses to larger-scale variables, such as climate
anomalies. Consequently, a greater focus on the role of resource levels as
endogenous cues for reproduction might help explain the frequent aseasonal
phenological patterns observed among tropical plants, particularly those
showing high intra-population asynchrony.
Methods
See 'Methods' section in main text.
Usage notes
Population-level phenological time series have missing values for the months of December 2008, October and November 2012, December 2014, and April 2015 (See 'Methods - Population-level Analyses' on the main text).