Journal article
Estimating tree phenology from high frequency tree movement data
Agricultural and forest meteorology, Vol.263, pp.217-224
12/15/2018
DOI: 10.1016/j.agrformet.2018.08.020
Abstract
•Accelerometer data matched the expected seasonal change in tree mass.•Estimates from accelerometers correlated with ground observations of leaf emergence.•Sample frequency had a substantial impact on detecting a phenology signal.
Shifts in forest tree phenology are one of the most important and conspicuous plant responses to climate variability. However, systematically documenting changes in phenology of individual trees across large areas at high temporal frequency is often prohibitively labor- and resource-intensive. Here we present a new method that uses accelerometers to overcome challenges of measuring high-frequency tree phenology in the field. Accelerometers are small, portable devices that can be attached to trees to record movement due to forcing by wind. Time series analyses of tree movement data recorded by accelerometers can yield an approximation of tree mass. Because leaf emergence and leaf drop alter aboveground tree mass, these phenological events are expected to be detectable from accelerometer data. To test how well accelerometers can be used to measure phenological dates, we deployed 20 accelerometers on balsam poplar (Populus balsamifera) trees across a variety of sites during the 2016 growing season and assessed how well phenology derived from accelerometers matched visual observation of phenology recorded by citizen scientists. We found that accelerometer measurements fit the theoretical expectation for the seasonal change in tree mass associated with leaf phenology; specifically, an increase in tree mass in the spring, and a decline in the autumn. Furthermore, we found that accelerometer-derived phenology matched visual observations for leaf emergence, with a strong correlation between the dates of first observed full leaves and accelerometer-derived phenology (r = 0.82, p < 0.01). Estimates of leaf drop from accelerometers and visual observations, however, were not significantly correlated (r = 0.16, p = 0.69). Our work shows that accelerometers can reliably be used to detect spring phenology of forest trees, and have the potential to overcome some of the challenges related to documenting spring tree phenology at high spatial and temporal resolution in the field.
Details
- Title: Subtitle
- Estimating tree phenology from high frequency tree movement data
- Creators
- Andrew V Gougherty - University of Maryland Center for Environmental ScienceStephen R Keller - University of VermontAnton Kruger - University of IowaCathlyn D Stylinski - University of Maryland Center for Environmental ScienceAndrew J Elmore - University of Maryland Center for Environmental ScienceMatthew C Fitzpatrick - University of Maryland Center for Environmental Science
- Resource Type
- Journal article
- Publication Details
- Agricultural and forest meteorology, Vol.263, pp.217-224
- Publisher
- Elsevier B.V
- DOI
- 10.1016/j.agrformet.2018.08.020
- ISSN
- 0168-1923
- eISSN
- 1873-2240
- Grant note
- DOI: 10.13039/100000001, name: National Science Foundation, award: IOS-1461868
- Language
- English
- Date published
- 12/15/2018
- Academic Unit
- Civil and Environmental Engineering; Electrical and Computer Engineering
- Record Identifier
- 9984197419602771
Metrics
6 Record Views