Mechanistic modeling of seed dispersal by wind over hilly terrain

A. Trakhtenbrot*, G. G. Katul, R. Nathan

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

43 Scopus citations

Abstract

Seed dispersal is the main movement mechanism used by plants. The last decade saw rapid progress in understanding the underlying processes, especially for dispersal by wind, in part due to new mechanistic modeling approaches that account for turbulent fluctuations. Yet, current wind dispersal models stop short of explicitly incorporating the effects of landscape topography on the main transporting vector - wind, so that the effects of wind variability over hills on dispersal patterns remain by and large unstudied.A new mechanistic model was developed that combines Eulerian wind statistics derived from a simplified analytical approach of flow over gently sloped forested hills with a Lagrangian seed trajectory model. Model runs were used to explore the effects of seed release location along the hill on dispersal kernels predicted by the new model in relation to their flat-terrain counterparts. The model was parameterized for a Pinus taeda plantation, and a range of seed motion capacities represented by terminal velocity and release height, and realistic topographic variation were then explored.To evaluate model performance, computed kernels were compared to kernel measurements collected in a large flume for spherical 'seeds' released near the top of a rod canopy covering gentle cosine hills. The evaluation showed that the model reproduced the key experimental differences in dispersal patterns for releases at the hill crest and bottom.The simulations revealed several novel findings. For seeds released within the canopy, both median and 99th percentile dispersal distances on the hill upwind side were up to two times longer than on flat terrain for the same motion capacity. Seeds released on the lee side traveled mostly toward the hill crest - following the local within-the-canopy wind direction. This direction was contrary to the 'regional' wind direction set by the flow conditions above the canopy. There, the directionality of the long-distance dispersal was additionally dependent on uplifting probability, affected by seed motion capacity.It was demonstrated that neglecting the effects of even gentle topography in mechanistic seed dispersal models can lead to biased estimates of dispersal distances and directionality on hilly terrain. These results are pertinent to plant population demography, connectivity and spread on hills. More broadly, the approach developed here can be extended to movement of pollen and various airborne organisms over hills.

Original languageAmerican English
Pages (from-to)29-40
Number of pages12
JournalEcological Modelling
Volume274
DOIs
StatePublished - 24 Feb 2014

Bibliographical note

Funding Information:
This work was supported by Israel Science Foundation grants ISF-474/02 , ISF-FIRST 1316/15 and ISF-150/07 ; and National Science Foundation (NSF) (through NSF-IBN-9981620 and NSF-DEB-0453665 ). A.T. acknowledges the additional support of Vaadia-BARD Postdoctoral Fellowship Award No. FI-470-2012 from the United States – Israel Binational Agricultural Research and Development (BARD) Fund. R.N. acknowledges the support of Friedrich Wilhelm Bessel Award, Humboldt Foundation; the Minerva Center for Movement Ecology; and Adelina and Massimo Della Pergola Chair of Life Sciences. G.G. K. acknowledges the support from BARD (Award No. IS-4374-11C), the United States Department of Agriculture (USDA #2011-67003-30222), and NSF (NSF-AGS-11-02227). We thank Melissa Chernick for providing us with the tree height and DBH data for A. rubrum from the FACE database.

Keywords

  • Anisotropic dispersal kernel
  • Connectivity
  • Coupled Eulerian-Lagrangian closure (CELC) approach
  • Landscape heterogeneity effects on seed vector movement
  • Mechanistic seed dispersal model
  • Seed dispersal by wind over non-flat topography

Fingerprint

Dive into the research topics of 'Mechanistic modeling of seed dispersal by wind over hilly terrain'. Together they form a unique fingerprint.

Cite this