Winching studies to directly measure tree windfirmness

From the very earliest studies of wind damage in forests, it was apparent that there was substantial tree-to-tree variation in levels of damage.  Some of this variation is of course due to different winds in different locations, or to site factors such as soil saturation.  But most likely a large portion of the variability results from inherent differences between trees in their windfirmness.  This tree characteristic can be estimated based on theoretical expectations, but empirical measurement is essential to truly document tree strength.  Such empirical data can be generated in only one way:  by winching trees to failure, and measuring the force needed to cause uprooting or trunk breakage.  This is a tedious and time-consuming endeavor, which may explain why (to our knowledge) there is only one other lab in the U.S. that actively uses this technique (an arboriculture lab in Massachusetts that studies street & residential trees).  Winching studies have been conducted extensively in Canada and the U.K., and to a lesser degree in Japan and Finland.  Prior to our work, there were only two published winching studies with forest trees in the U.S.

The procedure for such winching studies (called static winching) consists essentially of a winch and strap system attached to the base of an “anchor” tree and just below the canopy of a winched tree; the system includes a strain gauge (or load cell) that continuously measures the force in the winch/strap system and outputs that force to a file on a laptop computer.  Simultaneously, two tilt sensors attached to the tree also continuously output the tilt away from vertical of the tree during the winching process.  The setup is illustrated schematically in this graphic.


The winching itself it shown in this photo, which illustrates one person on the winch, and another ensuring that the steel cable that moves through the winch does not get tangled. 


For a subsample of the trees, it is necessary to cut them into 1 m sections for weighing, as shown here.


Using funding from the California Dept. of Water Resources, we completed our first winching study on levee trees (valley oak Quercus lobata, and western cottonwood Populus fremontii) in central California in 2011; this was published as Peterson & Claassen 2013.  We found that due to large site to site variation, the interspecific differences between these two species were not significant, although they were substantial.  Not surprisingly, there was a strong increase in windfirmness with increases in tree sizes.  On the sandy soil of the levees, soil moisture levels had little influence on tree windfirmness, although it is worth noting that the soil moisture conditions during our winching tests did not approach saturation; saturated soils may have led to much lower windfirmness of winched trees.

In 2012, as part of setting up the artificial gaps for Jeff Cannon’s dissertation (wind & fire), we completed a winching study on loblolly pine (Pinus taeda) and tulip poplar (Liriodendron tulipifera) at the Piedmont National Wildlife Refuge in central Georgia.  One of our 2012 undergraduates, Meredith Barrett, is leading the effort to write up these results for publication; we expect a manuscript to be submitted in November 2013.  Surprisingly, we did not find any substantial difference between the critical turning moments of similar-sized pines and tulip poplars; this contrasts with the general received wisdom that conifers are more vulnerable to wind damage than hardwoods. 



In Fall 2013, we are embarking on a new collaboration with Dr. Brady Hardiman and Dr. Michael Dietze, both of Boston University, to conduct a winching study in northern Wisconsin at the UNDERC field station.  This undertaking will help Brady and Michael establish an artificial gap experiment in a northern hardwoods forest for their studies of ecosystem carbon dynamics.