Little Elm Woods History
Waste Management's Mesquite Wildlife Management Area


Introduction
On Waste Management's Mesquite Wildlife Area there are several woods but one grove appears much different than most natural woods.  On the property, there are mixed mesquite woods, mixed hillside woods, mixed bottomland woods, and the unique isolated elm grove.  This elm grove attracted my attention because of its young look, almost single species nature, and its seemingly random placement.  I at first thought it could be a rapidly spreading invasive situation.  See figure 1 and figure 2.


Purpose
The purpose of this study was to determine if the tree species was an invasive species, whether the grove was rapidly spreading, and when the grove began.  If possible I also wanted to get some insight into why the grove grew where it did.


Methods
To address the species issue I compared the various appearances of the elms to many identification sources.  To address whether the grove was rapidly spreading I used two approaches, increment coring of selected elms, and a study of an array of sequential air and satellite views of the grove.  Increment coring, backed up by the air photo sequence established when the grove began.  The older air photos hinted at why the grove began where it did.  To see how the grove expanded, tree age was linked to tree diameter through a derived relationship from cored elms and applied to hundreds of non-cored elms to give age distribution throughout the grove.

The first step in determining the age of the grove and  growth rate was to get age directly from some of the older trees using tree rings from increment cores, and to also get age from a couple of the little two to six foot elms on the periphery of the grove.

Using a GPS enabled mapping unit I traced off on the ground the perimeter of the grove and the also the elm tree core.
 
Surveying the tree location was a two step process.  Since GPS signals within the woods are poor, I surveyed in several stations outside and within the woods as reference points and then used a compass and tape for azimuth and bearing from survey stations to the trees to be aged.  Using this technique provided multiple points for each survey station, allowing the GPS satellites to pass overhead, between and around limbs.  This gave numerous slightly erroneous locations for each station which could then be averaged to provide greater accuracy of those reference survey points.  Accuracy of the survey stations went up from about 2.5 meters to about 1.5 meters.  It would have been good to use triangulation to further improve both the survey reference points accuracy and the actual chosen tree locations.  However with the woods, too few of the stations and trees could be seen form other stations and trees so the simpler azimuth and distance method had to be used.  The compass I used was a sight-through model readable to better than a degree. I measured the distance with a long field tape.  The azimuth direction was corrected for magnetic declination.  This measuring routine decreased the tree trunk accuracy from the 1.5 meters of the survey stations to about 2.5 meters for the tree trunks.  Even so, the two step process yielded better tree location accuracy than would have been acquired by trying to read a GPS location directly within the canopied woods.  Many of the larger-trunked trees would have blocked so many GPS satellites as to provide no useable location at all, had a GPS location been attempted from within 1 meter of the trunks.  Chosen trees were marked with temporary pink tape. The grove perimeter, inner elm core of the grove, and the cored tree locations are shown on figure 3.

Cores of older trees were taken with a 14 inch 5.15mm Haglof corer while smaller trees were cored with an 8 inch 4.3mm Haglof corer. For consistency of tree ring work, all cores were taken at a four-foot height.  Normally tree cores are taken at “breast height” but I prefer to use a height of four feet tin central Texas.  To later convert the ages determined at four-foot height, two small trees were sawed at both four-foot height and at ground level to determine the number of years the trees took to reach four-foot height.

Each cored tree was given an identification number and each core or disc taken from the trees was given the number corresponding to the tree it was taken from.  Each cored or sawed tree was also photographed and those photos were  given the same ID number of the trees cored or sawed.

Each core was glued into a grooved wooden tree core form and the ID number transferred to the core form.  After glue drying, the cores were sanded with three successively finer sandpaper grades.   The small tree-sawed discs were also sanded down similarly.  Tree discs and cores were then polished on heavy paper.

I tried three core inspection methods, visual, jeweler’s visor at 2.75X, and binocular microscope at 10 and 30X.  The middle magnification, the jeweler’s visor was the most appropriate for detecting annual growth rings on this set of cores.  Growth rings were marked and counted for each sample.  Conservative and liberal counts of each sample showed a variance of about 2 tree rings in thirty rings, or an age accuracy of six percent.
 
The two small trees sawed for age difference from ground to four feet tall showed that they took three and five years respectively to reach four feet tall.  All tree ages were later adjusted by four years to bring the ages taken at four feet to the equivalent ages at ground level when the trees sprouted.

Each tree cored was also measured in diameter at the same four-foot height level.  The diameter was measured with a forestry circumference tape that read out tree diameter directly in centimeters.  In Excel, a graph was made of tree age versus diameter.  This chart, figure 4, provided a means to then get a reasonably useful elm age from many other trees by measuring their diameter at the four-foot height. 

The graph seems to exhibit two growth rates even though all cored elm trees were on the same acre of flat ground with deep soil.  The slow growing ones always had the thicker rougher bark regardless of age and were distributed randomly amongst the faster growing elms.  There may be some Winged Elms amongst the Cedar Elms.  Leaves from both slow and fast growing and from smoother and rougher barked elms had the identical small leaves.  See figure 5.  Rather than try to treat these differences with separate diameter-to-age charts, I averaged the two for a single growth rate curve.  The best fit for this was the equation below

Age at 4 ft =((tree diameter in centimeters)^0.6)*4.6

In figure 4, the pink dots are the estimated age from this best-fit formula, while the true age at four-foot height from cores are the dark-blue dots.  From discs taken at both ground and at 4 feet height, there is about a four year time span to reach four feet in height, so this four years is added to the data obtained at four feet.

Age at sprouting =(age at 4’)+4

To get age distribution of the elms in the grove, I measured diameters of each elm within a 25-foot radius of chosen circles and also kept track of the number of elms within each circle.

Elm diameters were converted to estimated age with the above formula in Excel and median as well as average ages were computed.  Figure 6 shows a view of the median age for each area within the grove at four-foot height.  Note the young trees near the most northeast point have skewed this area to a younger value.  I will get to this a little later.  The elm density within the grove, (number of elms within the 25-foot radius), is shown contoured in figure 7.  This contoured elm density map was used to compute the number of elms within the grove.  There are about 360 elms in the grove.

Within Excel, age distribution of elms in the grove was divided into 5-year segments. The ratio of elms in the grove, compared to the number elms measured provided a calibration factor for the numbers of trees in each age bracket.  After adjusting the age distribution by this calibration factor, and adding four years to bring the age to ground-level sprouting age, the age distribution chart seen in figure 8 shows reasonably accurate age distribution of the elms within the grove.

With age distribution of the elms, it was then possible to reconstruct the chronology of the grove’s establishment.  Figure 9 shows the number of trees in the grove from 1965 through the present in 5-year increments.

I used air photos and satellite images to independently collaborate results of the tree core data.
 
Photos of the area were taken by Tobin International in 1938, 1961, and 1988.  Tobin became P2 Energy corp.  I purchased segments of these air photos.  Satellite images taken in 1973, 1976, 1979, 1981, 1983, 1987, and 1999 were downloaded from either the USGS satellite portal of the TNRIS digital download site.  I obtained additional digital air coverage at 1-meter resolution as DOQQS from TNRIS in 1996, 2000, and 2004.

The sequential aerial sequence clearly showed the birth of this grove and its evolution into a closed-canopy woods, consistent with the tree ring core information.  The air photo sequence also showed the independent growth of the other woods on the property.


Results
The grove’s dominant species is Cedar Elm.  There are probably a few Winged Elms within the grove but only flower and seeding in the spring will differentiate them in my mind from the late summer-fall flower and seeding of the Cedar Elms.  Chinese Elm was ruled out.  Siberian Elm as a pure species is nearly ruled out as well.  I found the grove was not expanding.  There are very few elm trees less than 11 years old.  From core examination, nearly all trees, even the younger ones, have grown more slowly in the last ten years than previously.  Sprouting of new elms nearly stopped back in 1995 and has not resumed.  See figure 9.

From cores, the first tree in the grove sprouted in the savanna in 1955.  This tree is shown in figure 10.  Another tree, hit by lightning in 1980-1981, sprouted between 1953 and 1960.  See figure 11.  With its core burned out, its exact age is indeterminate.  Both these trees are near the center of the current grove and their location, (or future location), is highlighted in figure 12 in 1938, figure 13 in 1961, figure 14 in 1988, and figure 15 in 2004.   By 1970 there were about 10 elms.  By  1975, there were about 40.  By 1980 there were about 100.  By 1995, when sprouting declined, there were about 340 elms.  Today there are about 360 elms in the one-acre grove.  Elms occupy 0.85 acres with the remaining periphery being mostly hackberry and mesquite but with minor numbers of Live Oaks.  The median age of the elms calculates as  22.9 years old, and is seen in the age distribution chart of figure 8.
  
The grove began in a savanna between two plowed fields (see figure 12 and figure 13).  Air photos going back to 1938 suggest that this grove was never associated with any other grove or woods and that there were never any woods at this location prior to the first two elms in about 1955.  In contrast, the mixed slope woods by Schwatzlose and the mixed bottomland woods towards the 1101 and Kohlenberg intersection were well underway by 1938.   See figure 16.  Mixed Mesquite woods, not a part of this study, are a continuing development of the area from grassland through Mesquite savannah to Mesquite woods. 


Discussion
Even though there is a core of almost exclusively elm in the little elm grove, there is a periphery of hackberry and mesquite.   The elms, especially the larger ones are surviving the draught very well while the hackberry trees in the periphery are not.  Since elms are usually relatively long-lived, the grove will likely remain another half century, barring changes in climate or land use, even without any new young elms.

Birds are using the grove.  David Pueppke and I saw several species of migrant warblers feeding within the elms in May.  Bob Norris and I saw White-winged Doves in trees adjacent to the grove many times in spring.  I found numerous hatched dove eggshells within the grove in late June and into July suggesting it was used this year as a high-density dove nursery, probably White-winged Dove.  See figure 17.   Several Great Horned Owl feathers found also in the June through early July time period indicate one of those owls was using the grove also, possibly for the doves.   See figure 18.


Conclusions
This elm grove is not spreading and began with two elms sprouting around 1955 in a savanna.  It was never associated with the much older mixed woods on the property.  It was also not associated with any corral or yard.  The nearly 360 elms within it range mostly from 11 to 40 years old with a median age of 22.9 years old.  The first few elms got their start in a savanna between plowed fields, but most of the elms sprouted after 1975 from airborne seeds from these few earlier elms.  This corresponds to Freedom Lake’s first filling after its construction.  It is possible the lake provided just a little moisture to the river floodplain, (of Mesquite Creek),on which the grove is perched, but not in all years.  The lake went dry around 1986 and was still dry in 1987 even while the rate of elm sprouting remained high.  The lake was filling again by 1988 but the sprouting rate tapered off around 1995.  Today sprouting is only occurring in a local place on the northeast edge of the grove.  Even without a renewal of sprouting in the remainder of the grove, it should remain healthy for several more decades, as its trees are mostly still relatively young.


Acknowledgments
Tobin International took air photos of the area in 1938, 1961, and 1988.  I purchased portions of them from its successor, P2 Energy.

Satellite images from 1972 through 1990 were downloaded from the USGS website.

Various DOQQs (corrected very accurate 1-meter resolution air photos) from 1996 through 2004 were downloaded from the Texas Natural Resources Information System website.

Increment coring techniques were taught by Dr. Butler and Dr. Lynn Ressler at Texas State Univ. around 2002.

Field mapping techniques pertaining to quantifying woods was taught by Dr. Earl at Texas State Univ. around 2003.

Integrating mapping grade GPS into field mapping and into GIS and the required quality control was taught by Karen Steed Terry around 2000 in Austin Tx.

Air photos, satellite images, elm locations, grove perimeters, and field-gathered data associated with elms was brought together in ArcGis software.

Statistics on trees and on the grove were done in Excel and ArcGis software.