Click on image to enlarge.

Scientific name: : : : :
Common Name: Information Sheet, Jason Robertson

Country: USA
State/District: DC
County: not applicable
Date (D-M-Y): 1 - 12 - 2003

Photographer: E. M. Barrows

Identifier: E. M. Barrows
Collector: not applicable
Location: Washington, D.C., Area
Keywords: A FE2003R Forest Ecology
Additional Information:

Instructors' Note

A student in Forest Ecology (fall 2003) at Georgetown University (GU), Washington, D.C., produced this report as an individual class project which had a limit of about 25 hours.

Goals of the course projects included:

(1) learning about a relatively natural forest (Glover-Archbold Park which is adjacent to Georgetown University, in Washington, D.C.), the "open urban forest" of GU Campuses, and the "urban forest" in residential neighborhoods near GU.

(2) learning about the scientific process while working on a hands-on field project.

(3) learning about answering scientific questions and testing hypotheses.

(4) providing information about these subjects to interested parties via the Internet.

These are short projects that lay groundwork for further investigation in their respective areas.

Therefore, these are pilot projects.

The first emphasis was on working with the scientific process, and the second emphasis was on conclusively answering questions (or conclusively testing hypotheses) as the allotted time allowed.

To see all of the 2003 online projects, please use the keyword “FE2003R” on this Website.

Projects of future Forest Ecology students might continue lines of investigation of these and past projects.

E. M. Barrows and Kyle M. Brown, instructors

Beechdrops Growth on American Beech Trees of the Glass Hill Area of GAP, Washington, D.C.

Jason O. Robertson
Department of Biology
Georgetown University
Forest Ecology
355 Fall 2003


The objective of this study is to describe the changes in number and growth of Beechdrops (Epifagus virginiana) that occur with increasing distance from American Beech tree trunks and with various grade slopes. Beechdrops commonly parasitize American Beech trees in Glover-Archbold Park (GAP), but it is unknown how root depth affects Beechdrops’ ability to attach to tree roots and subsequently to grow. If root depth were a factor in Beechdrops growth, one could suggest that some American Beeches evade some Beechdrops by having deep roots. For the purpose of this study I examined only Beechdrops and American Beech trees in the Glass Hill area of the southern part of GAP near Georgetown University. I found that Beechdrops numbers increase on the downward slopes from the trees, and I suggest that this is because the tree roots are closer to the surface. They are therefore more accessible to Beechdrops at greater distances from the tree compared to roots in the upward slope and the level sides. I found no correlation between Beechdrops height or branch number and distance from the tree (on all four sides).


Beechdrops (Epifagus virginiana) are in the family Orobanchaceae, which is commonly known as the Broomrape Family. It consists of 150 species in 17 genera, and about 90% of the species are natives of North America. This family primarily thrives in northern warm and temperate zones; although, about 10% of its species can be found in the tropics, and one species has spread as far north as the Arctic (Thieret, 1971).

The majority of the parasitic broomrapes can live only off of one or two host plant species. Epifagus virginiana lacks chlorophyll, and it completely derives nutrients from the roots of Beech trees, including American (Fagus grandifolia) and European Beech (Fagus sylvatica) trees. Beechdrops live directly on their hosts by attaching haustoria to Beech tree roots and penetrating the root tissue (Rutherford 2003; Mitich 2002).

Due to their dependence on a host, broomrapes have developed mechanisms that ensure that they begin to grow only when there is a plant nearby that can support them. Broomrape seeds are predominantly disseminated by the wind, and they have been known to survive in the soil for up to 13 years. The fact that they can lie dormant enables them to function in forest succession. However, parasitic broomrape seedlings must quickly find a suitable host if they will survive because, while they can easily penetrate the soil, they are unable to derive sustenance from it. As a result, broomrape seeds will not germinate until they are exposed to a biochemical exudate produced by their host plant’s roots. Upon germination, broomrape seeds develop a small radicle that grows chemotropically towards host roots and firmly connects to a host rootlet. The radicle immediately saps nutrients from the host and stores them as starchy reserves in its upper part, forming an engorgement. The radicle gradually forms into a nodule on the host root and some developing roots emerge from the nodule surface. These roots envelop the nodule and attach at other places along the host roots. The large bulb-like swelling then develops into a shoot and elongates and becomes the above ground portion of the broomrape. This process takes several weeks for completion. Beechdrops can grow to about 20 inches tall, and they are distinguished by clusters of whitish or whitish-purple flowers scattered along brown stems (Mitich 2002).

Broomrapes do not consume enough nutrients to kill their hosts, and this is beneficial to them because they are obligate parasites. Yet, parasitic broomrapes annually cause great economic losses around the world by reducing crop yields. In the 16th and 17th Centuries, broomrapes were also prescribed as medicinal herbs, and Beechdrops, specifically, were used as a powerful astringent (Mitich 2002).

In my study, I collected data concerning the patterns of Beechdrop growth at increasing distance and slope from the trunks of American Beech trees. I hypothesized that fewer Beechdrops are found on uphill slopes because the seedlings potentially die before growing downward through the soil enough to attach to a root, and I thought that those Beechdrops that were found further from the tree would be shorter and contain fewer branches due to the fact that more energy and time would be required to penetrate through the top of the soil. This study will give scientists, working in Glover-Archbold Park, a better understanding of the patterns of Beechdrops growth, and it will contribute to the wealth of general ecological knowledge by determining whether or not distance from American Beech trees and root depth (in a preliminary fashion) affect Beechdrops growth in forest environments.

Materials and Methods

I collected data from the Glass Hill area of GAP, which is populated by many American Beech trees, in October of 2003. I selected five American Beech trees for my study in a non-random manner based upon three criteria: 1) large size (greater than 10-inch trunk diameter), 2) isolation (5 feet or more away from all other American Beeches), and 3) location on a steep (approximately 45̊ or more) downward slope. All American Beech trees in this area had Beechdrops. Isolation of a study tree from other Beeches was important in order to ensure that at increasing distances from a study tree the Beechdrops weren’t growing due to attachment to roots from other trees. Notably, American Beech trees can also be found in the Ant Hill area of GAP, but due to some unknown circumstance no Beechdrops grow in that area of the Park.

After I selected a tree, I collected along four transects, one for each of the four different sides of the tree. The transects ran for 60 inches outward from each side. One transect was run on the downslope side, one on the upslope side and one on each of the sides, left and right, oriented from the upslope side. I collected data on all Beechdrops that were located within 3 inches of either side of the transect lines. Measurements for each of the Beechdrops included height, branch number, and distance from the tree, and they were determined using a meter stick and a 12-inch ruler. I tabulated the total numbers of Beechdrops along each transect after I finished gathering data.

I also averaged the Beechdrops height over the intervals of 0–10, 11–20, 21–30, 31–40, 41–50, and 51–60 inches in order to create one data point per interval per tree. I did the same for the Beechdrops branch number. I then analyzed the data for the upslope, downslope and sides with linear regression and calculated P to test the alternate hypothesis that the slope of the regression line ¹ 0. I analyzed other data from boxplots and histograms using Tukey-Kramer ANOVA tests. All assumptions were met for each statistical test.

Results and Discussion

The number of Beechdrops within 60 inches of an American Beech trunk on the downward sloping side is, on average, more than double the number found on either the upward slope or the left and right sides, which were fairly similar differing only in that there is a narrower range of Beechdrops numbers on the left side of trees (Figure 1). There is a statistical difference between the upslope and downslope data, between the left side and downslope data, and between the right side and downslope data (Tukey-Kramer ANOVA, P < 0.05). There is no statistical difference between the numbers of Beechdrops on the two sides or between either of the two sides and the upslope data (Tukey-Kramer ANOVA, P > 0.05). The variation that occurs is possibly due to chance alone. The above data suggest that some condition of the downward slope favors Beechdrops growth.

I hypothesized that the reason for this difference was that roots were more accessible to Beechdrops on the downward slope due to shallower soil and erosion, which would allow a greater number of Beechdrops to easily parasitize the roots. If this were the case, then one would expect to find a uniform distribution of Beechdrops numbers with increasing distance intervals from the tree on the downward slope, whereas one would expect to see a decrease in number of Beechdrops with distance from the trunk in the upslope and left and right side directions. This would result from the fact that in the latter three cases, roots would become progressively deeper and more inaccessible with distance. On the downslope, however, the roots are most always still exposed 60 inches from the tree. Indeed, I observed a more marked decrease in Beechdrops numbers with distance on the two sides and the upward slope than on the downslope. At each interval up to and including the 51–60-inch distance interval, there are more Beechdrops on the downslope side than any other side (Figure 2).

There is a statistically significant decrease between the 0–10- and 31–40-, 41–50- and 51–60-inch distance intervals for the upslope data (Tukey-Kramer ANOVA, P < 0.05). There is no such statistically significant decrease for the downward sloping data except for a difference between the 31–40- and 51–60-inch intervals. All differences are determined at " = 0.05. The numbers collected for each interval from the sides are more comparable to each other and to the upward slope data than they are to the downslope data. Furthermore, regression analysis of the downslope intervals is significance at P = 0.045 with an R2-value of 10.5, whereas regression analysis of the upslope intervals rejects the null at P = 0.001 and an R2-value of 30.8. Therefore, about three times more variation in the numbers of Beechdrops can be explained by distance on the upward slope than on the downward slope, and it should be noted that P for the upslope data is much closer to 0.05 than in the cases. Regression that I performed for the left side yielded P = 0.009 and R2 = 19.0, and regression that I performed for the right side yielded P = 0.003 and R2 = 24.7. As expected, the distance variable is most significant on the upward slope and least significant on the downward slope. Thus, a negative correlation between Beechdrops height and distance from tree trunks is observed that seems to be affected by root depth, as the grade of the slope affects Beechdrops growth.

However, it is important to note that data collection could not lead to an ideal analysis. No tree was perfectly isolated. That is, the American Beech trees of Glass Hill are fairly close together and they are large enough that it was impossible for me to guarantee that the roots of neighboring trees didn’t interweave with those of my sample trees. If such interweaving occurred then I could have counted Beechdrops from a neighboring tree’s roots as parasitizing my experimental tree. This was more of a problem on the sides and on the upward slope because often the degree of the downward slope precluded tree growth for greater distances from the tree. It was also impossible to determine what course the roots follow underground because the National Park Service did not allow me to excavate roots. It is likely that they dip up and down at different intervals. Therefore, it is possible that the roots do not get progressively deeper as we move up hill or to either side, which would confound my analysis. As it stands, I have to say that on average it is likely that with increasing distance the roots follow a downward path.

I was unable to establish any correlations between Beechdrops height or branch number and distance from American Beech trees. In all cases I did not reject the null hypothesis that the slope of the regression line was equal to zero. For the height variable, I obtained R2 values from 0.7 through 3.0 and for the branch number variable I obtained R2 values from 0.1 through 0.7 (Table 1). Clearly, I see no variance in either height or branch number as distance increases from the tree. This suggests that as long as Beechdrops can tap into a root they are capable of growing normally at any distance or root depth. Again, however, the above data could have been confounded by roots from nearby trees extending into the periphery of my sample tree or by undulations in the paths of the roots underground that prevent clean data.

In all data analyses my conclusions were limited by low sample sizes. Time constraints for this project, as well as the availability of only a few American Beech trees that met my isolation criteria, prevented heavier sampling. Future studies should be more expansive, and it may also be worth controlling for the differences in the degree of the slopes in all directions from the American Beech trunks.

Additionally, my study was burdened by the fact that I was able to sample only a small population of Beechdrops in one local area of GAP (Glass Hill). The other nearby area containing American Beech trees, Ant Hill, was not parasitized by Beechdrops. Random selection of American Beech trees from the whole of GAP would be recommended for more conclusive future studies. It might also be beneficial to make the isolation criteria for acceptable trees stricter so as to reduce the possibility of confounding variables in the analysis.


I would like to thank Dr. Edward M. Barrows for his guidance and the expertise that he contributed to this project. I also thank Dan S. Kjar for his assistance in publishing this project on the Biodiversity Database of the Washington, D.C., Area (BDWA).

Literature Cited

Mitich, L. W. 2002. Orobanche – The Broomrape. Internet file. (11 December 2003)

Thieret, J. W. 1971. The genera of Orobanchaceae in the southeastern United States. Journal of the Arnold Arboretum 52: 404-413.

Merops. The Peptidase Database. 2003. Epifagus virginiana (Beechdrops) Taxonomy. Internet file. (11 December 2003)

Rutherford, J. 2003. Epifagus virginiana. Internet file. (11 December 2003)

 Table 1.  Branch number and height of Beechdrops are not correlated with distance from Fagus grandifolia.

Linear regression R-squared-value P Significance (alpha = 0.05)
Beechdrops height      
— Upslope 3.0 0.482 no
— Sides 0.8 0.661 no
— Down slope 0.7 0.653 no
Beechdrops branch number      
— Upslope 0.1 0.923 no
— Sides 0.1 0.853 no
— Down slope 0.7 0.657 no

update template
�Copyright 2009 Georgetown University