Several years ago Dr. Ilhan Temizsoy, director of the Museum of Anatolian Civilizations, expressed concern about erosion on the Midas Mound at Gordion (Fig. 0. It occurred to me that the most effective way to reduce soil loss would be to have an uninterrupted cover of plants grow on the mound surface. At that time, the vegetation looked much like the surrounding overgrazed pasture, which is characterized by bare ground, many spiny plants, and plants that de­fend themselves against grazing goats and sheep by producing bitter compounds in their leaves (Fig. 2). I also began to envision an even grander project—recovery or partial restoration of the vegetation native to this part of Turkey. The cen­tral Anatolian steppe grassland is highly diverse, and includes many perennial grasses, wildflowers, and small shrubs that stay green all or most of the year and that are palatable to grazers (Fig. 3). Restored vegetation would not only be beautiful, but would also improve the rangeland.

In the short term, maintaining and expand­ing the plant cover would be an effective way to reduce erosion. Plant cover would keep the strong Anatolian winds from blowing soil away, would diminish the force of water reaching the soil sur­face, and underground, the roots would reduce the total volume of water reaching the bottom of the mound by absorbing rainwater. This approach would not solve all the problems, because the steep slope of the mound naturally limits plant growth. Also, the mound is the high point of the village, which makes climbing it almost irresist­ible to local young people and tourists.

Following the concept, “first do no harm,” it seemed best initially just to encourage the exist­ing plant cover to spread; thus, in April 1996 the General Directorate of Monuments and Mu­seums erected a fence to keep animals and people off the mound (Fig. 4). By the summer of 1997. a luxuriant cover of grasses and other plants on the lower slopes demonstrated that the fence had begun to work at the base of the mound. Further up the mound, however, deep channels were still bare of any fresh vegetation.

Mudbrick for Erosion Control

In the summer of 1997, the project benefited greatly from a site visit by Kurt Bluemel, an ex­pert in ornamental grasses and landscaping. As he and I walked all around the mound, it became clear to us that the most serious problem was now the erosion channels. The force of water flowing in the channels was strong enough to move even fairly large stones down the slopes, so plants could not gain a foothold. I asked Kurt what he thought about lining the channels with sunbaked mudbricks, the traditional Near Eastern building material; by absorbing water and dis­solving gradually into the channels as the winter rains progressed, they would effectively slow the rate of water flow and allow seedlings to become established. Kurt thought this idea was worth a try and showed us how the mudbricks should be set.

The Gordion Project arranged to have 1000 mudbricks from a dismantled village structure placed in two of the erosion channels. The Gor­dion site caretaker, Remzi Yilmaz. was drafted to supervise the laying of bricks in the fall of 1997 (Figs. 5, 6). The work had to be done by hand because heavy equipment would destroy the soft surface of the mound. As I described the plan to Remzi, the watchman for the dig house listened in. I explained that it was just an “experiment”; if it did not work out, I didn’t want Remzi to blame himself. But then I caught the eye of the watchman. “You think I’m crazy, don’t you?” I asked him; he smiled and said, “Well, if it works, you’re not crazy.”

By the spring of i998, plants were thriving in both channels where the bricks had been laid (Fig. 7), and I have counted and mapped over 125 different types of plants. distributed accord­ing to their own growth requirements (Fig. 8).

The mapping demonstrates that there is no one plant or even group of plants that is best for the mound (Fig. 9). For example, one perennial grass, Stipa arabica, does very well in the southeast sector, while another one, Bromus cappadocicus, is most abundant in the north sector.

In the summer of 1999, we set mudbricks in two more erosion channels, and we continue monitoring the vegetation to assess our progress and fine-tune our methods for the long-term survival of plants on the Midas Mound surface. It is possible that in the total absence of grazing, ac­cumulating dead plant material might catch fire. In addition, there is no longer a source of animal droppings to fertilize the mound, which might adversely affect the steppe vegetation. Within the next few years, it may be desirable to allow some animals in to graze under controlled conditions.

Clearly, however, encouraging native plants works. As it stands, the variety of plants on the Midas Mound can survive the fluctuations of wind and weather that occur from year to year. Even when the spring wildflowers are gone, many of the other plants survive well into the summer or year-round, so there should always be some color on the mound.  Gordion’s Changing Landscape Climatically. Gordion lies at the upper limit of the central Anatolian steppe, and this region might actually be capable of supporting a savannah-type vegetation (dominated by grasses but with widely scattered trees). In this part of Turkey, precipitation increases with elevation. Trees, largely absent near Gordion outside of gardens, do grow unwatered and untended as close as 10 kilometers from the site today.

As the archaeobotanist for the Gordion project, my research (which includes both archaeo­logical and modern plant studies) informs the conservation work. Plant remains from excavations directed by Dr. Mary M. Voigt on the settlement site (approximately I kilometer from the mound) reflect how humans have changed the vegetation over a long period. The remains consist of wood charcoal and charred seeds. Many of the seeds come from plants consumed by animals and excreted in dung. (In wood-poor environments, dung is a common alternative fuel.) Therefore, both seeds and wood tell us first about fuel use, and secondarily allow us to infer the landscapes of the past.

At the beginning of the Gordion archaeological sequence (Late Bronze Age, about 3200 years ago), scrubby juniper was scattered amidst the grassland and cultivated fields. The flocks grazed on a steppe dominated by grasses, but which also produced many other nutritious plants. Over time, the nearby juniper was cut, and people went to more distant areas for firewood of oak (today about 10 km from the site) and pine (about 50 km). Indeed, by the time Midas’s tomb was constructed of large pine and juniper logs, scrubby juniper was no longer available for fuel. Overall, the tree cover within easy collecting distance of the settlement was reduced. In addition. pasture quality seems to have declined, as some of the best forage plants (egg., clover-like Trigonello) made up a smaller proportion of the seed assemblage.


Future Plans for Vegetation Restoration

Successful as this first phase of the work ap­pears to have been, it should be stressed that these plants do not constitute a “reconstruction” of the natural vegetation. The mound surface of ancient times, on which pollen or other evidence of growing plants might have been preserved, is long gone, so it is impossible to restore the vegetation cover to some hypothetical original state. In fact, from an archaeological perspec­tive, we have no idea what, if any, plants grew on the tumulus in antiquity. although our research is continually providing new information on the ancient landscape (see box, “Gordion’s Changing Landscape”). It is somewhat misleading even to use the word “natural” in the context of a land­scape that people have inhabited fairly intensively for at least 5000 years.

The conservation work at the Midas Mound does, however, have significance well beyond Gordion itself. We have shown that erosion con­trol can be inexpensive and can use locally avail­able technologies, while maintaining the diverse flora of the region. Our methods may begin to demonstrate the value, and relative ease, of re­storing rangeland by keeping animals off over­grazed pasture for just a few years. After all, the spiny and inedible plants are themselves compo­nents of the native steppe vegetation and only come to the fore because the flocks preferentially eat the rich pasture grasses and legumes.

It is clear that despite several thousand years of farming, herding, and woodcutting, much of the vegetation can heal itself with limited hu­man intervention. There is little doubt that the potential beauty of the landscape can be brought forth with a modest amount of thoughtful plan­ning. Modern farming and herding can flourish even as tourists come to enjoy the open spaces, interesting plants, and of course, the archaeo­logical ruins and museum that any visitor to Gor­dion should see and experience.

As a Senior Research Scientist in MASCA (archaeobatanical section), NAOMI F. MILLER studies ancient environ­ment and land use in the Near East. She has been working at Gorton, Turkey, on plant remains from the archaeological site. Her appreciation of ancient and modern vegetation in the region has also inspired the steppe restoration project for the Midas Tumulus, reported here. Other current projects include archacobatanical investigations along the Euphrates river in Syria and Turkey and at Anau, Turkmenistan. You can visit her website at