The story of Roman Arabia begins with the death of the Nabataean king, Rabb’el II soter, in 106 CE. The emperor Trajan annexed the Nabataean kingdom and organized its territory within the new Roman province of Arabia Petraea that occupied a large area of northwestern Arabia, extending from the Hauran in the north, to the Negev and Sinai deserts in the west, and into the northern Hejaz in the south. The administrative district of Perea, formerly under the Herodian dynasty was also added. The provincial capital of Arabia Petraea was located at Bostra which replaced Petra as the capital of Nabataea in the last decades of that kingdom’s independence. Petra continued to function as an important cultural, political and economic center of the province for at least a century following annexation. During the early Byzantine period (4th-6th centuries CE), the major cities of the Roman era continued to flourish, and Christianity gradually became the accepted religion of the area. Petra remained an important center in the early Byzantine period when it was inhabited by a substantial Christian community and was the seat of a bishopric.
The region of Arabia Petraea is primarily a desert environment that belongs to the eastern Mediterranean weather system. Its inhabitants experienced marked seasonal contrasts with dry, hot summers and cold, wet winters. Rainfall occurs during the winter months and ranges from over 600 millimeters in the north-west to as low as 100 millimeters in the south and east of the region.
Petra, known to its ancient inhabitants as Raqmu, is located approximately 80 kilometers southeast of the Dead Sea in modern Jordan. The ancient city is situated on the western slopes of the Sharā Mountain Range, in a wide basin surrounded by deeply eroded mountainous ridges of sandstone, with outcrops of limestone and porphyry that form a natural fortification.
The development of Petra as an important political, economic, and cultural center in the Near East during the Hellenistic-Roman periods (1st century BCE-2nd century CE) was ultimately the result of its strategic position on the crossroads of the major trade and communication routes that crossed the region (Fig. 1). Wadi Mousa, the narrow valley along which Petra developed, provided one of the few convenient and negotiable routes through the mountainous barrier allowing caravans to pass between the high desert plateau to the east and Wādī ʻAraba to the west. This strategic position on the crossroads of the desert trade routes made it possible for the Nabataeans to establish control over the caravans that traveled between Arabia and the Mediterranean.
The Southern Terrace is a series of monumental built terraces constructed along the Colonnaded Street on the southern slope of Wadi Musa, which bisects the city center of Petra. In the early 20th century, prior to the onset of archaeological excavations, the major components of the Southern Terrace were assigned the names (from west to east) the “Great Temple”, the “Lower Market”, the “Middle Market,” and “Upper Market.” The “Lower Market” has since been identified as a garden and pool complex and the author proposes that it is one component of a large palatial complex that occupied the entirety of the Southern Terrace and dominated the City Center (Figs. 2 and 5).
Petra Garden and Pool Complex
- artificial islands
- artificial water channel
- castelli aquae
- exedrae (site elements)
- garden pavilions
- groves (plant communities)
- horticultural containers
- lead pipelines
- reflecting pools
- reservoirs (water distribution structures)
- root (plant material)
- swimming pools
- terrace gardens
- terracotta pipelines
- water features (landscaping)
The only example of a garden investigated using specialized garden archaeology techniques in the region of Arabia Petraea is found at Petra. A large open area was believed to be the site of a marketplace, the so-called “Lower Market,” until excavations revealed the remnants of a formal garden and pool complex. The garden is situated on an artificial terrace on the southern slope of the Wadi Musa, overlooking the main colonnaded street and flanked by the so-called “Middle Market” to the east, and the Great Temple complex to the west. To the south of the garden is a residential quarter on top of a high sandstone ridge (Ez-Zantur), where several private villas have been excavated (Fig. 5). Primary access into the garden was apparently from the west, through a triple colonnade that marks the boundary between the Great Temple’s “lower temenos” and the garden terrace.
The garden is composed of two distinct parts: on the north is a large earthen terrace that is elevated approximately six meters above the main street and is level with the Great Temple’s “lower temenos”; to the south of the earthen terrace and elevated another two and a half meters above it is a monumental pool with a central island-pavilion (Fig. 3). A metrological analysis of the site reveals that the Roman pes was employed in the layout of the garden terrace and pool.
The garden’s earthen terrace measures 65 by 53 meters and is supported on the north by a double retaining wall. Prior to archaeological study of the site, the terrace appeared devoid of visible architectural features, with the exception of a short stretch of a low wall at its center. The terrace’s southwest quarter has a raised earthen area, bounded on the north and west by reused architectural fragments, attesting to the modern use of the site as an agricultural field by the local Bedouin inhabitants of Petra.
The application of ground-penetrating radar (GPR) provided valuable information about the buried garden (Fig 4). It revealed that the terrace is primarily earthen with a series of stone structures aligned along its central north-south axis. Follow-up excavations uncovered two stone platforms—one at the center of the terrace and one several meters to the south, between the central platform and the pool. Each platform is constructed of a core of unhewn fieldstones faced with sandstone ashlars, typical of Nabataean masonry. There is also evidence that the southern platform was enlarged and redesigned, probably during the Roman and Byzantine periods. At the northern end of the terrace, aligned with the platforms, two soundings exposed the foundations of a large structure that appeared as a pronounced rectangular stone outline (ca. 8.5 x 11 meters) in the GPR data. The gravel interior floor surface and lack of evidence for substantial walls or a roof suggest that it was an open structure, perhaps a pergola, designed to provide shade while viewing the surrounding garden. Remnants of a gravel pathway were uncovered along the central east-west axis in Trench 17 (Fig. 6). The pathway’s border is defined by the foundations of a stylobate, suggesting a colonnaded or trellised walkway with bays for constructed views of the garden (Fig. 7). Fragments of a planting pot (Fig. 8), perforated for drainage and aeration, were found near the pathway along with several root cavities, which provide evidence for plantings along the path’s edge. A similar planting pot fragment was uncovered during preliminary excavations in the so-called “Upper Market” in 2009 (Alcock et al. 2010: fig. 13.2) and may be indicative of a second garden site on Petra’s Southern Terrace.
Soil samples taken from the garden terrace produced archaeobotanical specimens for analysis. The data from the macrobotanical remains conducted by Jennifer H. Ramsey (Table 1) can tell us little about the plants growing in the Petra Garden but does identify the types of plants – grape vines, and fruit and nut trees – that would be expected to be growing in a pleasure garden in the Hellenistic/Roman Near East and depicted in a frescoe in the “Painted House” at nearby Siq al-Barid (“Little Petra”) (Fig. 9). A phytolith study conducted by Carlos E. Cordova has shown evidence for date palm trees (Phoenix dactylifera), possible a grove, and the presence of Panicoideae grasses which grow only in the summertime in wet areas and therefore suggests frequent irrigation of the gardens during the dry, hot summers. The cereal and weed taxa identified in the macrobotanical remains (Table 1) are interpreted as evidence for what was happening in the larger agrarian landscape of the site and the region.
To the south of the garden terrace, the sandstone hillside was quarried out to create a space for a monumental pool encircled by a promenade that is elevated 2.5 meters above the garden. A staircase provides access from the southwest corner of the garden terrace to the pool’s west promenade (Fig. 3). The area of the pool-complex measures 65 x 32 meters and is bounded on the south and east by vertical sandstone escarpments rising up to 16 meters above pool-level; to the west is the Eastern Perimeter Wall of the Great Temple complex. The pool is a monumental construction with massive walls built to hold approximately 2069 cubic meters (547,000 gallons) of water. Stretching at least forty-six meters across the width of the site, the pool’s north wall measures three and a half meters wide and is preserved to a height of six courses (2.2 m). The wall’s exterior face is constructed of close-fitting sandstone ashlars typical of Nabataean masonry; between them are alternating rows of roughly hewn sandstone blocks and rubble, bonded into a solid mass with an impervious white mortar. Large rectangular pavers preserved in situ indicate there was once a nicely paved promenade around the pool’s perimeter. Porticoed promenades were commonly associated with ornamental pools of the Hellenistic and Roman periods. To the rear of the pool, a screen wall masks the quarried face of the naturally veined bedrock and displays a central niche and flanking exedrae (only the western exedra has been excavated) that provide focal points at different locations around the pool.
The interior dimensions of the open-air pool measures 43 x 23 meters, and it is two and a half meters deep. Its interior walls and floor are lined with a thick coat of concrete such as is typically used to line Nabataean reservoirs and cisterns. A stone staircase is built into the northeast corner to facilitate entrance into the pool. At the center of the pool is an island-pavilion perched on top of a massive stone plinth (8.5 x 11 x 2.5 m). The pavilion is rectangular in plan with broad, open doorways on at least three sides. Interior columns with marble-faced bases supported the pavilion’s roof. Numerous fragments of painted red stucco, as well as some blue and white, provide some indication of the pavilion’s decorative color scheme. An interior channel or drain cuts diagonally across the floor to connect with a channel that encircles the exterior perimeter of the pavilion, just above the maximum water level of the pool. It is likely that this exterior channel was originally capped and functioned as an overflow channel between pavilion and pool.
Excavations revealed an elaborate water distribution system incorporated into the design and construction of the pool. A reservoir perched on top of the eastern escarpment fed water to the pool and garden below. The pool’s north wall functioned as an aqueduct with channels conducting water across the top of the wall to a castellum auquae, or central holding tank and redistribution point built into the wall. Water then exited the castellum and was fed into stone channels that form part of an irrigation system for the garden that was laid out on the earthen terrace. In addition, two large capped channels cut into the bedrock along the eastern edge of the complex may be intended to direct excess or waste water into the wadi. What appears to be a large subterranean cistern fed and drained by pipelines has been identified through GPR in the southeastern quadrant of the terrace, but remains unexcavated. Additional pipelines of terracotta and lead transported water to designated locations on the site.
The archaeological record indicates that the large terrace upon which the garden was found was constructed during the reign of the Nabataean king, Aretas IV (9 BCE – 40 CE) and thus it is believed that the garden and monumental pool were initially created at the end of the first century BCE or early first century CE The role of the garden and pool complex within Petra in this earliest phase remains ambiguous. The site is closely associated with the neighboring Great Temple complex. However, the original function of the Great Temple monument remains unsettled, with theories ranging from religious/sacred (temple) to private/royal (possibly a palace or audience hall). Close parallels found in the gardens and pools of the contemporary palaces in neighboring Judaea during the reign of the Hasmonean Dynasty (168-40 BCE) and Herod the Great (37-4 BCE) support the interpretation of the Petra garden as part of a palace complex.
The Petra garden continued in use after the Roman annexation of Petra in 106 CE, at which time the Great Temple was converted into an odeion or bouleutereion. It is likely that, within this context, the garden functioned as a public park much like the urban parks in contemporary Rome. Several examples of evidence for renovations have been identified with this phase of the site. For example, the addition of a bridge allowed easy access to the island-pavilion that was originally only accessible by swimming or possibly boating. In addition, a reorganization of the hydraulic system may represent a new approach to water consumption influenced by the Romans, whose rules about the use of natural resources differed from those of local inhabitants.
A deep deposit of sediment above the pool floor indicates that the pool went out of use or was not well-maintained for a period of time before the collapse of the pavilion walls into the pool as a result of the earthquake of 363 CE that caused much destruction in Petra. The remnants of later walls, evidence for the reuse of the hydraulic installations, and a raised field that occupies part of the earthen terrace, testify to the continued use of the site for agricultural purposes through the post-Roman occupation of Petra and into the modern era.
Table 1: Taxa identified from the Petra Garden and Pool Complex. (from Ramsey and Bedal 2015)
End of the 1st century BCE - 2nd/3rd centuries CE (Late Hellenistic/Roman periods)
Alcock, S.E., M.L. Berenfeld, I.B. Straughn, C.A. Tuttle, and (with a ceramic report by Tali Erickson-Gini). 2010. “The Brown University Petra Archaeological Project: Report on the 2009 Exploration Season in the ‘Upper Market.’” Annual of the Department of Antiquities of Jordan 54: 153–166. (http://publication.doa.gov.jo/Publications/ViewChapterPublic/91)
Bedal, L.-A. 1999. “A Paradeisos in Petra: new light on the ‘Lower Market.’” Annual of the Department of Antiquities of Jordan 43: 227–239. (http://publication.doa.gov.jo/Publications/ViewChapterPublic/759)
ibid. 2000. “Paradise Found: Petra’s urban oasis.” Expedition 42 2: 23–36. (https://www.penn.museum/sites/expedition/paradise-found/)
ibid. 2001. “A Pool Complex in Petra’s City Center.” Bulletin of the American Schools of Oriental Research 324: 23–41.(https://www.journals.uchicago.edu/doi/10.2307/1357630)
ibid. 2002. “The Petra Garden Feasibility Study.” American Journal of Archaeology 106 3: 448–450.
ibid. 2002. “Desert Oasis: Water Consumption in the Nabataean Capital.” Near Eastern Archaeology 65 4: 225–234.(https://www.journals.uchicago.edu/doi/10.2307/3210851)
Bedal, L.-A., K.L. Gleason, J.G. Schryver, and (with reports by J. H. Ramsay and J.Bowsher). 2007. “The Petra Garden and Pool Complex, 2003-2005.” Annual of the Department of Antiquities of Jordan 51: 151–176.(http://publication.doa.gov.jo/Publications/ViewChapterPublic/1664)
Bedal, L.-A., L.B. Conyers, J.E. Foss, and K.L. Gleason. 2013. “The Petra Garden and Pool Complex, Ma’an Jordan.” In Sourcebook for Garden Archaeology: Methods, Techniques, Interpretations and Field Examples, edited by, A-A. Malek, 625–641. Berne: Peter Lang/Fondations des parcs et jardins de France.
Kanellopoulos, C. 2003. “The Layout of the Garden and Pool Complex in Petra. A Metrological Analysis.” Annual of the Department of Antiquities of Jordan 47: 149–158. (http://publication.doa.gov.jo/Publications/ViewChapterPublic/723)
Macaulay-Lewis, E. 2006. “Planting pots at Petra: a preliminary study of ollae perforatae at the Petra Garden Pool Complex and at the ‘Great Temple.’” Levant 38: 159–170. (https://www.tandfonline.com/doi/abs/10.1179/lev.2006.38.1.159)
Ramsay, J.H., and L.-A. Bedal. 2015. “Garden Variety Seeds? Botanical Remains from the Petra Garden and Pool Complex.” Vegetation History and Archaeobotany 24/5: 621–634. doi:10.1007/s00334-015-0520-4. (https://link.springer.com/article/10.1007/s00334-015-0520-4)
21 Apr 2021