For the birds — An environmental archaeological analysis of Byzantine pigeon towers at Shivta (Negev Desert, Israel)

Highlights

• Environmental archaeological evidence for arid agriculture in the Negev desert, Israel

• The role of pigeons in agricultural settlements in arid environments in antiquity

• A multidisciplinary examination of agro-archaeological evidence for desert agriculture in the Roman and Byzantine Near East

• A holistic look at pigeon diet and local environmental conditions in the Byzantine Negev

Abstract

Pigeon rearing was an integral part of the agricultural regime that dominated the Negev region in Israel throughout the Roman and Byzantine periods. Dozens of structures have been documented that relate to the raising of pigeons and the exploitation of their dung as a fertilizer as is attested in the literary sources (Pliny, Columella and Varro). Excavation of a dovecote near Shivta produced large quantities of pigeon dung and sediments. The material recovered was processed for floral remains (both macro and micro) and archaeozoological remains. We present a holistic look at pigeon diet and local environmental conditions in the Byzantine Negev through the archaeobotanical remains. Demographic and morphometric analysis of size and shape of the pigeon bones refine our understanding of pigeon species that were bred in antiquity as well as inform on their life-histories. This integrated examination from an agro-archaeological perspective illustrates the complexity of desert agriculture.

Introduction

The site of Shivta is located in the Negev Desert of Israel (Fig. 1). It was, according to current research, founded in the early Roman period (ca. 1st century BCE – 1st century CE) and reached its peak during the Byzantine period (4th – 7th centuries CE). It appears that the Roman military presence in the area of southern Israel and Jordan during the 4th century CE, following an empire-wide economic collapse, likely played a significant role in causing an increase in agricultural production throughout the region (cf. Barker, 2002, Rubin, 1991). Farming was the primary occupation of the inhabitants of the site, who were prosperous enough to build large and complex agricultural installations such as wine presses, aqueducts, cisterns, water pools and dovecotes (Segal, 1983, Negev, 1993, Hirschfeld, 2003), as well as three Byzantine churches.

A good deal of recent scholarship on the Roman–Byzantine period has increasingly focused on the nature and capacity of its agricultural economy in arid environments (Kehoe, 1990, Barker, 1996, Barker, 2002, Barker, 2012, Avni et al., 2009, Avni et al., 2013, Ashkenazi et al., 2012, Erickson-Gini, 2012, Haiman, 2012, Kamash, 2012, Ramsay and Parker, 2016). Understanding the empire's economy is indeed a formidable task as in general there is a lack of surviving documentary evidence that relates to agricultural production and trade. Therefore archaeological investigations of agricultural sites are one of the only ways to elucidate information about arguably the most important economy in the ancient world.

Evidence for regional rain-water harvesting agriculture in southern Israel and Jordan appears to have had a long history and has been documented as early as the Neolithic. It appears to continue in relation to precipitation levels through to the recent past (for example see: Kedar, 1956, Bruins and van der Plicht, 2004, Hunt and el-Rishi, 2010, Barker, 2012, Bruins, 2012). Shivta was supported by sophisticated systems of water collection and irrigation during the Roman-Byzantine period that allowed large-scale agriculture. These included dams, channels cisterns and reservoirs. Early research at the site determined that the agricultural fields coincided with the watershed lines of the basin of the valleys and tributaries (Kedar, 1956). Kedar's (1956) study determined that agriculture was based on a hydrographic system and distribution of rainfall. He noted that all fields depended on episodic rainfall for irrigation as there were no perennial or seasonal water sources in the area. Cereal agriculture was still being carried-out on these systems when Kedar's study was conducted as the water-carrying ditches were still intact and the ancient fields still had flat surfaces. In antiquity these large-scale agricultural field systems were also supported by intensive pigeon-raising and four large dovecotes remains are scattered around them.

According to written accounts, archaeological evidence and artistic representations, pigeon-raising was a widespread livestock industry in the Roman-Byzantine world. Pigeons were raised primarily for the production of fertilizer and as a year-round source of meat (Tepper, 2007a, Tepper, 2007b). All of the dovecote remains were excavated between 2000 and 2004 by Hirschfeld and Tepper (2006; Fig. 2). On a small hill 500 m north of the site a round dovecote was excavated (Fig. 2:1; Fig. 3) and the structure's debris buried under the collapsed stones consisted of large quantities of pigeon (Columba livia) remains, as well as pigeon manure, and various botanical remains. Though numerous dovecots are known, no research has so far been carried out on the remains themselves.

Barker has stated that a major problem with research on desert settlements has been a lack of interdisciplinary research (Barker, 2002, Barker, 2012). Therefore the aim of this paper is to contribute to our understanding of desert agricultural in the Roman-Byzantine period by analyzing the various well preserved floral and faunal materials that are part of the complex agricultural regime in this arid landscape of the Negev and use them to reconstruct pigeon diet, home range, taphonomy, demographic profiles of pigeons and environment (< 100 mm mean annual precipitation) (Fig. 4).

The excavation of the dovecote was carried out in two seasons during 2003–2004 by Hirschfeld and Tepper on behalf of the Institute of archaeology of the Hebrew University, Jerusalem. A round dovecote was revealed with an external diameter of 5.2 m and preserved to a maximum height of 1.65 m. The internal diameter is 4 m divided by Y-shaped walls into three rooms, with open doors at the interior wall (Fig. 5). Two main strata were excavated under the surface layer at the top of the ruined tower. The first stratum was a collapse layer (1.2–1.3 m) of the walls and the second was a pigeon dung layer, that attained a depth of 0.4 m above the dovecote floor (Fig. 5). The debris layer and the dung on the floor probably point to the destruction of the dovecote (Hirschfeld and Tepper, 2006). The pottery retrieved from the second stratum can date the construction and use of the building to the Byzantine period (Hirschfeld and Tepper, 2006: 102, Fig. 25: 10–13). Likewise, a Greek inscription (ΔΟΡΟΝ = present) with an associated cross was found on top of the first strata and supports our assumption of the destruction period of the dovecote during the Byzantine period (Hirschfeld and Tepper, 2006: 97, Fig. 14). Although we cannot be completely certain, all the above evidence, supports the abandonment of the dovecote (with the dung layer above the floor), having been related to a critical event that happened in the middle of the 6th century CE.