An estimate of the time taken by a voyage between trade centers not only depends on the distance but also on the moment of departure, the time of the day that could be used for traveling, the time needed for visiting ports, and the changing weather conditions. Ships heading for Africa south of the Sahara and southwest India had to deal with the monsoon winds, which determined both the departure and arrival time. Those that made return trips between Berenike and the ports along the Somalian coast did not suffer from these cyclic events. The rough estimates of the voyages presented in Figure 5.1 are based on interpretations by Warmington and Casson . Commodities loaded in Ostia or Puteoli could reach Berenike in about 45 days. The voyage from Berenike to Muziris took some 50 days and to Rhapta as many as 90 to 140 days. As ships leaving for Rhapta could not depart from Cape Guardafui earlier than mid-October or the beginning of November because of the prevailing winds, their prolonged voyage in the Red Sea and the Gulf of Aden could have been used for visiting ports along the coast . A return voyage between Berenike and Muziris would have lasted three-quarters of a year, including a stay at Muziris for three up to four months, procona valencia buckets which could have been used for collecting new cargo. A return voyage to Rhapta, on the other hand, would have lasted a year and a half and included a forced stay of about eight months in its destination port.
To trade with ports along the Red Sea coast and the Gulf of Aden must have been an attractive alternative for the Berenike traders, considering the availability of a wide range of commodities brought there by Arab and Indian traders. Such return voyages would have taken not more than two months. The Nile Valley and the Gebel Elba area can be considered as the most important areas where food was obtained. Staple food was accessible within short time periods and with low risk from both locations . Import from the Mediterranean area, on the other hand, took almost the same time span as from India. Such long distance transport, combined with the continuous demand of food supply, must have had implications for the preservation of the food items. On a limited scale, these logistic problems could have been overcome by local cultivation of some food plants.Small-scale cultivation practiced by dwellers of the Eastern Desert, described in Chapter 2, highlights the question whether the inhabitants of Berenike also cultivated plants themselves. This consideration is supported by classical sources, which suggest that agriculture was more respected than trade . Furthermore, the inhabitants of Berenike had lived with the tradition of agriculture, and living on the desert fringe would not automatically have blocked experiments in the field of subsistence farming. The most likely groups of plants that come into consideration for local cultivation are vegetables and some desert trees that produce edible fruits. Fresh vegetables, tubers, and condiments could have been cultivated in kitchen gardens. Several species that were found in Berenike could have been grown in those kitchen gardens: garlic , fenugreek , fennel , coriander , dill , cumin , watermelon , cucumber , and, revalued as a cultivated plant, also beet . This may be especially true for plants from which the surplus value is determined by its fresh condition, such as coriander and dill. Additionally, also pulses such as the faba bean could have been among the garden plants, as is still the case in modern kitchen gardens in the desert.
The presence of several fruit fragments of the caper bush just outside the walls of some buildings indicate that this plant once grew out of the walls, probably beyond the reach of browsers. But its weedy character does not rule out its penetration into kitchen gardens. Just like today, protection against browsing and grazing must have been necessary. Archaeozoological research has revealed that camels, donkeys, horses, sheep, goats, chickens, domestic fowl, cattle, and probably also pigs were locally bred or imported on the hoof from the Nile Valley . The scattered pattern of trenches and the complex architecture within these trenches make it as yet unable to judge possible locations of kitchen gardens.Conclusive evidence for the presence of trees in ancient Berenike, such as plant parts that have no direct economic value, has not yet been found. The only indication of tree growth at Berenike concerns two independent observations of a coconut tree by two seamen somewhere at the end of the nineteenth century, but most likely one or more date palms were mistaken for this exotic palm, as is argued in the section dealing with the coconut. One might suspect that in classical times the date palm was cultivated at Berenike. Dates were appreciated very much by the Romans as can be deduced from the recovery of date palm seeds in many Roman settlements throughout the Empire. In Berenike, the date palm is represented by many different plant parts; not only seeds but also flowers, vessels, and woody parts from the leaves were frequently found. Furthermore, the date palm can withstand brackish conditions to a certain extent, an advantage in this coastal settlement. Although two attractive reasons for planting trees are the shade that they provide and their edible fruits, it should be realized that the local cultivation of trees was not always possible and might have been hampered by the salty condition of the water. The original surface on which Berenike was built appears to be no higher than +2 to +3 m above sea level with the lowest levels at about sea level .
In addition to this relatively shallow, brackish groundwater table, there might also have been a considerable influx of windblown salt crystals into the soil from fl at areas along the sea during low tide, as could be demonstrated by Aldsworth and Barnard . Today, the salt marsh extends several hundred meters inland, as can be judged by the presence of the halophyte Zygophyllum coccineum. Leaving the influence of the possible brackish condition around the site in classical times aside, one might wonder that other trees might have grown in Berenike. Based on the present-day practices and also on the number of sub-fossil plant remains , the following species come into consideration: nabq , fig , and possibly the Egyptian plum , the balsam tree , and Cordia nevillii/sinensis.Because of the limited storage life of vegetables and condiments, local cultivation would have had an advantage over their import. Grains, however, can be stored for longer periods, and for this reason importing them was also a viable option. Cultivating cereals locally would only have been carried out if the logistics of import were too complex, which is not plausible for Berenike, procona buckets which depended on an established trade system between the Nile Valley and the Red Sea. If local grain cultivation was practiced on a reasonable scale, yields would have placed large demands on the water supply. Possible water sources that may have been used in Roman times were the large watering station of Kalalat, 8.5 km southwest of Berenike, and the surface water present in Shenshef, which could have been supplied over a distance of more than 25 km. The most likely area for crop cultivation near Berenike is the western part of the coastal plain, out of reach of the salt spray and sea seepage. It is in this particular part of the plain that Ababda nomads cultivate sorghum today when winter rains have been sufficient. Possible areas for crop cultivation in Shenshef are located near the narrowing of the wadi where surface water is exposed and at the eastern part of the settlement, where a flat elevated area is present, which conspicuously does not contain remnants of buildings. Two corings in the latter location did not reveal remnants of a tilled layer, but it is questionable if changes in the soil texture would have developed in the loose sand. Moreover, the plant cover that developed after the heavy rainfall in October and November 1996was relatively poor and was dominated by Zygophyllum simplex, a species with a wide ecological range with respect to moisture content. The last argument is of particular interest as it is the only one offered that related to edaphic factors. They may be right in this respect, but dung could have been used as a fertilizer, assuming that it was available in sufficient amounts. An argument against local cultivation of cereals is also found in the Periplus Maris Erythraei . Several times it mentions the export of grain from Berenike to harbors along the African and Arabian coasts and even to India, with wheat particularly mentioned for Muza and Kanê , both located on the coast of modern Yemen. Thus, the supply of grain to Berenike was not only meant for local consumption, but also for export. Despite the presence of threshing remains of wheat at Berenike, it is naïve to assume wheat cultivation near Berenike, Kalalat, or Shenshef under unfavorable conditions, even though this type of grain is exported. A more plausible explanation is wheat import from the Nile Valley, part of which was consumed by the Berenike dwellers and part used as an export item.
An archaeobotanical clue to determine the ultimate origin of the cereals may be found in the analysis of the weed flora. Because sorghum, barley, and wheat were already cultivated over a large area for some millennia during the Roman period, the detection of the possible area of origin, either local or from remote areas, is facilitated by the identification of weed species having a more restricted dispersion. For this purpose it is necessary to have samples at one’s disposal in which one of the grain species predominates and also in which some diagnostic weed species are present. When several samples meet this condition, it is possible to determine the associated weed flora of each of the cereals. Unfortunately, samples from all excavation seasons that were rich in both cereal remains and weed species were not characterized by a predominance of one of the grain species. Although the archaeobotanical record contains a reasonable number of weed species, their indicative value is still limited. Fifteen weed species have been found so far that are still growing in the surroundings of Berenike. It is, however, unclear as to what extent they can be considered as weed species from local fields. Another 15 were found whose present geographical distribution in Egypt is especially concentrated in the Mediterranean coastal strip, the Nile Valley, and the oases in the Western Desert . Taking into consideration the trade route between Berenike and Rome, it is most likely that they are indicative of grain supply from the Nile Valley. A possible weed with a more southerly distribution is kodo or kodra millet , represented by some spikelets and a rachis segment. Weighing the arguments, it seems most likely that local cultivation of cereals was only practiced on small plots because yields could not have been in proportion to investments, limitations, and risks. Most probably, local cultivation of cereals was mainly limited to sorghum and barley and had an opportunistic character, as the water supply was critical. Threshing remains of cereals were mainly extracted from dump areas and middens. Surprisingly, threshing remains of barley and wheat were mostly found together with some of their grains. It is conceivable that even in the case of the delivery of so-called clean grain samples, threshing remains may still have been present to some degree. After the final cleaning of these grain samples at Berenike, small amounts of threshing remains could have accumulated in special deposits such as dump areas in the course of time. It is also possible that grain was, at least partly, transported in the ear. This would not have been a problem judging by the capacity of the caravan transport between Berenike and the Nile Valley. Another possibility is that threshing remains were imported as such. It is valued for its use as a filling material, which could have been in great demand in trade centers such as Berenike. It is also used as a fodder, fuel, and temper for building materials such as wall plaster, mortar, and mud bricks.