Horticulture

A reliable interpretation of date consumption based on the number of seeds can also be hampered by the possibility of trading crushed dates. Such dates might have been destoned, as could be demonstrated by the Kellis agricultural account book, which mentions a price of 2.5 times that of ordinary dates . Crushed dates with their seeds still present were found in the Coptic monastery of Phoebammon . Finally, we also have to consider the possibility of the supply of destoned dates and olives. The recovery of secreted plant remains is facilitated by the sampling of concentrations of feces, such as found in cesspits. Judging by the scarcity of water, it is not likely that fl ushed latrines were present in Berenike and Shenshef. During the excavation seasons, relieving oneself is done either in special tents used as lavatories or behind tamarisk hillocks, both situated at isolated spots to diminish the nuisance of fl ies. If this was practiced in a similar way in Roman times, clone rack it implies that only a fraction of fruits such as pomegranate and fig would be retrieved later on by archaeological research from the dump areas in the city proper.The original vegetation in the vicinity of Roman Berenike would not have differed much from the present one.

Climatic change that resulted in modern aridity came about between the First and Fourth Dynasties, in which the period from 2350 to 500 BC, prior to the Persian and Greek rule, was exceptionally dry . In the Eastern Desert, most of the vegetation would have been confined to the wadi branches and the coastal plain. Wadi branches offered occasional runoff water and had a more permanent supply of underground water, and in the coastal plain the vegetation could benefit from seepage and morning dew. Just like today, the sparse herbaceous vegetation had a limited carrying capacity for wild and domestic animals in most years. This picture is supported by the description of the Eastern Desert by Theophrastus and Pliny , who stated that there were no trees except the acacia, which was even sparsely present due to the lack of water. Although the analysis of wood samples from Berenike, Shenshef, and Kalalat does not allow a quantitative interpretation of possible shifts in the arboreal part of the local vegetation, some interesting conclusions can be deduced from the composition of the samples. The charcoal analysis from an early Ptolemaic industrial dump area of brick-making debris revealed that the dominant species was mangrove , in association with a minor proportion of members of the goose foot family . It is most likely, that we are dealing here with Suaeda monoica, the only woody chenopod of a reasonable size that is still present in small populations on the coastal plain. Its massive exploitation for firewood during the earliest habitation period of Berenike can be explained by the clearance of the local area to facilitate the layout of the area meant for habitation and harbor facilities.

This cutting down offered a temporary supply of low-quality firewood, as was evidenced by several fire experiments. This kind of exploitation is supported by the puffing structures in the charcoal of the Goose foot species, which indicates that fresh, living wood was used . Also in Roman contexts, charcoal and even worked wood and wood chips of Avicennia marina are represented in reasonable quantities. A second mangrove species, Rhizophora mucronata Lam., is also evidenced by wood remains , confirming that a mixed mangrove vegetation was present, possibly even at a closer distance to Berenike than that of today. The most dominant tree in Roman contexts that has been used for making charcoal is the acacia. This tree is still highly valued by nomads as its dense wood easily ignites, has a high caloric value, and does not easily fall apart in a charred condition. Judging by the present vegetation, it is most likely that A. tortilis was used as a local source of fuel. Additionally, charcoal from other acacia species might have been imported from the relatively densely wooded Gebel Elba. It remains an open question as to what extent local acacia trees were exploited. If only dead branches were gathered, as is the current practice among the nomads, the exploitation would have had no impact on their proportion in the vegetation. At the same time, such a selective exploitation would not have matched the demand. Harvesting of most of the tree is even possible, as the plant coppices well, so that there is no need for replanting. But if trees are exploited that way, a harvesting schedule based on a 10-year rotation period is necessary . Moreover, a reasonably dense population should have been available to fulfill the demand, a situation that certainly did not exist. Over exploitation of trees would have resulted in their virtual disappearance within a short period of time. In that case, the current trees have to be considered as recolonized specimens, whose spread must have been a gradual process because the mountain ridges are natural barriers for wind dispersal, and the alternative dispersal by camels takes some time.

According to Oedekoven , many wadis in the Eastern Desert and the Sinai still bear names of trees that once grew there but were intensively cut in the distant past when forests in the Nile Valley were cut in favor of agricultural land. The relict character of the sugar date , the twisted acacia , and the leafl ess tamarix in the Eastern Desert is mentioned by Zahran and Willis . The relict nature of these trees is supported by Hobbs , who refers to bedouins who remember former times in which many more trees were present in particular parts of the desert. Large-scale cutting by local tribes is considered the main reason for over exploitation of densely wooded areas, and prolonged drought periods have a negative impact on the tree cover as well.The Diversity of Cultivated Plants. The cultivated plant species that have been found at Berenike and Shenshef can be categorized according to their possible use: cereals, pulses, vegetables, edible fruits, condiments, oil-yielding plants, and plants used for dyeing and tanning . This categorization is based on the proxy value of the recovered plant remains Although many species can be used in medical treatments, this category is left aside as no clear evidence is available as yet for such use. This is not to say, of course, that as the occasion arose, particular plants might have been used in such a way. Five different cereals have been found at Berenike and Shenshef: hulled six-row barley , hard wheat , sorghum , Johnson grass , and rice . Barley, hard wheat, and sorghum can be considered as staple foods. Despite the relatively low numbers of Johnson grass that were recovered from both sites, the correlation between this potential weed and sorghum is weak, so that it is justifiable to consider it as an independent crop. Also the number of rice kernels and rice chaff is limited, but their presence in both early and late habitation layers indicate that this exotic cereal must have been of special importance. The supply of rice might partly have been intended for an Indian community that possibly dwelled at Berenike in the early Roman period, 4×8 tray grow as suggested by substantial quantities of unearthed Indian-made fine wares and a Tamil-Brahmi graffiti . Papyrological evidence, on the other hand indicates that, at least in the early Roman period, rice was also available in the Fayum. Transport of rice from Berenike further northward to Rome seems, therefore, most likely. Both barley and wheat are represented by grains and threshing remains. The threshing remains consist of chaff and sometimes also fragments of the culm. A quantitative comparison between barley and wheat can be done with respect to both the grains and the rachis nodes. The lemma and palea of both cereals are heavily fragmented in most samples, making it almost impossible to use these fragments for a quantitative comparison of barley and wheat. As for rachis fragments as a basis of comparison, it makes sense to quantify the rachis nodes and not the rachis internodes because the spikelets, in which the grains develop, are attached to the rachis nodes. Therefore, the number of rachis nodes is interchangeable with the number of grains. In comparing six-row barley with hard wheat, each rachis node of barley corresponds with three-grain kernels and each rachis node of hard wheat with an average number of three grain kernels that develop within a spikelet. As shown in Table 4.1, the numbers of grains and rachis nodes are not consistent.

The numbers of grains are relatively low with the exception of barley grains from the early habitation period, which even outnumber the rachis nodes. The rachis nodes seem to present a more reliable picture, all the more so because their numbers are much higher. Based on these rachis nodes, the general picture for Berenike is that the proportion of barley to wheat is 0.6 to 0.8. In Shenshef, barley predominates and its proportion to wheat is 1.2 to 1.0. In contrast with barley and wheat, the inflorescence of sorghum is a panicle in which the fertile spikelets are almost sessile. With the exception of a single grain kernel, all the threshing remains of sorghum that were found consisted of lemmas and paleas only. Lemmas and paleas of sorghum are tougher than those of barley and wheat and for this reason are mostly complete specimens. Hence, the quantification of sorghum could be based on the counting of the lemmas and paleas . Sorghum was only available during the second habitation period. The single chaff fragment recovered from an early context is considered as an intrusion. Judging by the absolute numbers of barley and wheat, the highest consumption of these cereals can be ascribed to the first and second centuries AD. In the second habitation period, barley and wheat were partly replaced by sorghum. Although there is some overlap with respect to the dating of fourth to early sixth centuries AD, it seems that the consumption of cereals had diminished in the last part of this habitation period, which is especially true for Berenike. Although cereals might have been locally cultivated on a limited scale, in which sorghum in particular might have produced some decent yields in years with sufficient winter rains, the bulk of these staple grains would have been obtained from the productive Nile Valley. Based on the analysis of the cereals and their accompanying weed flora, it appears that samples from Berenike that resemble those from Shenshef the most originate from trench BE94-1 , trench BE95-10 and trench BE96- 16 . These trenches are located in the dump area in the northwestern part of the site and all samples are dated to the fi fth century and early sixth centuries AD, the period that coincides with the occupation of Shenshef. The group of pulses is represented by the lentil , the white lupin , the faba bean , the bitter vetch , the mung bean , the grass pea , the chickpea , and the Abyssinian pea . Only the lentil, the white lupin, the faba bean, and the bitter vetch are recorded from both Berenike and Shenshef, and their number of plant remains indicates that the lentil and the white lupin were the primary staple. So far, the other four pulses are only found at Berenike and include import from abroad. This is true for the mung bean, which originates from India, and the Abyssinian pea, which originates from Ethiopia. Like the cereals, the pulses that were part of the staple foods also would have been brought in from the Nile Valley. Vegetables are underrepresented at Berenike and Shenshef, as is the case in most archaeobotanical records. This is especially true for green vegetables, which are mostly harvested before seeds are produced, such as lettuce . Seeds from this species have been found only occasionally and might concern seed stock or, more likely, are produced by bolted lettuce. This makes it difficult to decide to what extent such vegetables were available to the inhabitants of Berenike and Shenshef.

Therefore, longer-term research is required to develop non-chemical alternatives, and their adaptation will require effective integration with other methods of disease, pest, and crop management . In the post-methyl bromide era, Verticillium wilt is likely to reemerge as a major disease for conventional strawberry production. The disease already is a major problem in some organic production fields. In strawberry, symptoms begin to appear during early to mid-season, with outer leaves on infected plants turning yellow, drooping, and later turning brown and dry. Yield from these affected plants can be dramatically reduced and infected plants usually die before the end of the season . The fungus survives in the soil as microsclerotia for many years, and survives better in sandy loam soils typical of strawberry production fields in coastal Califtornia than in other types of soil . Large numbers of microsclerotia are formed in colonized Thissue of susceptible crops, and a few are formed even on non-hosts . Whether microsclerotia are formed on infected strawberry plants is not known. Resistance to Verticillium wilt is unavailable in currently used commercial cultivars and tolerance in these cultivars is low. With the phase out of methyl bromide and possible future loss or restrictions on the use of alternative fumigants, resistance to Verticillium wilt has now become a selection criterion in some breeding programs.

As a result, flower harvest buckets resistance to Verticillium wilt in locally adapted strawberry cultivars may increase over time. The concept of rotating crops to manage plant diseases is perhaps one of the oldest cultural practices in agriculture . The utility of this practice in reducing Verticillium dahliae inoculum and subsequent disease intensity has been equivocal . Microsclerotia of V. dahliae survive in the soil up to 10 years, and the extensive host range and lack of host specificity reduce the usefulness of some crop rotations for Verticillium wilt management . However, recent work has shown that rotations with broccoli dramatically reduce microsclerotial numbers and Verticillium wilt incidence in susceptible crops. If rotations of broccoli are successful in strawberry, they will be equally applicable to both conventional and organic strawberry production systems. Although the benefits of rotations are numerous and quantification of these benefits in dollar terms is difficult, simple cost-benefit analysis of adapting rotations will inevitably lead to a better understanding of their composite benefits. Such information also may lead to increased adoption of crop rotations. The objectives of this study were to determine the effect of crop rotation on soil borne fungal inoculum density, disease severity, and strawberry growth and yield; to assess the effectiveness of crop rotation in soil with no detectable Verticillium spp. to improve strawberry growth and yield; and to obtain a cost-benefit analysis of this method of managing Verticillium wilt in strawberry.

Experiments on vegetable–strawberry rotations were conducted at two sites from winter 1997 to fall 2000. One site was located at the United States Department of Agriculture, Agricultural Research Station at Salinas, CA with no history of either strawberry cultivation or Verticillium wilt ; however, the site had other soilborne pathogens such as Pythium spp., Rhizoctonia solani, binucleate Rhizoctonia, and Cylindrocarpon spp. The other site was located at the MonTherey Bay Academy in Watsonville, CA and has a history of strawberry production and disease pressure from V. dahliae as well as the above-noted root rot pathogens. Both sites lie in the major strawberry production areas of the central coast of Califtornia and prior investigations confirmed that black root rot pathogens were present and causing disease . Soil at the Salinas site was a Chualar loam with 57% sand, 31% silt, 1.1% organic matter, and a pH of 5.6. Soil at the Watsonville site was an Elder sandy loam with 62% sand, 26% silt, 0.6% organic matter, and a pH of 6.1. Average yearly precipitation at Salinas and Watsonville was 429 and 757 mm, respectively, and mean daily temperature ranged from 8.3 to 19.9°C and 8.3 to 19.7°C , respectively, during the study. The experimental design was a randomized complete block design with four replications. At both locations, two crops per year of vegetables were grown followed by rotation with annual strawberry. There were three rotation treatments at each location in 1997 and 1999. The crop sequences were broccoli-broccoli-strawberry, lettuce lettuce-strawberry and Brussels sprouts-strawberry or cauliftlower-cauliftlower-strawberry . The first cycle of vegetables, including broccoli , cauliftlower , Brussels sprouts , and lettuce , was grown in January or February and harvested in June in each year. The second cycle of vegetables was planted in July and harvested in September or October.

Because of varying maturity times, planting dates were staggered so that all crops could be harvested and incorporated at the same time. At the Watsonville site, the individual plots consisted of two beds of 7.6 m in length with a single plant line for Brussels sprouts and two seed lines on the 1-m-wide beds standard for lettuce and broccoli production in coastal Califtornia. Plant spacing was approximately 30 cm between plant lines and 20 cm between plants within a plant line for lettuce and Brussels sprouts and 10 cm between plants for broccoli. At the Salinas site, the individual plots were 9 m long and eight 1-m beds wide and consisted of two plant lines for lettuce and broccoli production in coastal Califtornia. The beds contained a single plant line for cauliftlower. Plant spacing was approximately 30 cm between plant lines and 30 cm between plants within a plant line for lettuce and cauliftlower. Plant spacing was 15 cm between plants for broccoli. In both years, 4- to 5-week-old transplants of broccoli, cauliftlower, Brussels sprouts, and lettuce were planted. At the Salinas site, however, lettuce was direct seeded in two rows per bed and thinned to a spacing of 30 cm 3 weeks after emergence. After transplanting, the experimental sites were irrigated with sprinklers for 3 to 4 days to ensure proper establishment of plants. The rotational vegetable planting cycles were timed to include two rotational crops of broccoli, cauliftlower, or lettuce per year and one Brussels sprouts crop. All experiments were maintained using standard cultural practices for coastal Califtornia vegetable production. Preplant fertilizer at 450 kg ha–1 was applied to all beds before transplanting or direct seeding, and all plots were drip-irrigated twice a week. Two side dressings at 67 kg ha–1 followed by fertigation at the rate of 11 to 22 kg ha–1 were provided during each season. At maturity, all marketable vegetables were harvested and remaining residues were flail shredded, air dried on the soil surface for 48 h, and incorporated into the soil to a depth of 15 to 20 cm using a rototiller. Three to four weeks after incorporation, the beds in all plots were reworked for the next vegetable or strawberry production cycle. The second crops of vegetables were grown and incorporated similarly.Standard cultural practices for strawberry production were followed for bed preparation . Beds in the control plots were fumigated with methyl bromide plus chloropicrin at the rate of 450 kg ha–1. Fumigants were injected at a depth of 15 to 25 cm through two hollow shanks, and black or brown high-density polyethylene tarps were used to cover the soil immediately after application. Vegetable rotation plots also were covered with black or brown high-density polyethylene tarps before planting strawberry. Two weeks after fumigation, slow release fertilizer was applied at the rate of 672 kg ha–1 at both sites to all treatments. Fertilizer was band placed about 10 to 15 cm deep in each planting row and covered with approximately 4 cm soil to prevent direct contact with plant roots. Bare-root transplants of strawberry cv. Selva were planted 15 cm apart in all plots at both locations in two rows per bed spaced 40 cm apart in mid November of 1997–98 and 1999–2000 . The drip irrigation system consisted of two drip tapes with emitters spaced 30 cm apart, placed 8 cm from the bed center at a depth of 2 to 4 cm, with an emitter flow rate of 0.87 liter min–1 at 70 kPa. Approximately 30 mm of water per week was applied to each of the experimental plots. After plants were established, round flower buckets drip irrigation was applied once or twice per week for 1.5 to 2 h depending on soil moisture conditions. Hand weeding was done periodically when necessary. All management and harvest of the strawberry crop was done by cooperating commercial growers. To determine the densities of propagules of V. dahliae and Pythium spp., soil samples were collected at the beginning and end of the rotation crop and every month after the start of strawberry production.

Soil samples were taken to a depth of 6 to 10 cm from 10 random sites in each treatment, in the center four beds at the Salinas site and in both beds at the Watsonville site. Samples from each treatment were bulked and placed in paper bags and air dried for 4 weeks on greenhouse benches. The dried soil then was assayed for V. dahliae propagules by direct plating onto improved NP-10 medium using the modified Anderson sampler technique . After 3 weeks of incubation in the dark, the plates were washed in a gentle stream of water and examined under a sThereoscope . The number of microsclerotial colonies on each plate was counted and expressed as microsclerotia per gram of dry soil. For the determination of the number of CFU of Pythium spp., 1 g of air-dried soil was added to 10-ml sterilized water blanks and vortexed to mix the soil, and 500 µl of this soil solution was distributed evenly over a petri dish containing a semiselective Pythium medium . The medium consisted of corn meal agar amended immediately after autoclaving with 0.1% Tween 20 followed by pimaricin , ampicillin , rifampicin , rose bengal , and Benomyl 50WP after the medium had cooled to 50°C . After 24 h of incubation at 25°C, the surfaces of the plates were washed free of soil under a gentle stream of water and the number of colonies counted. Plates were placed back into the incubator and subsequent counts made again after 24 h. The total colony counts were expressed as CFU g–1 of dry soil and the presented data reflects the mean of the three replicates. Prior investigations at these sites revealed that Pythium ultimum was the most commonly encounThered species, but other species such as P. irregulare were recovered as well .To determine the relative effects of different rotation treatments, strawberry plant growth was monitored by recording the plant canopy diameter of 20 arbitrarily chosen plants per replication at least twice per season in each replication. Measurements were made in both the east-west and north-south directions because there were significant directional effects early in the season. Development of Verticillium wilt at Watsonville was assessed based on foliar symptoms . The symptoms of root diseases such as black root rot complex were similar but rarely progressed past stunting with leaf chlorosis later in the season . Assessments were made at 2-week intervals starting from June until the end of the season . No Verticillium wilt was observed at the Salinas site. Twenty plants per plot were visually rated to monitor disease progress. The disease severity estimate was made based on a scale of 1 to 8, where 1 = healthy plant, 2 = moderately stunted, 3 = moderately stunted and slight outer rosette of dead leaves, 4 = moderately stunted and moderate outer rosette of dead leaves, 5 = significantly stunted and slight outer rosette of dead leaves, 6 = significantly stunted and moderate outer rosette of dead leaves, 7 = significantly stunted and significant rosette of dead leaves, 8 = dead plant. Data on the cost of production of strawberry fruit, including fumigating, preparing of land, mulching, planting, irrigating, fertigating, weeding, and fruit harvesting, were obtained from four growers that produce both strawberry and vegetables. Similarly, total yields per unit area of production, corresponding revenue earned, and so on also were obtained. Similar data were obtained for strawberry production without fumigation. The cost of producing two crops of broccoli, revenue from broccoli, and the following strawberry crop also was obtained.

In fact, such an analysis could illustrate a shift in the proportion of smaller to larger seeds, demonstrating an increase of out breeding. And even if such an increase could be demonstrated, it still remains to be seen whether the increase of out breeding, which is responsible for the increase in seed size, could be considered as the result of artificial pollination or of a more favorable proportion of male palms to female ones in a population. The interpretation of the variation in seed size may be further obscured by the presence of palms that have been propagated by seeds. Both seeds and basal suckers are used for propagating the date palms, as is, for example, mentioned by Theophrastus . Preference is given to the basal suckers, as they guarantee both the sex of the tree and the quality of the dates . In contrast, a seed will either produce a male or a female plant and because cross-fertilization has occurred, an unpredictable genetic variability will be produced. The combination of outbreeding pollination and vegetative propagation has resulted in many local varieties and just as many date varieties. The size of the seeds from Berenike proved to be quite variable. Seeds that were retrieved from the 1995 season measured 2.27 cm . This indicates that the dates originated from groves that were at least partly propagated by seeds.

The smallest seeds possibly even originated from unfertilized flowers. Trees developing from discarded seeds appear to be half female and half male specimens. If plants are intentionally propagated by seed, plastic growers pots several seeds might be put together to ensure the appearance of female plants. According to Täckholm and Drar , this is especially practiced in the Aswân province. It has also been observed by the author in the southern part of the Eastern Desert. Theophrastus gives another explanation for planting several seeds together. Both roots and stems will become entangled and in this way prevent the development of a weak tree. Theophrastus does not take into consideration the dioecy of the date palm, although he is aware of this character. A concrete building near the old settlement of Berenike was established, which is used as a working space, a temporary stockroom for archaeological finds, and it also offers accommodation to a museum in which a standing ethnographic collection of the Ababda nomads is displayed. After the building was finished, an experiment followed to grow some date palms. Permanent guards were to take care of the water supply. Unfortunately the young trees died, probably because the soil was over manured and not watered enough. That dates were highly prized by the Romans is evidenced by the number of dates that have been found at Berenike and Shenshef. The sugar content of ripe dates ranges from 70 to 80 percent, making it a tasty and nutritious food. It also lends itself well to preservation, a favorable condition in a desert environment. Spread over 284 samples, some 2,700 seeds of date palm have been found.

As this amount of seeds only represents the proportion that has been sub-sampled from a small area of both Roman settlements, it is clear that dates have to be considered as one of the staple foods. It is even possible that dates were partly used as a ballast commodity. From more recent periods, it is known that dates were used as ballast by ships on their way to India, Africa, and southeast Africa . Judging by the huge amount of dates found at Berenike and Shenshef, it seems likely that most of these fruits originated from the Nile Valley. It may not be excluded, however, that groves in the Eastern Desert also were exploited in an opportunistic way. In this respect the substantial grove in the Wadi Gimal estuary may have been visited, as it is located along the coastal route that connected Berenike with Quseir al-Qadim. Furthermore, import from northern Sudan may not be ruled out. Despite the huge production of dates in Egypt today, which are partly meant for export, Egypt also imports dates from northern Sudan . Only a fraction of the recovered sub-fossil seeds are charred. They were not concentrated or found near ovens or fireplaces, ruling out the possible use of date seeds as charcoal. Their use as fuel could be demonstrated in a monastery at Kom el-Nana in the Nile Valley, where charred date rachilla and perianth were found in an oven . The use of date seeds for making charcoal is recorded from Iraq, were they are in demand by silversmiths . It is noteworthy to mention that this practice had already been described by Strabo for the bronzesmiths of this area.The emblic is a member of the spurge family . The Latin name Phyllanthus, which means “leaf-flower,” is in fact misleading because the sessile flowers, and at a later stage the fruits, are not attached to the stalk of a pinnate leaf, but to a branch that bears many small leaves that are arranged in two rows. Wild fruits have a diameter of 1.5 to 2.5 cm; cultivated ones are larger.

The fruit is classified as a drupe. The thick mesocarp is edible. The hard endocarp consists of six carpels that form three compartments, each of which contains two seeds. A total of eight endocarp fragments were found in a trash dump of Berenike. The genus Phyllanthus comprises some 500 tropical and subtropical species. Three species of this genus are native to Egypt: P. reticulatus Poir, P. maderaspatensis L., and P. rotundifolius Willd. All three species grow in the Gebel Elba area and the last one also in other parts of Egypt . The fruit fragments found at Berenike are similar to those of the edible fruits of the emblic , which are much larger than those of the Egyptian species of the genus Phyllanthus. Emblic grows in the Maskarenen, India, China, Indochina, and Malaysia. It is common in most parts of India, extending from the base of the Himalayas to Sri Lanka. It is especially abundant in deciduous forests of Madhya Pradesh and Uttar, located in central India . As a sacred tree for the Hindus, emblic is associated with Vishnu, Shiva, Parvati, and Lakshmi . For this reason, it is also planted on the south side of a temple or home. But most fruits of emblic are today gathered from wild populations. Another edible species is P. acidus Skeels. Although this species is cultivated in gardens today throughout India, it is considered to be native to the coastal region of northeast Brazil and has erroneously been ascribed to Indian, Madagascar, and Malaysian origin . The fruits of emblic belong to the ones with the highest content of vitamin C: 100 g of juice contains 600 to 1,300 mg vitamin C . A similar amount of orange juice, for example, only contains 35 to 50 mg vitamin C. The extremely high vitamin C content is responsible for its acidic, astringent, and somewhat acrid taste and is expressed in its Indian name amlika, which means “acid, sour.” Fresh fruits are almost too sour to eat raw but can be made into pickles, preserves, and candy. Pickling, however, results in the loss of much of the vitamin C, although most can be retained by boiling the fruits and adding a large amount of salt . Due to the presence of tannin, blueberry in pot the oxidation of vitamin C is prevented or retarded. Probably, the consumption of emblic during the long sea voyages protected sailors against scurvy. Scurvy was successfully treated in the Indian army in Rajputana in 1837 with fruits of the emblic . Unripe fruits, bark, and leaves of the emblic are used for tanning. The pulp of unripe fruits may contain up to 35 percent tannin, which is more than the concentration in the other parts of the tree. The fruits of the emblic are often used as a tanning material in combination with fruits of two other so-called myrobalans: beleric myrobalan and black myrobalan . All three species have the same distribution area. In India this mixture of fruits is known as triphala or tinepala, which means “three kinds” . Triphala is also a well-known medicinal preparation of the Ayurvedic system, which developed during the Vedic period, about 500 BC . In present-day bazaars with a good assortment of herbal drugs, including ones in Khan al-Khalili in Cairo, emblic is still offered for sale. It is traded as dried fruit, known as Myrobalani emblicae.

A whole dried fruit consists of 40 percent gray-blackish mesocarp, 40 percent endocarp, and 20 percent seeds. As only the mesocarp of the fruit is of economic importance, Myrobalani emblicae samples predominantly consist of mesocarp particles, endocarp fragments being only a minor contamination. In a sample of 100 g, obtained from an India-Pakistan shop in Amsterdam, only seven carpels were found corresponding with 1 1/6 fruit. The total number of fruit specimens of that sample was estimated at 70. Assuming that the emblic at Berenike was also traded with the endocarps and seeds removed, this implies that the eight carpels found stand for a more substantial number of fruit specimens. Today, perishable emblic fruits are also sold in tins or pickled in glass jars, in which most specimens are still intact. So far, emblic is only known from Berenike and four other archaeological sites in India: Neolithic Budihal, Chalcolithic Inamgaon and Navdatoli, and from Narhan in historical times .Many pine species are of economic importance, especially valued for their wood, which gives good timber, and their resin, which can be used as an adhesive, as an ingredient of perfumes, and was formerly used in mummifi cation. Only a few species yield edible seeds rich in proteins and oil, including the stone pine with its characteristic, umbrellashaped crown. In fact, the quality of the wood from this particular pine is poor and was therefore not used much in antiquity . The stone pine occurs in the northern and eastern coastal part of the Mediterranean area, extending into Syria, and the southern coastal part of the Black Sea. In Egypt, only a few specimens are present in gardens . Mature cones of the stone pine are about 10 to 15 cm long and 10 cm across. On the upper side of each cone scale, two unwinged seeds develop. One cone consists of about 100 scales so that it produces some 200 seeds. Cones can be harvested when they are still green and have to be exposed in the sun for further ripening. In this way the yield is maximized since no seeds are lost during harvesting. Alternatively, seeds can be gathered easily under the trees because the unwinged seeds are not dispersed over long distances. Most archaeological finds of the stone pine date back to the Roman period and are frequently reported from countries where the tree is not indigenous, such as Egypt and England . Kislev points to the religious significance of the tree on the evidence of archaeological contexts and literary sources. In Egypt, as well as in England, whole cones and scales have been found in connection with temples. It is possible that at Berenike the stone pine was also used in religious ceremonies. However, in trenches BE95-4 and 7 in a putative temple area and BE95-6 in which religious effects were found , no remnants of this species were recovered. Besides its religious use, whole cones or seeds of the stone pine also may have been traded for consumption. The nuts were not only used as garnish or filling, but also as a substitute for the expensive condiment asafoetida, obtained from the roots of the Persian umbellifer Ferula assa-foetida L. The Romans used asafoetida as a substitute for the highly valued North African plant laserpitium , after laserpitium became extinct in the second half of the first century AD due to intensive gathering. In order to make use of the desirable taste on a larger scale, Apicius recommended mixing some asafoetida with pine seeds so that the latter could absorb the flavor of asafoetida . The Roman predilection for the taste of laserpitium and asafoetida could explain the concentration of pine nuts from archaeological contexts dated to the Roman period. After the fall of the Roman Empire, it seems that pine nuts ceased to be used as a substitute for these condiments. Today, shelled seeds of the stone pine are offered for sale as a rather expensive delicacy.

To determine whether archaeobotanical remains of coriander were used for such treatments is in most cases impossible, as the archaeological context does not support such interpretations. An exceptional case is, however, a find of coriander fruits in a stockroom of a Roman military hospital in Neuss . According to Dioscorides, the consumption of the fruits positively affects one’s intellectual capabilities. Rutten suggests that fruits infected with the fungus Puccinia petroselini might cause this. In Egypt, the fruits are still sold for medical purposes. Pounded fruits are used for vertigo and as a carminative and tranquilizer . An aberrant use of coriander concerns adding fruits to incense mixtures, as was observed in many incense samples sold today at Egyptian spice markets. It is not known if this was also practiced in antiquity. In any cases, the specimens from Berenike do not support this kind of use. The only charred fruit was found together with 64 desiccated specimens in a trash-dump sample that originated from trench 13.Hazel is an Euro-Siberian species that is native to the Caucasus, west Asia, and Europe. In the Mediterranean area, this tree only grows on the mountains,large plastic garden pots such as on Mount Tmolos and the Mysian Mount Olympus, as mentioned by Theophrastus . It does not grow in Egypt.

The nuts of this tree thus belong to the food items that have been brought from the Mediterranean area to Berenike and Shenshef. The number of sub-fossil hazels that is recorded from Egypt is limited and, furthermore, confined to the Greco-Roman period. A much earlier record, dated to the predynastic period , concerns pollen that was found at Tell Ibrahim Awad in the Nile Delta . This location is outside its distribution area. Moreover, hazel is almost confined to the cool mountain slopes in its southern distribution area, like the situation in Crete. Its presence in Tell Ibrahim Awad, just like that in sediments near Carthage, can be explained as a result of long-distance transport, probably from substantial hazel populations in Italy and Turkey. Only fairly large pollen values or an increase in the hazel curve may indicate that the plant was locally cultivated or was part of the natural vegetation .Based on the distribution of its possible wild relative Cucumis sativus ssp. hardwickii Royle in the foothills of the Himalayas and other parts of India and in Arabia, it is widely assumed that the cucumber was taken into cultivation in India . If this is true, then its domestication must have been started earlier than 3000 BC, judging by the archaeobotanical records of cucumber from Sharh-i-Sokhta in Iran .

Cucumber was cultivated by the Greeks and Romans, who were also responsible for its further spread over the Roman Empire, as is evidenced by sub-fossil remains from Germany, Belgium, France, and Great Britain. The find of cucumber fruits in Deir el Medineh, which are dated to the Eighteenth Dynasty , indicates that the fruit was introduced in Egypt much earlier. Its identification on the basis of fruits only is, however, rather dubious as fruits of the cucumber are extremely irregular in shape. Theophrastus already describes three different forms, which are also referred to by Pliny : the Laconian , the cudgel-shaped , and the Boeotian. Pliny also mentions a quince-shaped fruit that was newly cultivated in Campania. Although the other Egyptian records of cucumber are small in number, they are more reliable as their identification is based on seed morphology and are all dated to the Greco-Roman period. The juicy cucumbers must have been a highly appreciated food in Berenike and Shenshef. They might have been locally cultivated in kitchen gardens, although a more substantial supply would have been imported from the Nile Valley. Cucumbers were not only eaten, but also used in other ways. According to Pliny , myrrh was adulterated with cucumber juice to give it a bitter taste. As cucumber has an insipid, watery taste, which is in fact mentioned by both Theophrastus and Pliny , it is more likely that the bitter gourd is meant. Pliny also mentions a variety of medical uses of pounded cucumber seeds, partly in conjunction with seeds of cumin . Cucumbers were also used in the production of cosmetics, such as a depilatory cream that consisted of boiled, crushed bones of a bird, fl y dung, sycamore juice , gum, and cucumbers . Today, seeds of cucumber are still sold in spice markets in Egypt and are recommended for the treatment of diabetes as well as for the treatment of the swelling of the colon and as an antiseptic .

Cumin is a desert plant that grows in oases but cannot withstand severe dry heat or heavy rainfall . The plant produces many fruits, which contain 2 to 4 percent essential oil. The fi nely sulcate leaves, on the other hand, are of no special value. Cumin is, therefore, one of the kitchen herbs that is exclusively cultivated for its fruits. Only a few sub-fossil remains of cumin have been identified so far. They are recorded from the second millennium BC Tell ed-Der in Iraq , Iron Age Deir Alla in Jordan , and, in addition to Roman Berenike and Shenshef, from three other sites in Egypt, namely, pharaonic Deir el Medineh , Roman Mons Claudianus , and Islamic Kom el-Nana . Furthermore, a dubious identification of an imprint is recorded by Soderstrom from pre-Islamic Hajar Bin Humeid in south Yemen. Remarkably, the distribution of the sub-fossil remains does not coincide with that of the wild forms of cumin, which grow in central Asia along the river Kisilikun in Turkestan, considered to be the area of origin of this spice . This discrepancy can be explained by the backlog of archaeobotanical research in central Asia and the poor preservation conditions of the charred etheric fruits. Pliny mentions both cultivated and wild cumin, the latter one distinguished by its slender habitus and four or five serrated leaves. This description of the leaves, however, indicates that we are not dealing with a close relative of the cultivated cumin. The common name cumin is derived from the Greek kuminion, which in turn probably originates from the old Babylonian ka-mu-nu . As a common name, cumin is rather confusing as it is used for a variety of species, including the following ones that originate from the Old World and represent no less than three different plant families: Nigella sativa L. , Allium spp. , Cuminum cyminum L. , Pimpinella anisum L. , Trachyspermum ammi Sprague , and Carum carvi L. . Obviously, all these species have been evidenced for the Roman period from archaeobotanical remains. The fruits of cumin found at Berenike and Shenshef might either have been imported from the Nile Valley or have been harvested from locally cultivated plants. They could have been used for flavoring food as well as for all kinds of medical purposes. According to Pliny ,raspberry plant pot cumin was one of the most agreeable seasonings. Today, this kitchen herb is largely replaced by caraway . Nevertheless, fruits of cumin are still offered for sale in Egyptian spice markets and are recommended for the treatment of intestinal spasms, rheumatoid arthritis, and kidney stones .In this explanation, only the description of the grape is in concordance with current anatomical descriptions. Both the sycamore fig and the common fig can only be successfully pollinated by the symbiotic wasp Ceratosolen arabicus Mayr. Unfortunately, this wasp has probably not been present in Egypt since the early Holocene . Fruit production is, however, still possible by the development of fruits that have not been pollinated, a process that is called parthenocarpy. Although in both species specific forms exist that are parthenocarpic, in Egypt the only parthenocarpic forms found are those of the common fig . The production of accessory fruits without the small fruits by the sycamore fig in Egypt was already mentioned by Theophrastus . Therefore we can reliably say that the cultivation of the sycamore fig in Egypt, which must have already started in predynastic times and is nowadays restricted to the lower Nile Valley, has always been propagated by cutting. The presence of the fruits at Berenike and Shenshef indicate, therefore, that we are dealing with the common fig . Most likely, these figs were cultivated in the Nile Valley, and we must consider them parthenocarpic plants. The recovery of the fig in many excavations is not only facilitated by its suitable preservation properties and long-distance transport, but also by taphonomic processes.

The archaeobotanical remains of the fig mainly concern the small fruits, which are usually called “seeds.” These small fruits are swallowed and pass through the intestinal canal largely undamaged. Especially in archaeological feature types where dung is concentrated, such as cesspits, conditions for recovery are favorable. For two reasons, the number of figs found at Berenike and Shenshef is relatively low. In order to be able to process large quantities of samples, it was decided to sieve only a small number also through a 0.5 mm sieve, which is capable of retrieving the fruits of fig. Furthermore, trash layers that have been investigated so far are located in the built-up areas and it may be assumed that such locations were not considered as suitable latrines. During the excavations, in order to minimize the serious annoyance caused by flies, relieving oneself was facilitated by special latrines or one could go behind the high hillocks of the Tamarix nilotica, both situated at some distance from the campsite. Although advanced flushing latrines became a standard accessory of large settlements from the beginning of the Roman period onward, it seems highly unlikely that they were also constructed at Berenike or Shenshef due to the scarcity of water. It may be assumed, therefore, that in Roman times, too, most of the human dung was deposited far beyond the inhabited area. Due to the high sugar content of the fleshy part, the fig can easily be preserved by drying. In their dried state, figs can be stored for a considerable period and transported over large distances. These properties made possible the easy availability of the fig throughout the Roman Empire, including the northern part, which is outside its distribution area.This kitchen herb is very similar to dill , and Linné even classified fennel in the same genus as dill . What is characteristic of fennel are its life span and the fruit morphology, the latter being advantageous for archaeobotanical research. Fennel is a perennial herb and has oblong, channeled fruits. Dill, on the other hand, is an annual and has flatter fruits, although intermediate forms do exist. Jansen , who by the way argues for using its original Latin name, mentions that seed samples offered in Ethiopian markets mostly consist of a mixture of fennel, dill, and cumin . Seed samples that were obtained by the author in spice markets of Egypt turned out to be pure. Fennel and/or dill fruits have been recorded from Roman Italy, Switzerland, Germany, France, The Netherlands, Great Britain, Libya, and Egypt. Obviously, fruits of dill were found in more than twice as many sites than fennel. In a number of sites both herbs were found. This pattern is also reflected in the sparse sub-fossil record of Egypt. The archaeobotanical evidence of both species supports to some extent the reliability of classical sources that sometimes also list both species together, such as Theophrastus . On the other hand, it is rather conspicuous that in describing the hair-like leaves of a particular plant, both Theophrastus and Pliny only present fennel as an example. Fennel has been cultivated for its roots, leaves, and fruits, the most valued being the aromatic fruits that contain 2 to 6 percent essential oils. Those from the first-produced umbels are the best. To prevent loss of the essential oil, the fruits should be dried in the shade . Besides its use for flavoring all kinds of food, in particular fish courses, it has also been used as a medicine. Because fennel is also cultivated for its fruit, it should have a good chance of recovery in archaeobotanical research. Nevertheless, the sub-fossil remains are sparse. This is not only true for Berenike, where only three fruits have been found so far, but also for other settlements in Egypt.

Although it may not be excluded, it seems unlikely that A. tortilis once grew on the site proper. This tree penetrates the coastal plain via the main wadi branches but is not found today in the near vicinity of Berenike. Both its salt non-tolerant status and its deep, penetrating root may prevent successful establishment on this limestone outcrop in a salt-marsh area. More to the south, however, a large population of A. tortilis is present in the coastal plain area. Such a population of trees would certainly have been more accessible than the more scattered specimens in the middle and upstream part of the wadis. Another possibility that might explain its presence at Berenike is that we are dealing with plant remains that may have partly entered the trash deposits by natural agents. The use of large, spiny branches for fencing off kitchen gardens to protect them against grazing animals, as has been observed in Marsa Alam, 140 km north of Berenike, probably offers the most plausible explanation. Such branches will have disintegrated in the course of time and will have supplemented the soil with their diagnostic parts.

A similar use has been described by Wood for dead branches of nabq ,vertical grow which are used to make thorn bush hedges in Yemen.One seed from a dump area belongs to the baobab. With its unusually thick trunk and slender branches, giving the impression that these above ground parts are out of proportion, this tree has a characteristic appearance. The baobab has its natural habitat in the dry woodland savanna south of the Sahara, where it is frequently associated with tamarind , a tree that has also been attested for Berenike by its seeds. Two baobab trees are also recorded from two different localities in north Yemen . Whether these are relics of a larger native population or have to be considered as introductions is not clear. The current distribution of the baobab in India is confined to the northern part, including the northwest coast along the Gulf of Cambay, where it grows in and around the old ports of Janjira, Chaul, and Surat . According to Burton-Page, the tree has to be considered as an early fifteenth century introduction. This rules out the possibility of an import from the Indian continent. On the other hand, the presence of a baobab seed in Berenike clearly demonstrates that it was traded in much earlier times and that an earlier introduction into India cannot be ruled out. If the seed unearthed from Berenike originates from the East African savanna, the most probable area of origin is northeast Sudan, from where it could have been offered for trade in Ptolemais Thêrôn, a port located at lat 18°40′ N and which was only accessible by small vessels .

As the baobab is almost not recorded from the Horn of Africa, a second possibility is that it originates from east tropical Africa, from where it could have been exported from Rhapta , according to the Periplus Maris Erythraei the only port of trade on the East African coast south of Opone. Because a return voyage from Berenike to Rhapta would take about a year and a half, its seems more likely that trade items from this area were subjected to distributive trade and offered for sale in the so-called “far side” ports in Somalia. This might also have been the case for the baobab. On the other hand, the uncertain status of baobab in north Yemen does not exclude the possibility that the baobab was imported from a far less distant place. The baobab is a highly valuable tree because all parts can be used in a variety of ways. The hard, indehiscent fruits of the baobab measure 15 to 20 cm in length and contain a great number of seeds imbedded in pulp. The seeds can be eaten fresh or roasted and contain up to 15 percent oil. The fruit pulp contains tartaric acid, which is also edible and can be mixed with water or milk. Additionally, both seeds and fruits have medicinal properties . Today, whole fruits, mostly sprayed with fancy colors, are offered for sale in fl orist shops and are used for making bouquets of dried flowers.Bulb remains of garlic have been evidenced from the Second Intermediate period until the Islamic period. In Berenike, a reasonable number of bulb bases and bulb scales from garlic were found in different trash dumps, giving the impression that garlic was cultivated locally and consumed in reasonably large quantities. Both garlic and onion are well represented in the archaeobotanical record of Egypt.

Within the spectrum of vegetables and herbs, both species belong to a small group of species that have a good chance of ending up in an archaeobotanical record. Edible parts from vegetables and herbs such as lettuce , thyme , and mint have a very small chance of being discovered, even in Egypt where preservation conditions are excellent, due to the extreme arid climate. If they are found at all, it is their seeds that are unearthed. This, in turn, is exceptional because these plants are normally harvested before seed setting. In fact, the few seeds that have been found can be considered as those meant for sowing or as seeds that originate from plants that have bolted, which makes more sense. Other vegetables and herbs, such as saffron, the orange-red stigmas of the crocus , are rare because they concern plant parts that cannot be propagated. When the edible parts of plants are also used for propagation, they are adapted to living in the soil, which favors a good preservation. This is the case with bulbs of garlic and onion and with the seeds and fruits of fennel , coriander , and dill . Almond is one of the fruits that has been imported to Berenike from the Mediterranean area. Although no large quantities were found, fragments are present in many samples and suggest that this fruit was available on a regular scale. In addition to the records from the Greco-Roman and Byzantine periods, almond fruits as well as walking-stick handles made from the almond tree are recorded from several archaeological contexts dated to the Eighteenth Dynasty . It is a point of discussion whether the almond was cultivated in Egypt during this period. Germer , for example, is of the opinion that the tree was once cultivated in Egypt because the small pot found in Tutankhamen’s tomb not only contained fragments of the endocarp , but also some whole fruits. Apricot has its origin in eastern Asia and was introduced in the Near East around the first century BC. From there it was introduced into the Mediterranean Basin where it became a well-established horticultural element . Archaeobotanical records of this perishable fruit are extremely rare. In fact, the only finds from within the Roman Empire are recorded from Egypt. One of them originated from the sacred animal complexes near Saqqara. The building of these cemeteries was started in the Nineteenth Dynasty and they remained in use in the Ptolemaic period. As is suggested by Germer , the kernel unearthed must have come from the Ptolemaic period. A second kernel is recorded by Thanheiser from Kellis in the Dakhla Oasis and is dated to the Roman period . In Berenike, two fragments of apricot were found in trash layers in trench 10. The most likely origin is the Mediterranean area,rolling grow table although it may also have been cultivated in northerly-located Fayum. The presence of a perishable fruit in such a remote corner of the Roman Empire clearly refl ects the luxury status of the food supply, though it may not have been available on a regular scale.In Egypt, mangrove vegetation consists predominantly of the Avicennia marina species and is restricted to the Red Sea coast, especially south of lat 25° N. This kind of vegetation is for the most part confined to the tropics and is indicative of muddy tidal waters. In the delta of Wadi Gimal, some 100 km north of Berenike, this plant is partly covered by sand hillocks, as a result of the silting of the shoreline zone. Apparently, A. marina can withstand the absence of nutritious mud to some extent. Along the Egyptian Red Sea this plant has mostly a shrubby habitus. It is possible that, in antiquity, the mangrove vegetation was more widespread. According to Strabo , mangrove vegetation was present along the whole coast of the Red Sea. The comparison of the mangrove tree with the olive tree and the laurel is, however, misinterpreted by Strabo and Pliny . Both writers are talking about olives when they describe the mangrove vegetation.

This exploitation of A. marina has had a long tradition. In the past, leaves were used as camel fodder during the summer, and shepherds ate the soft, green fruits . This kind of exploitation would certainly not have endangered this plant species. According to Mandaville the leaves are merely second choice because their salty taste. Additionally, Drar mentions of the use of the mangrove as fuel, resulting in the destruction of the mangrove vegetation, especially between Quseir and Wadi Gimal. Also Schweinfurth had already mentioned the use of A. marina as firewood, pointing to the high quality of the charcoal because it remains hot for a considerable period of time. Additionally, Schweinfurth reports the suitability of the tough and branching stems for constructing houses and fences. Today, this kind of exploitation is probably nonexistent. Ababda nomads still use branches of the twisted acacia in combination with mats or driftwood for making their shelters, although more and more concrete houses are becoming the norm. In the Gebel Elba area among Bisharin nomads, it was observed that wooden houses are built from a variety of woody species that grow in that particular environment, including arched branches of Cocculus pendulus. Furthermore, A. marina can also be used for tanning purposes. Exploitation of the tree for this purpose has resulted in the destruction of several mangrove forests . The presence of both leaves and wood of A. marina indicates that whole branches were harvested. Huge quantities of charcoal and some leaves of A. marina, intermixed with a minor fraction of Suaeda sp, were found in trench BE96-11, located in an industrial area southwest of the central town of Berenike . Most probably, this early Ptolemaic dump can be related to the cutting down of mangrove vegetation in the near vicinity of Berenike. Strabo, who visited Egypt just after the annexation of Egypt by the Romans, mentions that there were some convenient landing places for ships in Berenike . This indicates that at that time the bay was not silted up and, therefore, would have been too deep for the mangrove vegetation. Leaves used as fodder would have been a welcome by-product in the cutting down of the mangrove vegetation. It is also possible that mangrove wood was exported to Arabia, where it is still used for various kinds of construction, including houses, and as firewood. The mangrove was once the main cargo of specially constructed dhows that exported it from the East African coast . The only other archaeobotanical record of A. marina originates from Abu Sha’ar, a Roman settlement along the Red Sea coast 20 km north of Hurghada . Large quantities of leaves have been found here.Although some of its English names suggest membership of the palm family, the sugar date or desert date in fact belongs to the Balanitaceae, of which it is the only representative in Egypt. The shape of sugar dates is quite irregular and resembles that of the real date. But this resemblance does not hold for the anatomy of the fruit or the morphology of the seed. The inner layers of the fruit are woody and resistant to decay. The outer soft part of the fruit is not present anymore in sub-fossil specimens. The fruit pulp has a high sugar content of about 45 percent, while the relatively small seed is quite soft and yields 40 to 60 percent balanos oil. Balanos oil was used for the production of all kinds of perfumes. Like alcohol, fats and oils are capable of absorbing and retaining odors.

Only isolated specimens in inaccessible spots of rocky affluents succeed in flowering and fruit setting in such years. Other grasses remain small in size, even in favorable years, and it seems as if these grasses are present at the fringe of their distribution area. A good example is the presence of dwarf specimens installations, only one could be assigned to this variety. Several specimens were found between stones in a small affluent, which are relatively large in comparison with the variety typicum. The variety ehrenbergii Henr. is characterized by its scabrous lemmas, which are covered by thick hyaline-curved hairs. This variety is not mentioned for Egypt. According to Henrard , its distribution area stretches from Ethiopia and Eritrea to Arabia and Iran. In the more southerly Gebel Elba area, which is characterized by the presence of the Sudanese flora, in addition to the variety typicum, only two other varieties have been recorded, viz. pumila and aethiopica . Thus the presence of the variety ehrenbergii in Shenshef clearly points to a connection with more-southerly-located areas. One may wonder if the presence of this plant in the vicinity of Berenike can be related to the former trade contacts with Arabia, Ethiopia, and Eritrea.

The investigated localities in the Eastern Desert have many species in common, especially when they are connected by the same wadi system. In Khesm Umm Kabu and Qariya Mustafa ‘Amr Gama,plastic seedling pots which are located about 115 km apart, mutual species are rare, namely, Tamarix aphylla, Salvadora persica, and Trichodesma africanum. The number of unique species for a particular site is relatively small. Examples are Ifl oga spicata and Rumex simpliciflorus near Khesm Umm Kabu in the upstream part of Wadi Gimal. Obviously, both species were also found in the delta of this wadi and this shows that the dispersal within a specific wadi is not problematic.The high percentage of species that is present in more than one locality seems to indicate that seed dispersal, which is mainly dependent on wind and animals, is quite effective. Additionally, strong water currents after heavy rainfall transport a lot of sediment including the existing seed bank. Dispersal by the agency of water is, however, always restricted to the specific catchment area of a wadi system. Moreover, seeds are transported in either the direction of the Nile Valley or the Red Sea coast. The effectiveness of this kind of dispersal can be observed, for example, in drift litter along the Red Sea coast. Fruits of Zilla spinosa, Balanites aegyptiaca, and Neurada procumbens, for example, were frequently found in such concentrated organic debris.

Due to wind and animals, seeds succeed in reaching new wadi systems. Probably, the fl at coastal plain plays an important role in this respect as seeds that arrive on this coastal strip can easily be transported in a northerly or southerly direction, depending on the prevailing wind direction. Wadi branches in the coastal plain proper constitute no serious obstacles in this respect. I once witnessed a huge Zilla spinosa of almost 1.5 m in diameter rolling over the coastal plain. The wind was strong enough to lift this plant over a steep slope of a wadi branch of about 3 m high. The mountain slopes, however, constitute insurmountable barriers, resulting in large concentrations of tumbleweeds on specific spots . Dispersal of seeds by means of tumbleweeds, which lose their seeds during rolling, is an effective method in arid and semi-arid environments. Due to the scarcity of water, no closed vegetation is developed, and most plants can reach their optimal shape. This enables plants such as Zilla spinosa to develop into a spherical shape. The root of this species is twisted alternately to the left and right and weakens at the ground level after seeds are ripe. As a result, the plant is easily uprooted and becomes a tumbleweed. Not only spherical-shaped plants behave as tumbleweeds. For example, the procumbent Aizoon canariense also is uprooted and becomes a tumbleweed, even with part of the root still attached. The small seeds can easily escape from the cracks in the capsules. Finally, the whole plant becomes disintegrated. A special case in this respect seems to be Aristida funiculata, which might be described as “tumbleseeds.” Unlike the other two Egyptian species of this genus, namely, A. adscensionis and A. mutabilis, both column and awns of A. funiculata are long and stiff.

The total length of a spikelet is almost 8 cm. It has been observed several times that seeds form a cluster in such a way that the awns are interwoven and the seeds are projected on the outside of the clump. Another advantage of this species in this respect is that the articulation point is just below the base of the column. Due to this position, the shape of the clump of seeds is not affected when seeds break off during tumbling. Two such seed tumbles were found entangled in Convolvulus hystrix, showing that only part of the seeds were lost during their probably short trip. Dispersal by animals is possible by external carriage and by secretion via the guts. Spiny fruits, such as those of Arnebia hispidissima, or sticky ones, such as those of Boerhavia repens, are primarily transported via the outside of animals. The seeds of Acacia tortilis may serve as an example of the second mode of dispersal. Both unripe and ripe fruits of A. tortilis are partly found beneath the foliage, partly covered by sand. Such seeds need a secondary dispersal agent to find a safe site to germinate. More effective is the dispersal by camels internally. Camels have no problems with the long, stiff spines and are able to browse from the branches . Whereas twigs, leaves, and fruits are digested, the extremely hard seeds are excreted with the dung. Frequently, seedlings can be observed that germinate from inside camel dung . Owing to the wind, dung particles are quickly covered with sand, which in turn improves the growing conditions for the young tree. The number of seedlings that develop in a dung particle may vary considerably and will depend on the number of fruits excreted in a dung particle, as well as the number of mature, unaffected seeds in those fruits. Failed fertilization, abortion of seeds, and the attack of insects all contribute in reducing the number of viable seeds in a fruit . A concentration of 17 seedlings of A. tortilis proved to have originated from camel dung containing 18 seeds and covered by 10 cm of sand. Only when such camel dung is dropped at the edge of a wadi can one of the seedlings become a full-grown tree because strong water currents periodically destroy the vegetation in the central part of a wadi. For the same reason, these trees have a better chance of surviving along convex banks. One of the few shrubs that has frequently been observed in the central part of the wadi and is capable of surviving flooding is Chrozophora tinctoria, though its appearance is little more than that of a heavily damaged dwarf specimen.A variety of trees and shrubs were found in the vicinity of the Roman installations: Lycium shawii, Tamarix nilotica, T. aphylla, Suaeda monoica, Acacia tortilis, Salvadora persica, Balanites aegyptiaca, Calotropis procera, and Leptadenia pyrotechnica. The most common tree proved to be A. tortilis, found along the edges of many wadis, both in the mountains and in the coastal plain. Also around Berenike, this tree is frequently present in the small wadi branches. More substantial populations of acacias have been observed in the coastal plain south of Berenike. Despite its predominance, Acacia tortilis may be absent locally, such as in Wadi Gimal near Khesm Umm Kabu. Here, Balanites aegyptiaca and Calotropis procera are the predominating trees. Ababda nomads confirmed that Wadi Gimal, in particular, was characterized by B. aegyptiaca. In wadi branches dissecting the coastal plain,container size for raspberries this tree is a rather strange phenomenon. This is not the case with C. procera, which can frequently be observed in the coastal plain, especially south of Berenike. Tamarix nilotica and Suaeda monoica are halophytes and occur in the near vicinity of Berenike. On the higher levels of the salt marsh north of Berenike, several large hillocks of T. nilotica are currently present.

When several rows of such hillocks are present, the population functions as a natural barrier against penetrating windblown sand. In the salt marsh south of Berenike, T. nilotica is intermixed with S. monoica. Here, T. nilotica does not form hillocks and is present on the moreelevated levels, whereas S. monoica is concentrated in the lower levels of the salt marsh, together with the halophytic Halopeplis perfoliata. On both sides of the Red Sea coast, S. monoica gradually replaces the halophyte Nitraria retusa in a southerly direction. Halfway between Wadi Gimal and Ras Banas there is a transitional zone, whereas south of Ras Banas only populations of S. monoica are present . The Suaeda population south of Berenike dropped most of its leaves after the heavy showers in November 1996. The distribution of the salt-requiring S. monoica is confined to locations with a specific chloride content and the water table . Laboratory experiments with artificial solutions containing only traces of sodium resulted in the yellowing of leaves of S. monoica, as well as curling and wilting, whereas field observations revealed that seasonal fl uctuations have apparent effects on the distribution of this species . Other salt-tolerant species present in the salt marsh around Berenike are Cyperus conglomeratus, Zygophyllum album, and Z. simplex and an herbaceous Suaeda sp. Zygophyllum album has a wide ecological range with respect to the salt content of the soil, and Z. simplex is even less demanding as it also grows in the mountains, far away from the influence of the sea. The ecological preference of Z. coccineum, which is also rather common in the surroundings of Berenike, is less clear. The species has a large tolerance with respect to salinity and it is said to be indicative of limestone . In fact, both environmental conditions are present around Berenike, and each condition could explain its presence in itself. With respect to the salt content, possibly two ecotypes exist: a salt-tolerant and a nonsalt-tolerant one. It could be that the specimens growing in the vicinity of Berenike concern the salt-tolerant ecotype. Obviously, these specimens are full sized, whereas those found in the vicinity of the Roman settlements in the Red Sea mountains were less frequent and also of a reduced size. The population of Z. coccineum near Hitan Rayan might be of the nonsalttolerant type as this inland area does not witness the maritime influence of the Red Sea. Furthermore, the area is of Precambrian crystalline rocks and no limestone is present in Hitan Rayan or any of the surrounding hills . Salt-tolerant plant species are present near the shoreline on the supratidal flat , which consists of sandy, silty, and clay-like sediments and on a broad strip of the coastal plain adjacent to the sabkha. The latter is characterized by a salty horizon near the surface, which is quite difficult to penetrate. Mangrove vegetation is absent in the near vicinity of Berenike; its nearest population is located several kilometers north of Ras Banas. In Egypt this kind of vegetation mainly consists of Avicennia marina. South of lat 23° N it is also mixed with Rhizophora mucronata . Mangrove is present in calm, shallow water and on a fl at, muddy soil, permanently flooded or during high tides. Both protected lagoons and bays are suitable habitats. The northeast boundary of the present distribution is marked by some scattered specimens about 20 km north of Hurghada and becomes quite dense south of lat 25° N. At some localities this species also grows on sandy hills along the coast.Living in a desert environment essentially leads to its exploitation. Due to the limited carrying capacity of the environment, which is basically determined by the availability of water, people mostly live in relatively small groups that have adopted a nomadic life style. As a consequence, the material culture is strongly related to the bare essentials of life. The current arid climate of Egypt was established during the early second part of the Holocene and has had an intrinsic influence on the desert culture.

The series editors would like to thank the following organizations and individuals for enabling the eight years of archaeological work at Berenike on which the annual reports and final compilations such as the present volume are based: National Geographic Society, Netherlands Foundation of Scientifi c Research , Gratama Foundation, donors of the Berenike Foundation, Mallinson Architects of London, the University of Delaware, the American Philosophical Society, the Dorot Foundation, the Samuel H. Kress Foundation, Lotus Hotel in Cairo, Utopa Foundation, Organon Egypt, Eastmar Travel in Cairo, Philips Egypt, 3-Com Computer, Dionysus Systems, and in alphabetical order the following private donors: Carmine Balascio, Millie Cassidy, Bruce Gould, James Harrell, Charles Herndon, Susan Hodge, Dorothy Johnson II, Norma Kershaw, Carol Maltenfort, W. Weissman and W. Whelan.For a long time Berenike has been on the list of desiderata of many archaeologists. It was not until 1818, however, that G. B. Belzoni,raspberries in pots a weightlifter engaged in hunting antiquities in Egypt, identified Berenike south of the large Ras Banas peninsula.

Several visitors became interested in the site, and the expeditions that followed were mainly focused on excavating the temple dedicated to the Greek-Egyptian god Serapis, situated on the highest ground at Berenike . These visits, however, were short-lived due to the extreme desert conditions and related logistic problems. Food and water were not locally available and could be only delivered over large distances. In addition to the logistic problems, security reasons were also a serious hindrance to the realization of a large-scale excavation. The strategic location of the Greek-Roman harbor of Berenike has its modern counterpart in the presence of a large military base on Ras Banas and scattered bunkers on the limestone outcrop on which Berenike was founded. Furthermore, both Egypt and Sudan have a claim to the Hailab area, a triangular piece of land bordered along the Red Sea by Shelateen and Hailab. For these reasons, the Eastern Desert south of Marsa Alam was until recently only accessible with special permission. A couple of years after the Gulf War , it became possible to start the Berenike project, which included excavations at Berenike and Shenshef and the survey of the southern part of the Egyptian Eastern Desert. The excavations and surveys were directed by Prof. Dr. S. E. Sidebotham from Delaware University and Dr. W. Z. Wendrich from Leiden University.

After a pilot excavation season in 1994, large-scale excavations began and continued until December 2003, when the threat of an invasion of Iraq was at hand. Including the pilot study, a total of eight excavations seasons were conducted. The aim of the research project was to study the organization of trade and the subsistence economy at Berenike. Berenike was, together with Myos Hormos, an important harbor along the Red Sea coast. It was founded approximately 275 BC for the import of all kinds of luxury commodities from Africa south of the Sahara, Arabia, and India. Only at the beginning of the seventh century AD before the Arabs arrived in Egypt, was Berenike abandoned by the Romans. Several written historical sources deal with this trade, and these documents have been used for a long time to reconstruct this foreign trade. It remained uncertain, however, how reliable this picture was. Written sources are certainly not complete in their enumerations, and the interpretation of certain exotic commodities is still under discussion, as can be judged by the different translations of the Greek and Latin texts. The available information on the ancient trade contrasts sharply with that dealing with the subsistence economy of Berenike. Although located along the Red Sea, Berenike faced a desert climate that must have drawn heavily on its food supply. Much progress was therefore to be expected from the Berenike project, which had the cooperation of all kinds of specialists. This book synthesizes the results of the archaeobotanical research carried out over the years at Berenike and Shenshef and includes information that has been published in previous interim reports. The archaeobotanical research is based on the study of plant remains that have been preserved in both sites.

In Egypt, plant remains are predominately preserved by desiccation owing to the arid climate. Additionally, plant remains may become charred as a result of deliberate burning, such as in the offering of food items, or by accidental fire. The desiccated plant remains, in particular, are still in an excellent condition and facilitate, in most cases, identification to the level of species. In this way, a detailed list of plant species could be produced from Berenike and Shenshef, providing a solid basis for a reconstruction of the international trade of plant products and the food economy. The sub-fossil plant remains from Berenike originate from 794 samples and those from Shenshef from 86 samples. The samples from Berenike represent 34 different trenches, which are partly related with buildings and partly with trash deposits. Those from Shenshef originate from 10 different middens. Basically, each soil unit from a particular trench has been dry-sieved over a sieve with a mesh size of 4.0 mm. All sieve residues have been sorted out and have yielded many so-called “hand-picked” plant remains. A disadvantage of this procedure is that it is highly biased in favor of plants that produce large fragments. Cereals, for example, will only be secured in this way by the long rachis fragments in such sieve residues, whereas most of the grain kernels, threshing remains, and accompanying seeds of weed plants will get lost. To obtain a representative record of plant remains from Berenike and Shenshef, botanical samples were collected that have been processed by using sieves with a standardized mesh size of 5.0 mm, 2.0 mm, 1.0 mm, and 0.5 mm. The botanical samples from trash deposits, in particular, proved to be rich in well-preserved plant remains. Samples that were secured from the inside of buildings, on the other hand, were in most cases soil samples with only a low concentration of plant remains. Because these sediments were, for the greater part, considered as wind-blown secondary fillings of buildings, the plant records from these samples could not be used for an interpretation of the use of these buildings. For the reconstruction of agricultural practices, it is desirable to have at one’s disposal a considerable number of samples that presents one particular crop and associated weed plants. Unfortunately, almost all samples did contain at least two crops, such as wheat, barley, or lentil. This makes it frustrating to reconstruct the specific agricultural practices and also hampers the tracing of the origin of the crops. Most crops have acquired a broad tolerance to environmental factors as a result of selection processes, which is, for example, expressed by the many land races that have evolved. In most cases it is not possible to identify the area of origin of a cultivated plant found outside its production area, such as in Berenike, by studying the morphological features of its remains. Weed plants, on the other hand, are partly indicative of specific environmental conditions. Such plants can be used to trace the source of a particular crop plant indeed,blueberries in containers growing though its association with a particular crop should be evidenced by the archaeobotanical record. Samples with only a few plant species, rather than those with a high diversity, are suitable in this respect because one can usually unequivocally match these crops with their associated weed plants. Because the samples from Berenike and Shenshef did not meet this condition and also because of the large number of samples, it has been decided to categorize all plant records in several tables. Although Berenike was inhabited throughout the Ptolemaic and Roman periods, the archaeobotanical results are confined to the Roman period. It appeared that most of the Ptolemaic remains are located eastward of the Roman settlement. The few Ptolemaic contexts that have been unearthed had suffered seriously from salt seepage and did not yield identifiable plant remains. It was realized at an early stage in the project that the interpretation of the archaeobotanical records would benefit from the study of the current natural vegetation around the Roman installations and from observations of the local Ababda nomads who live in the southern part of the Eastern Desert.

For that reason, inventories have been made of the desert vegetation around the Roman installations during the various excavation seasons. Additionally, a separate visit was organized to the Gebel Elba area, located some 260 km south of Berenike, which was one of the source areas of some of the trade products. The ethnoarchaeobotanical research of the Ababda nomads was focused on the basic necessities of desert life with special emphasis on their material culture, their exploitation of the natural environment, and the possibilities of local food production. The reconstruction of the former natural vegetation is based on the archaeobotanical records, evidence from written sources, and the current vegetation. The interpretation of the cultivated plants found at Berenike and Shenshef summarizes their possible use, including the possibility of local food production, the differences in habitation periods, and the identification of their possible source of supply. The identification of the possible area of supply proved to be rather problematic because of the botanical richness of most of the samples. Determining the supply area is based on the natural distribution of the plant species, its archaeobotanical record from Egypt and the surrounding areas, and on the evidence from written historical sources, if available. A special topic related to the long-distance transport and the hot climate in Egypt is the preservation of food and the luxury status of food items. Based on the different methods of food preservation known from historical sources and those practiced today, the possible ways in which the various plant products might have been treated are discussed. Finally, the archaeobotanical record of exotic plant species from ancient Egypt is discussed, with special emphasis on those evidenced from Myos Hormos, the other important harbor along the Red Sea.Knowledge of Indian trade comes from various historical sources. Two of these sources, the Alexandrian Tariff and the Periplus Maris Erythraei, are of particular interest since they include a considerable list of botanical commodities that might have been traded through Berenike . Both sources provide first hand information. It is assumed that the author of the Periplus was an Egyptian Greek who had traveled at least along the African coast south to Rhapta and to the west coast of India. Contrary to other periploi that have survived, this Periplus is exceptional in its emphasis on trade items, in addition to information on the trade route proper. The Alexandrian Tariff was issued between AD 176 and 180 by Marcus Aurelius and enumerates 54 items subject to import duty at Alexandria on their way to Rome. It includes 20 different plant products, only half of which are also mentioned in the Periplus: costus, cassia, aloe, lykion, myrrh, malabathron, black and long pepper, nard, and aromatics. The Periplus, dated between AD 40 and 70, describes in great detail the trade routes from Myos Hormos, now identified as Quseir al-Qadim, and Berenike to India, including many harbors along the African, Arabian, and Indian coasts. The Erythraean Sea formerly included the Red Sea, the Gulf of Aden, and the Indian Ocean. The Periplus enumerates a vast number of import and export commodities, including 34 products of botanical origin, of which 18 are reported as import items from Berenike. The trade items mentioned in both documents represent a wide variety of plant parts: root; wood; bark; plant secretions such as resins, gums, and oils; leaves; flowers; seeds; and fruits, as well as whole plants . Plant secretions and seeds or fruits are the best-represented categories. A few trade items are either unspecified or not documented in such a way that a reliable identification can be made . The arrangement according to plant part is in some cases arbitrary. Aloe, for example, may be either a resin or fragrant wood, and both rhizomes and leaves were traded asnard. In some cases, trade items are reduced to a common trade name, such as “cassia” for “cassia turiana” and “xylocassia.” The scientific plant names are mainly based on Warmington , Miller , and Casson and have been updated where necessary.

Data were analyzed using a 3-way ANOVA with density of seed , application of B. bassiana , and date as factors . The commercial blueberry test site selected was located north of Bakersfield in Delano, CA. The trial began in August of 2008 and was conducted post blueberry harvest. The V. corymbosum varieties contained within the test area were, ‘Santa Fe’, ‘Jewel’, and ‘Star’. The most susceptible variety of blueberry to citrus thrips damage grown at the test site was the ‘Star’ variety and ‘Star’ was used consistently for evaluation of thrips numbers for all aspects of the trial . Our cooperator was interested in alternatives to traditional pesticides as the farm regularly was dealing with extremely high citrus thrips populations. For example, in 2008 the grower sprayed 5-10 times per field , rotating with traditional chemicals to reduce thrips impact on the subsequent year’s fruit set. Irrigation in all fields took place via drip irrigation with one water delivery emitter per line at each plant base , but additionally, one portion of the blueberry field was equipped with 360° overhead sprinklers. This irrigation setup provided the ideal situation to test B. bassiana under two watering regimes. The commercially available GHA strain is formulated to be mixed with water and for application via chemigation or as a foliar spray.

The label states that no surfactant is needed to keep the spores in suspension. However,plastic plants pots agitation alone in the 1,892.7 L holding tank was not sufficient to keep the material from precipitating, therefore 312.3 ml of Silwet L-77 was added to the tank mix. Mycotrol O® was applied directly to the soil surface with a gas-powered sprayer with a hand spray gun equipped with an adjustable flow meter. The dimensions of the plots were used to calculate the amount of material needed for both B. bassiana formulations . Plants in the test field were spaced every 0.92 m down each row, 3.35 m between each row, and each row was about 165 meters in length. Our studies were conducted inan 18-row section of a 4.04 ha field. The overhead sprinklers were spaced every 7 meters in the row and were located every other row for 12 rows. We chose to investigate the effectiveness of the B. bassiana colonized millet seed versus a Mycotrol O® soil application under two watering regimes, drip-line alone versus drip-line with overhead sprinkler, because B. bassiana conidia are highly subject to desiccation. Comparing the soil drench in both irrigation types with the colonized millet elucidated the effectiveness of the treatments when compared to the control. The blocks were laid out in a 3 x 2 factorial design, with each block consisting of most of five rows of blueberries , each being 27.4 m long . The berm used to grow blueberries at the commercial farm was 1.21 meters wide and each plot was 27.4 meters long. The spacing between adjacent rows was 3.35 m, while the spacing between the plants down a row was approximately 0.92 m with 30 plants per treatment plot . These dimensions result in 0.157 ha treated with raw spores but because the top of the berm was where thrips activity was evident and would be sampled, only 36% of the soil surface area was treated.

The Mycotrol O® label states that the maximum field rate is 6.9 L/ha mixed in 935.3 L/ha water. We therefore chose to apply the entire 6.9 L of Mycotrol O® in 378.5 L of water per ha directly to the berm with no application between the rows, which resulted in 100% of the per ha rate of product being applied to 36% of the area and allowed the maximum amount of active ingredient to be applied to the area that would have almost all thrips activity . Our field trial was intended to determine the extent to which B. bassiana might fit into a program projected to both control citrus thrips effectively and provide rotation among available chemistries so as to reduce thrips resistance evolution. Thus, we felt it was important to operate under the best possible conditions for thrips infection by Mycotrol O® , regardless of financial considerations, i.e. application of product at the maximum label rate in the area where thrips were most likely to be active. The amount of millet seed used in the field trail was calculated based on the area of the berm to be treated and likewise with the Mycotrol O® treatment, only 36% of the total field area was treated. The amount of seed used was one colonized seed/ 2 cm2 over an area of 576 m2 ; the fact that 0.45 kg of seed was needed per 840 cm2 resulted in the application of 3.40 kg of colonized millet seed for the 8 treated plots . Every other plant within the middle ten plants of the middle row of each plot were sampled with pupation emergence cages . These cages were placed tight against the base of each set of canes on the east side . With 5 cages per block and 4 replicate blocks per treatment, a total of 20 cages sampled thrips pupation per treatment over two sample periods, i.e. for two consecutive 3-day periods after the Mycotrol O® soil drench. The treatments were: no B. bassiana with and without overhead sprinkler; colonized millet seed with and without overhead sprinkler; and a soil drench of Mycotrol O® with and without overhead sprinkler . In total, data were collected from 240 emergence cages over the duration of the trial .

The colonized millet seed was set to imbibe water and allowed to sporulate for three days before application and was applied using a hand fertilizer applicator . Four days post application of the millet seed, the soil drench of Mycotrol O® was applied and pupation emergence cages were placed in the field and left out for 3 days . After three days, the sticky cards from each emergence cage were collected and replaced with new cards and the traps were switched to the next plant on the east side. These traps were left in the field to sample thrips for another 3 days . Because the traps were placed out every other plant, this ensured that all of the middle ten plants were sampled over the two, 3-day sampling periods . For two weeks before through two weeks after the applications of B. bassiana , counts were taken of thrips levels on plants twice per week. Beat samples were taken by beating random canes of flush foliage such that the thrips would fall onto a 12 x 12 cm black acrylic beat tray. The numbers of thrips on the beat trays were counted quickly in the field. The counts were taken twice per week from each of the 10 central “data plants” from the ‘Star’ variety of each of the 24 test plots. The new green flush growth was measured on three dates to record the amount of growth since the beginning of the fungal treatment applications to determine if there were differences based on the treatments and amount of water applied to the different plots. Measurements were made of the average cm of new shoot growth over the 6-week trial period. Due to the complex nature of the experimental design, i.e. treatments nested in a 5-way ANOVA , beat count data were analyzed using PROC MIXED and means were separated using Tukey’s test . Figure 3-2 shows the location of late second instar citrus thrips at death in the greenhouse study as well as those that located pupal refuges on the plant. Based on where they dropped off the plant,blueberry pot data indicated that more than 92% of the thrips would have pupated off the plant, likely in the soil near the base of the plant. Numbers did not vary significantly by location over the seven sample dates of this study; therefore data were pooled . A key result was that the proportion of second instar thrips crawling down the base of the plant was higher than the proportion dropping off the plant at distances measured past the base the plant . The four emergence cages placed under the field blueberry plants in each cardinal direction provided a means of sampling late second instar thrips moving towards the soil to pupae versus adults emerging out of the soil following pupation . Total numbers of thrips collected were pooled for the four traps in each direction at each respective location to determine which cardinal direction showed the most activity, and therefore was the most appropriate location to sample for citrus thrips in the field trial. Emergence cage data were summarized in two ways; the number of thrips moving off the plant to pupate in the leaf litter and the number of thrips emerging from the leaf . Data from the nested ANOVA generated p-values for direction , as well as distance grouping from the base of the plant . The cage closest to the base of the plant had significantly higher numbers of thrips emerging from the soil . Numbers of thrips trapped from the eastern cardinal direction were significantly higher for both mean numbers of thrips moving to and from the soil , indicating that for the field trial, emergence cages should be placed directly next to the base of the plant on the eastern side to sample the location that would have the most thrips activity.

Of the proportion of thrips not finding pupal refuge on the plants with colonized seed , 100% infection was seen with each of the different quantities of seed, i.e. each of 0.5, 1, or 2 seeds/cm2 was a sufficient density to infect and kill all late second instar thrips in the greenhouse study. No thrips were infected in the control treatment. There were insufficient data to conduct a 3-way ANOVA because all recovered thrips were infected with the fungus. Because all three densities tested were effective, we chose to utilize the most economical density in the field trial, i.e. 0.5 seeds/cm2 . In the split-plot design model, the whole plot factor was water and the split-plot factor was fungus treatment in a type three analysis of variance . Water, time and treatment were the main effects in the full model. Thrips levels measured on pupation traps at 3 days after treatment were lowest with colonized millet seed, intermediate with Mycotrol O ® , and highest in the untreated control . Additionally, there was fewer thrips counted in the colonized millet seed treatment than in Mycotrol O® treated plots . However, at time two , thrips levels with Mycotrol O® were no longer significantly reduced in relation to the control . While thrips levels measured using pupation traps were significantly less than observed in the control in all plots , thrips levels on plants measured using beat samples did not show a significant decrease , although comparing data with no overhead water that with overhead water, there appeared to be fewer numbers of thrips in the overhead sprinkler plots . The measurements from the new green flush growth in the overhead sprinkler treatments showed that those plants had longer growth than those without overhead sprinklers , but thrips numbers were not significantly lower on those plants. At none of the times when foliar beat counts were taken were there significant differences in thrips numbers across any of the three treatments . The ultimate goal of this work was to determine if the GHA strain of Beauveria bassiana could be used effectively as an alternative to traditional insecticides in commercial blueberries in California. Laboratory and greenhouse trials with Beauveria bassiana have shown variable success in controlling thrips and several other insect species , whereas field trials have shown limited overall success, but very few field trials included Thysanoptera . This is mainly due to the fact that climatic conditions in the laboratory and greenhouse situations are stable and often more humid than the ambient field environment in arid areas like most of California.

The trial was initiated in late fall 2000 in a drip-irrigated vineyard near Greenfield, Calif., and continued through the 2005 harvest. The vineyard was established in 1996 with Vitis vinifera L. cv. Chardonnay on Teleki 5C rootstock. Vine spacing was 8 feet between rows and 6 feet within rows. Annual rainfall normally ranges from 4 to 10 inches. Soil is elder loam with gravelly substratum. The vineyard was drip-irrigated from April to October.Row weed control treatments were: cultivation, post-emergence weed control only and pre-emergence herbicide , followed by post-emergence herbicide applications . Cultivations and herbicide applications were timed according to grower practices and label rates. Cultivations were carried out every 4 to 6 weeks during the growing season using a Radius Weeder cultivator . The cultivator used a metal knife that ran 2 to 6 inches below the soil surface cutting weeds off in the vine row; it had a sensor that caused it to swing around vines. Pre-emergence herbicides were applied in winter with a standard weed sprayer, and post emergence herbicides were applied in spring through fall as needed with a Patchen Weedseeker light-activated sprayer . An early and late-maturing cereal were chosen for the cover-crop treatments; legumes were not considered due to aggravated gopher and weed problems.

Cover-crop treatments in the middles were: no cover crop ,plant pot with drainage earlier maturing ‘Merced’ rye and later maturing ‘Trios 102’ triticale . Cover crops were planted with a vineyard seed drill in a 32-inch-wide strip in the middle of 8-foot-wide rows just before the start of the rainy season in November 2000 to 2004 . They were mowed in spring to protect vines from frost, and both cover-crop species senesced by summer. Prior to planting cover crops each November, row middles were disked to incorporate the previous year’s cover crop and stubble and prepare a seedbed. Weed control and cover-crop treatments were arranged in a 3 x 3 split block design with three replicate blocks covering a total of 23 vineyard rows . Each block contained six vine rows and six adjacent middles. Weed control treatments were applied along the entire length of each vine row ; cover-crop treatments were established along one-third of each middle and were continuous across the main plot treatments in each block. Each replicate main plot-by-subplot treatment combination included 100 vines. Soil compaction. Soil compaction was measured in the vine row in November or December 2003, 2004 and 2005 with a Field Scout Soil SC-900 compaction meter . Ten sites in each plot were sampled to a depth of 15 inches.

Soil moisture. Soil water storage was evaluated from volumetric soil moisture measurements taken in-row and adjacent middles to a depth of 3.5 feet at 1-foot intervals using a neutron probe. The neutron probe readings were calibrated with volumetric moisture measured from undisturbed soil cores collected at the site. Rainfall and runoff. A tipping bucket rain gauge with an 8-inch-diameter collector was used to monitor daily and cumulative rainfall at the field site. Runoff was collected at the lower end of the plots into sumps measuring 16 inches in diameter by 5 feet deep. Each sump was equipped with a device constructed from a marine bilge pump, a float switch and flow meter, to automatically record the runoff volume from the plots during storm events. During the second and third years the sampling devices were modified to collect water samples for sediment and nutrient analysis. Vine mineral nutrition. One-hundred whole leaves opposite a fruit cluster were collected from each plot at flowering in May 2003, 2004 and 2005. Petioles were separated from leaf blades, and tissue was immediately dried at 140°F for 48 hours and then sent to the ANR Analytical Laboratory for nutrient analyses. Petiole and leaf-blade tissue samples were analyzed for nitrate , ammonium , nitrogen , phosphorus , potassium , sulfur , calcium , magnesium , boron , zinc , manganese , iron and copper . Soil mineral nutrition. Composited samples from 10 soil cores taken to a depth of 1 foot were collected from the vine rows and middles at flowering as described above. Samples were air dried and sent to the ANR Analytical Laboratory for analyses. Soil samples were analyzed for pH, organic matter, cation exchange capacity , nitrate, Olsen-phosphorus, potassium, calcium, magnesium, sodium , chloride , boron and zinc. Soil microbial biomass. 

Due to the limited capacity of the laboratory, microbial biomass assays were conducted on selected treatments. Ten soil cores were collected to a depth of 1 foot and then composite samples were made from each replicate of the pre-emergence and cultivation weed-control treatments and the adjacent middles of the ‘Merced’ rye and bare treatments. Samples were collected about four times each year from November 2001 to November 2005 for a total of 14 sets of samples. Soil samples were immediately placed on ice and taken to the laboratory for soil microbial biomass carbon analysis according Vance et al. . Mycorrhizae. Roots were collected, stained and examined as previously reported on April 16, 2003, May 3, 2004, and June 2, 2005. Grape yield, fruit quality and vine growth. Fruit weight and cluster number were determined by individually harvesting 20 vines per subplot. Prior to harvest a 200-berry sample was collected from each subplot for berry weight and fruit composition. Berries were macerated in a blender and the filtered juice analyzed for soluble solids as Brix using a hand-held, temperature compensating refractometer. Juice pH was measured by pH meter and titratable acidity by titration with a 0.133 normal sodium hydroxide to an 8.20 pH endpoint. At dormancy, shoot number and pruning weights were measured from the same 20 vines. Statistical analysis. Analyses of variance were used to test the effects of cover crop, weed control and year on the vine, soil and microbial parameters, according to a split-block ANOVA model in SAS . Cover crop, weed control, year and their interactions were treated as fixed effects. The main and interactive effects of block were treated as random effects. Year was treated as May Jun Jul 26 24 22 20 18 16 B 20 18 16 14 A Bare ground Merced rye Trios 102 Fig. 3. Average soil moisture at 6 to 42 inches due to cover-crop treatments during the 2004 growing season for middles and rows ; date-by-cover-crop interaction. a repeated measure. When necessary, data were log-transformed to meet the assumption of normality for ANOVA, although untransformed or reverse transformed means are presented. Changes in soil moisture among treatments during the winter and the irrigation seasons were determined from significant treatment-date interactions. Compaction evaluation We conducted evaluations with a penetrometer each fall to determine the impact of weed-control treatments on soil compaction. Soil compaction was not significantly different at any depth in 2003 . However, in 2004 and 2005 soil compaction began to increase in the cultivation treatment compared to the other two weed-control treatments. In 2004,growing blueberries in pots soil compaction at the 4- to 7-inch depth was significantly greater in the cultivation treatment compared to the standard treatment , but not more so than in the post-emergence treatment . In 2005, the cultivation treatment had significantly greater soil compaction at the 4- to 7-inch depth than both the post emergence and standard weed-control treatments . At the 8- to 11-inch depth, soil compaction was significantly greater than the standard treatment , but not greater than in the post-emergence treatment . The blade of the cultivator passes through the soil at 2 to 6 inches deep, which may explain why greater soil compaction was measured there. Cultivations often also occurred when the soil was still moist following an irrigation, which may have contributed to the development of compacted layers over time.Moisture. 

Average, volumetric soil moisture levels at the 6- to 42-inch depth increased after the first rain events of the season, such as in winter 2002-2003 . Soil moisture declined most rapidly with ‘Merced’ rye in the middles during periods without rainfall each year , presumably due to its greater early-season growth and greater potential evapotranspiration, compared to the ‘Trios 102’ triticale. Soil moisture levels were similar between the bare and ‘Trios 102’ triticale treatments until May for all years. During the irrigation season, average soil moisture levels at the 6- to 42-inch depths were higher in rows than middles. Soil moisture in the rows and middles steadily declined during the irrigation season for all treatments during all years . Moisture levels declined most in middles with ‘Trios 102’ triticale cover during each irrigation season, presumably due to the later growth of this cover crop . In addition, the row soil-moisture levels also declined the most adjacent to ‘Trios 102’ triticale for the 2003 and 2004 irrigation seasons , but not during the 2005 irrigation season . Runoff. Total precipitation at the field trial was 7.4 inches during the 2002-2003 winter, 7.6 inches during the 2003-2004 winter and 9.9 inches during the 2004-2005 winter. A majority of the runoff was collected during December and January for the 2002-2003 and 2004-2005 winters, and February for the 2003-2004 winter. Cumulative runoff collected from individual plots during the three winters ranged from 0.02% to 3% of seasonal rainfall. Runoff was usually collected during rain events greater than 1 inch per day. Runoff was highest during the second and third years of the trial. During three consecutive winters, runoff was significantly lower in the cover crop treatments . ‘Trios 102’ triticale and ‘Merced’ rye had significantly less runoff than the bare treatment . Suspended sediment and turbidity were also significantly lower in runoff collected from the cover-crop treatments than in bare middles during winter 2004, but nutrient levels were similar among all treatments .Vines. Weed control and cover treatments did not have any significant effect on the nutritional status of the grape vines as measured by nutrient levels of the leaf petiole tissues, as determined by ANOVA. Although the nutrient levels by year were significantly different, the interactions of weed control-by-cover and weed control-bycover-by-year were not significant . Weed control and cover treatment also had no significant effect on blade nutrient content with the exception of boron and phosphate content. Vines adjacent to cover crops had significantly lower boron and phosphate levels in the leaf blade tissue than vines adjacent to bare row middles. As with the petioles, there was an absence of significance between the interaction of weed control-by-cover and weed control-by-cover-by-year for all nutrients analyzed .Soil cores indicated that most of the vine roots at this site were located under the vine row and few of the roots extended out to the row middles. This root distribution probably occurred because irrigation water was applied under the vines, and low rainfall at the site does not facilitate root growth into row middles. Thus, the lower nutrient levels in vines near cover crops may have been accentuated by irrigation effects that reduced vine root exploration of the soil to a narrow band under the vines. Since cover-crop roots probably grew into this zone there may have been competition between vines and cover crops for some nutrients. Soil. Cultivated rows had significantly lower levels of nitrate-nitrogen . Although the nutrient levels by year were significantly different, there was an absence of significance between the interaction of weed control-by-cover and weed control-by-cover-by-year . The differences observed in nitrate-nitrogen in the cultivation treatment may be due to the impact of loosening soil on water movement and leaching. Weed control treatments had occasional impacts on soil mineral nutrition in the middles, but results were inconsistent from year to year . Cover-crop treatments had no effect on soil nutrients in the rows . The most significant impacts of the vineyard floor treatments were of the cover-crop treatments on soil parameters in the middles. Soil organic matter in cover-cropped middles was higher than in bare middles each year . Cover crops affected key soil nutrients in the middles; for instance, cover crops greatly reduced nitrate-nitrogen , and to a lesser extent, extractable phosphorus , which may be beneficial in reducing loss of these nutrients in runoff during winter storms, but which also may have reduced the phosphorus content in the vines. In addition, cover crops in the middles also significantly reduced soil boron , extractable sodium and pH , and increased chloride and zinc when compared to bare soil.Soil microbial biomass.