Total DNA was isolated using an EDNA HiSpEx Tissue Kit , following the manufacturer’s protocol. This method is non-destructive, allowing slide mounting and morphological examination of the specimen after extraction. After DNA extraction, two separate gene regions were amplified using PCR: the conserved 28S-D2 domain of the large rRNA subunit and the cytochrome c subunit I of mitochondrial DNA . Post DNA extraction, specimens were transferred into 70% ethyl alcohol to prepare for slide mounting. Specimens were cleared in 5% NaOH for 12 h, processed through an alcohol dehydration series, placed in clove oil, and slide mounted in balsam . Specimens were identified to species using the online and interactive key of the “Thrips of California” . A representative from each haplotype generated from DNA analysis for this work was slide mounted and placed at the University of California, Riverside Entomology Research Museum as vouchers, and their collection numbers are included in Table 6-1. Representative species-specific sequences are deposited in GenBank. A BLASTN 2.2.19 search was used to compare sequences to existing Scirtothrips and Neohydatothrips sequences deposited in GenBank. Sequences were initially aligned manually in Bio-Edit version 6.0.7 . The COI sequences were all of equal length and contained no stop codons, with the first position equal to codon position one,blackberry container making alignment simple. The 28S-D2 included six ambiguous regions that were variable enough to make alignment difficult.
The six regions were coded as single multi-state characters in the analysis, or INNASE coding . This mixed alphabetical and numeric multi-state coding was treated as unordered and combined with the other data. Parsimony analyses using 1,000 random addition sequences and random starting trees were done on 1) the complete molecular data set , 2) molecular data with only 28S-D2 and 3) molecular data with only COI. All analyses were performed using PAUP 4.0* , with the complete matrix verified using TNT with a new technology search . Bootstrap values were generated in PAUP 4.0* using 1,000 BS replicates with two random heuristic searches for each replicate. Maximum Likelihood analyses and associated bootstrapping, were conducted on the complete molecular data set with RAxML v.7.2.7 via the CIPRES Science Gateway . Two gene partitions were included , and data were analyzed with 1000 rapid bootstrap replicates. None of the specimens collected from citrus in Texas were Scirtothrips; all thrips in the vials collected from multiple types of citrus were Frankliniella occidentalis and were therefore not included in any analyses for Scirtothrips citri or near citri. All specimens from California and Arizona keyed to Scirtothrips citri. The distinguishing features for S. citri include: 1) both sexes fully winged, 2) body mainly yellow without dark markings, 3) antennae 8-segmented with segments III – IV having forked sensorium and segments III-VIII grey, 4) head wider than long with ocellar triangle and postocular region having closely spaced sculpture lines, 5) three pairs of ocellar setae present with pair III close together between the anterior margins of the hind ocelli, 6) pronotum with closely spaced sculpture lines and the posterior margin with four pairs of setae, 7) metanotal posterior half with irregular longitudinal reticulations and median setae originating behind the anterior margin, 8) first vein of the forewing with three setae on the distal half , the second vein with three widely spaced setae and the posteromarginal cilia wavy, 9) abdominal tergites III – VI with median setae close together, tergites II – VIII with lateral thirds covered in closely spaced rows of fine microtrichia, these microtrichial fields had four discal setae, the posterior margin had a fine comb and the lateral discal microtrichia extended medially and lastly 10) abdominal sternites were without discal setae and the posterior margins were without a comb of microtrichia .
Morphological differences from the above on specimens from Florida , Mexico and Nicaragua could not be found and all were keyed to S. citri. The Florida citrus thrips collected from Mimosa sp. did not fit well into the Thrips of California key. The metanotal median setae arising at the anterior margin are in contrast to that of S. citri, where the metanotal median setae arise behind the anterior margin. All other characters appeared to be in congruence with S. citri. The thrips collected from Mimosa in Miami, Florida were keyed to Scirtothrips near citri. The specimens from Turkey were keyed to Scirtothrips citri using the Thrips of California key. Relationships were largely congruent among all analyses with the following exceptions: the group ‘Turkey_Hatey_citrus1_5’ in the combined tree and in the 28S-D2 alone tree as a sister group to the S. citri clade but these were placed without support along with S. aff dorsalis . This group switched positions, depending on the type of analysis. Results regarding S. citri will be discussed in terms of the combined parsimony analysis. The parsimony analysis of the complete molecular data set, including ambiguous region coding, resulted in 114 most parsimonious trees . The overall = consensus solution of all of the trees had no significant areas of conflict and collapsed with strong support for branches and monophyly of the groups. Further, analyses based upon exclusion of one gene region at a time resulted in trees with the following monophyletic groupings 1) Arizona and California, 2) Mexico-Nicaragua and 3) Florida . The group collected from Miami, Florida was sister to Scirtothrips bounites Mound & Marullo in all analyses. The RAxML analysis resulted in a single tree with a final Ln likelihood of -9,623.28. The parsimony and RAxML analyses both produced similar trees for the specimens collected in Arizona, California, Florida , Mexico and Nicaragua.
Clades Arizona and California, Mexico-Nicaragua, Florida and Florida , while morphologically indistinguishable, are highly molecularly divergent differing by 10-40 base pairs. The analyses conducted on molecular data confirmed that Scirtothrips citri in the Americas is not a single species but a complex of at least three molecularly divergent groups . The parsimony and maximum likelihood analyses both produced similar trees for the specimens collected in Arizona, California, Florida , Mexico and Nicaragua with the specimens collected from Turkey switching positions, depending on the type of analysis. Arizona and California specimens were morphologically identified using the Thrips of California key as ‘California’ citrus thrips and the Quincy , Mexico and Nicaragua specimens also keyed out as ‘California’ citrus thrips but this is not surprising as the key was developed for the known thrips found in California. The Miami citrus thrips collected from Mimosa sp. did not fit well into the Thrips of California key, which is again not surprising. Our molecular results suggested the specimens collected from Mimosa in Miami are S. bounites, although these specimens were identified as S. citri by a collaborator . To our knowledge, S. bounites has not been found in California to date but was collected from mango in Mexico . This suggests that without any consistent morphological differences, a cryptic species complex is likely present. The specimens collected from Turkey were morphologically keyed to S. citri but were grouped differently based on parsimony analyses of the two gene regions separately. The Turkey specimens in the 28S-D2 alone analysis were placed as the sister group to the overall ‘Scirtothrips citri’ clade however, in the COI alone analysis, the Turkey specimens are grouped within the Scirtothrips dorsalis Hood clade. These specimens do not share S. dorsalis specific characters such as microtrichia covering the sternites and straight wing cilia. In the parsimony and maximum likelihood combined gene region analyses,planting blueberries in a pot the Turkey specimens are placed as the sister group to the overall ‘Scirtothrips citri’ clade. The Turkey specimens did not appear to be different than S. citri morphologically and were included in the analysis because the group may be a new/related species to the ‘Scirtothrips citri’ clade. A comparison of bootstrap values between the individual gene trees suggests that 28S-D2 is driving the backbone of the combined tree. There is a need for further research into the relationships between the groups of S. citri presented here. Based on our molecular data, citrus thrips in California and Arizona are clearly different than those found in Quincy and Mexico-Nicaragua but morphological data suggests they are the same. These conflicting pieces of information illustrate the fact that morphological identifications may not be accurate enough especially when dealing with some organisms of economic importance. For example in a hypothetical situation, if specimens were detected and incorrectly morphologically identified as a species of economic concern but actually are a part of a non-economic clade, not distinguishing these groups could have serious import/export implications. Utilizing molecular markers for species identification and separation could be an immensely useful tool if/when morphological identifications are unclear. There are a few instances of cryptic species investigations with thrips and the development of molecular keys for many different types of thrips pests have aided in our understanding that thrips populations from various areas, or thrips collected from different host plants, are not necessarily all the same species, even if morphological analysis suggests this is the case.
Between 1994 and 2001 a joint expedition of the University of Delaware and Leiden University conducted eight seasons of excavations at the Ptolemaic-Roman emporium of Berenike on the Red Sea coast of Egypt. It was a great opportunity to work in an area that is usually closed for research, because it is part of Egypt’s border area with Sudan and is protected by the military. The Berenike Project was created as an interdisciplinary research effort involving a large number of archaeological specialists, who discussed their findings and interacted both on and off the site. Since many of the specialists were present on-site at the same time, preliminary results and conclusions could be shared and discussed often while excavations were in progress. This valuable collaborative effort could then be shared with those conducting the actual excavations. The annual reports on the excavations and specialists’ results have been published in six volumes, while the seventh, and last, is in preparation . These reports provide comprehensive interpretations and conclusions that have emerged over the years. At the end of this phase of fieldwork, however, it is time to publish an overview, and in the coming years several specialists will produce final reports. Although such final publications concentrate on specific artifact or material groups, they have the benefit of ongoing discussions among various Berenike team members. The result, in each case, will be a specialist’s report rooted in a broad basis of multidisciplinary insights . As the present volume amply demonstrates, the archaeobotanical research at Berenike provides essential information for our understanding of the subsistence patterns of the inhabitants, as well as the long-distance trade. The occurrence of exotic plant remains at Berenike changed through its history, with clear peaks in the first/second and fourth centuries CE. The results of the archaeobotanical research, therefore, provide extremely important information that supplements, complements, and, in some cases, changes the image that emerges from the ancient written sources. The variety of plants traded was larger than expected. Analysis of the well-preserved archaeobotanical remains found in our excavations at Berenike also indicates the regions with which the traders were in contact and when. This book will not only be interesting for archaeologists and archaeobotanists, but the comparison with written “classical” sources provides a fascinating read for historical and classical scholars as well. In some cases the Berenike finds provide evidence to propose new translations for Greek or Latin botanical terms, and correct current interpretations. Perhaps the most important result of the archaeobotanical research conducted at Berenike is the attested correlation between the harbor function, where goods were transferred from ship to caravan, and the amount and variety of goods that can be recorded. The items lost, purchased, or pilfered over the centuries at Berenike represent a micro version of the goods that were in demand throughout the Roman Empire as a whole. To research the imports used in Rome effectively, it is effective to start the investigation at the periphery of her empire: at Berenike, the harbor town through which many of these products once passed.