Preprints (www.preprints.org) | NOT PEER-REVIEWED | Posted: 25 October 2018 doi:10.20944/preprints201810.0587.v1 Peer-reviewed version available at Genes 2018, 9, 595; doi:10.3390/genes9120595 1 Article 3 Trade in Zambian edible orchids – DNA barcoding reveals use of unexpected orchid taxa for chikanda. 4 5 6 Veldman, S.1*, Kim, S.-J.1, van Andel, T.R.2, Bello Font, M.3, Bone, R.E.4, Bytebier, B.5, Chuba, D.6, Gravendeel, B.2,7,8, Martos, F.5,9, Mpatwa, G.10, Ngugi, G.5,11, Vinya, R.10, Wightman, N.12, Yokoya, K.4 and de Boer, H.J.1,2, 2 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 Department of Organismal Biology, Systematic Biology, Uppsala University, Norbyvägen 18D, 75236 Uppsala, Sweden; sarina.veldman@ebc.uu.se; seoljong.kim@gmail.com; 2 Naturalis Biodiversity Center, P.O. Box 9517, 2300 RA Leiden, the Netherlands.; Tinde.vanAndel@naturalis.nl; 3 Natural History Museum, University of Oslo, Postboks 1172, Blindern, 0318 Oslo, Norway.; mariabellofont@hotmail.com; h.d.boer@nhm.uio.no; 4 Royal Botanic Gardens, Kew, Richmond, Surrey, TW9 3AB, UK.; R.Bone@kew.org; K.Yokoya@kew.org; 5 Bews Herbarium, School of Life Sciences, University of KwaZulu-Natal, Pr. Bag X01, 3209 Scottsville, South Africa ; Bytebier@ukzn.ac.za; 6 Department of Biological Sciences, University of Zambia, Box 32379, Lusaka, Zambia.; david.chuba@unza.zm; 7 Institute of Biology Leiden, Leiden University, P.O. Box 9505, 2300 RA Leiden, the Netherlands; barbara.gravendeel@naturalis.nl; 8 University of Applied Sciences Leiden, Zernikedreef 11, 2333 CK Leiden, The Netherlands 9 Institut de Systématique, Evolution, Biodiversité (ISYEB), Muséum national d’histoire naturelle, CNRS, Sorbonne Université, EPHE; CP50, 45 rue Buffon 75005 Paris, France; florentmartos@gmail.com; 10 School of Natural Resources, The Copperbelt University, PO Box 21692, Kitwe, Zambia; gmpatwa@gmail.com; royd.vinya@gmail.com; 11 East African Herbarium, National Museums of Kenya, P.O. Box 40658-00100, Nairobi, Kenya; grace.ngugi@yahoo.com; 12 Homegarden Landscape Consultants Ltd., P/Bag 30C, Chilanga, Lusaka, Zambia; homegarden.nicholas@gmail.com; * Correspondence: sarina.veldman@ebc.uu.se; Tel.: +46737828087 1 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 Abstract: In Zambia wild edible terrestrial orchids are used to produce a local delicacy called chikanda, which has become increasingly popular throughout the country. Commercialization puts orchid populations in Zambia and neighbouring countries at risk of overharvesting. Hitherto, no study has documented which orchid species are traded on local markets, as orchid tubers are difficult to identify morphologically. In this study, the core land-plant DNA barcoding markers rbcL and matK were used in combination with nrITS to determine which species were sold on Zambian markets. Eighty-two interviews were conducted to determine harvesting areas, as well as possible sustainability concerns. By using nrITS DNA barcoding, a total of 16 orchid species in six different genera could be identified. Both rbcL and matK proved suitable to identify the tubers up to genusor family level. Disa robusta, Platycoryne crocea and Satyrium buchananii were identified most frequently and three previously undocumented species were encountered on the market. Few orchid species are currently listed on the global IUCN Red List. Local orchid populations and endemic species could be at risk of overharvesting due to the intensive and indiscriminate harvesting of chikanda orchids and we therefore encourage increased conservation assessment of terrestrial African orchids. 46 47 Keywords: CITES; Chikanda; Conservation; DNA barcoding; Orchids; Species delimitation; © 2018 by the author(s). Distributed under a Creative Commons CC BY license. Preprints (www.preprints.org) | NOT PEER-REVIEWED | Posted: 25 October 2018 doi:10.20944/preprints201810.0587.v1 Peer-reviewed version available at Genes 2018, 9, 595; doi:10.3390/genes9120595 48 1. Introduction 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 Terrestrial orchids have been used for medicinal and culinary purposes for centuries [1], with the most notable example being the use of orchid tubers to make salep, a traditional Turkish creamy drink or ice cream, consumed in Asia Minor and several countries on the Balkan peninsula [2–5]. In south-eastern Africa terrestrial orchid tubers are mixed with peanut flour, salt, baking soda and chili powder to make a traditional Zambian meat-like cake known as chikanda or African polony [6– 8]. Although initially not highly regarded [9], chikanda has more recently become popular throughout the country. It is sold as a snack along the streets, on markets, in supermarkets and on the menu of high-end restaurants [10] and recipes as well as cooking tutorial videos can be found online [11]. Orchids used for chikanda are harvested exclusively from the wild, and although it is unlikely that traditional village consumption poses a serious threat to orchid populations, the increased popularity and subsequent commercialization of chikanda has led to the exhaustion of Zambian orchid resources [8]. Collecting tubers means the end of a perennial and generally longlived orchid, since the entire plant is removed in the harvesting process. Soweto market in the Zambian capital Lusaka is the hub of the chikanda trade. Surveys performed on this market have shown that a large part of the chikanda tubers sold are sourced from Tanzania and that Zambian chikanda orchids are collected from the Luwingu, Mporokoso and Kasama districts in the Northern Province and Serenje in the Central province [7,8]. According to local chikanda vendors, another region with a flourishing chikanda trade is the Kitwe region in the Copperbelt Province, but so far no surveys have been performed there. Despite international legislation (CITES) banning prohibiting cross-border trade, an estimated 2–4 million orchid tubers are transported annually from Tanzania to Zambia [8,12]. Import from the surrounding countries of Angola, DRC, Malawi and Mozambique is also documented [8,10]. Orchid species originally reported as ingredients for chikanda are Disa robusta N.E.Br. and Satyrium buchananii Schltr. [12,13], whereas recently at least 32 species belonging to the genera Brachycorythis, Disa, Eulophia, Habenaria, Roeperocharis and Satyrium were suggested to be used for chikanda production based on collections in the field [12,14–19] and one metabarcoding study of ready-made chikanda cakes [19]. To date, however, no study identified the orchids traded at the local markets, since the tubers lack sufficient morphological characters for taxonomic identification to species level [8,19]. Local classification systems categorize the tubers based on texture, harvesting locality, soil colour and phenology, but these are not likely to be congruent with scientific classifications [6,15]. Knowing which orchid species are currently being collected for the expanding chikanda trade enables the identification of species susceptible to overharvesting and can inform conservation planning . Molecular methods such as DNA barcoding can be applied to identify samples when morphological diagnostic characters for identification are lacking [20]. DNA barcoding and metabarcoding has proven to be effective in the authentication of commercial wood species (Jiao, 2018), medicinal plants [21] and salep-producing orchids on Iranian markets [5]. The analysis of ingredients in Tanzanian chikanda cake with DNA metabarcoding revealed the presence of 21 different orchid species [19], but a DNA barcoding approach has not yet been applied to individual orchid tubers used to make this product. The aim of this study was to test to what extent species delimitation using standard molecular markers yields robust identification of chikanda orchid tubers traded on Zambian markets. Molecular identification can enable mapping of harvesting and trade of specific Zambian orchid species and facilitate identification and implementation of targeted conservation strategies. Within that framework this study also aimed to identify conservation issues associated with the chikanda trade, and addresses the following questions: 1) Which species are used for chikanda production in the Lusaka and Kitwe districts of Zambia and what is their geographic origin? 2) Can chikanda tubers be identified up to species level using DNA barcoding? 3) How do local classification systems relate to scientific species concepts? 4) What are the main conservation issues associated with chikanda trade in the Lusaka and Kitwe districts? 97 Preprints (www.preprints.org) | NOT PEER-REVIEWED | Posted: 25 October 2018 doi:10.20944/preprints201810.0587.v1 Peer-reviewed version available at Genes 2018, 9, 595; doi:10.3390/genes9120595 98 2. Materials and Methods 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 2.1. Interviews and sample collection Fieldwork in Zambia was conducted in 2016 in the Kitwe, Kalulushi, Luanshya, Ndola, Mufulira, Chingola and Chililabomwe districts of the Copperbelt Province; the Kapiri Mposhi and Serenje districts in the Central Province and in the capital Lusaka (Figure 1). Semi-structured interviews were conducted with harvesters, middlemen and vendors to obtain insight in chikanda commercialization, harvesting times, preferences and availability. The questionnaires consisted of three sections, one on informant and interview characteristics, one with general questions about chikanda posed to all informants, and a third section with questions more specifically designed for each interviewee category: harvester, middleman and vendor. All research was conducted in accordance with the International Society of Ethnobiology Code of Ethics [22]. Ethical clearance was obtained from the Humanities and Social Science Research Ethics Committee of the University of Zambia. The interviews were performed in English or Bemba, with a translator affiliated with the University of Zambia and the Copperbelt University. Informants were selected using the snowball technique [23], by asking people whether they could direct us to people harvesting or selling chikanda. All informants were provided with information about the study and signed a prior informed consent sheet. Fieldwork took place during June and July, the peak season for chikanda [8], to ensure collection of both fresh and dried chikanda tubers on the market and in the field. A collection was made each time a specific vernacular chikanda type was bought from a specific vendor, and assigned a collection number (SJK1, SJK2, etc.). Each individual tuber within the collection received a subsample number within that collection (SJK1.1, SJK1.2, SJK 1.16, etc.). The fresh tubers were sliced and stored with silica gel in plastic bags. All chikanda samples were brought to Sweden under a CITES inter-institutional exchange agreement between the University of Zambia (ZM001) and the Botany Section of the Evolutionary Museum in the Evolutionary Biology Center in Uppsala (SE009). Export permission was obtained from the Zambian CBD and Nagoya Protocol focal point at the Ministry of Natural Resources and Environmental Protection. Upon arrival to Sweden some of the chikanda tubers had sprouted. Those were transferred to the Uppsala Botanical Garden for cultivation and subsequent sampling of fresh leaf tissue for DNA barcoding as well as morphological identification. 127 128 129 130 Figure 1. Overview of interview localities and reported provenance of the chikanda tubers. Dot size corresponds to the number of informants. Preprints (www.preprints.org) | NOT PEER-REVIEWED | Posted: 25 October 2018 doi:10.20944/preprints201810.0587.v1 Peer-reviewed version available at Genes 2018, 9, 595; doi:10.3390/genes9120595 131 132 133 134 135 136 137 138 139 140 141 142 143 144 2.2. Reference taxon sampling Herbarium specimens were collected with associated silica-dried material for DNA extraction and spirit collections during fieldwork in Zambia in January and February 2017. In addition, putative orchid mycorrhizal fungi were sampled from roots and tubers for isolation, culture and identification at RBG Kew (results not reported here). All material was collected and exchanged in accordance with national and international legislation. The collections are deposited at the Division of Forest Research (Kitwe, Zambia) and RBG Kew (UK) and field identifications verified at the Bews Herbarium (South Africa). A total of 94 novel Orchidaceae reference vouchers were collected for this study, representing 4 Brachycorythis, 9 Disa, 16 Habenaria, 6 Satyrium species and 26 species in other orchid genera. Voucher specimens of all taxa sampled are listed in Table S1 (Supplementary Material). In addition, 88 ITS, 71 matK and 45 rbcL Habenaria sequences generated for a forthcoming phylogenetic study [24], and 510 ITS, 522 matK and 213 rbcL sequences corresponding to 311, 325 and 100 taxa in the previously mentioned orchid genera downloaded from GenBank were included in the reference database. 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 2.3. From sample to sequence Out of the 1284 individual tubers in 48 different sample collections, 304 samples were selected for DNA extraction. A few tubers per sample (2-8) were extracted if the sample was morphologically homogenous, whereas more (6-33) were selected if the sample was diverse. DNA was extracted using a CTAB protocol [25] modified with 3 to 5 extra washing steps with STE buffer (0.25 M sucrose, 0.03 M Tris, 0.05 M EDTA) [26,27], to reduce the gelatinization effect of the large amount of polysaccharides in the starch-rich orchid tubers. Total DNA was stored in 70-100 μl 10mM Tris-HCl buffer, pH 8.0. DNA concentration was measured with a Qubit 3.0 fluorometer (Thermo Fisher Scientific, Oakwood, USA). The core land plant barcoding markers rbcL and matK were amplified using the primers and protocols described in Kress et al. [28] and Dunning and Savolainen [29] respectively. The reactions were performed in a total reaction volume of 25μl with 14.725μl ddH2O, 2.5μl DreamTaq Buffer (Thermo Fisher Scientific, Oakwood, USA), 0.5μl 25mM dNTP, 0.65μl 2% Bovine Serum Album (BSA), 0.125μl DreamTaq Polymerase, 2.5μl 5pmol forward and reverse primer and 1.5μl template DNA. Nuclear ribosomal nrITS was amplified using the Sun et al. [30] primers and protocol in a total reaction volume of 25μl containing 15.25μl ddH2O, 2.5μl DreamTaq Buffer (Thermo Fisher Scientific, Oakwood, USA), 0.5μl 25mM dNTP, 0.125μl 2% BSA, 0.125μl DreamTaq Polymerase (Thermo Fisher Scientific, Oakwood, USA), 2.5μl 5pmol forward and reverse primer and 1.5μl template DNA. For ITS an additional protocol was used with Q5 highfidelity polymerase: reactions were performed in a total reaction volume of 23.5μl including 10.875μl ddH2O, 5μl Q5 reaction buffer, 0.5μl 25mM dNTP, 5μl Q5 GC enhancer, 0.125μl Q5 highfidelity polymerase, 1.5μl 5pmol forward and reverse primer and 0.5μl template DNA. The PCR program for the ITS primers in combination with the Q5 polymerase was an initial heating step of 30s at 98°C; 35 cycles of 10s 98°C, 30s 56°C, 30s 72°C; and a final elongation of 2min at 72°C. PCR products were cleaned using eight times diluted ExoSAP (Thermo Fisher Scientific, Oakwood, USA) and analysed on an ABI3730XL Sanger sequencer by Macrogen Europe (Amsterdam, The Netherlands). The obtained trace files were assembled using Pregap4 and Gap4 [31] as implemented in the Staden package [32]. Sequences shorter than 200 bp were discarded from the analysis and all sequences have been deposited in NCBI GenBank (Table S2, Supplementary Material). The NCBI BLAST algorithm was used to assess the identification of all the obtained query sequences using the Python BLAST tabular parser script (https://github.com/SLAment/Genomics/blob/master/BLAST/BLAST_tabularparser.py). Similarity scores, query coverage, expect value (E-value), and max identity percentage were calculated, and all the information of the top 5 hits were automatically mined and tabulated per marker (Table S3-S5, Supplementary Material), and summarized per individual tuber sample (Table S6, Supplementary Material). 180 181 2.4. Phylogenetic analysis and species delimitation Alignments for nrITS, matK and rbcL were made using AliView [33], combing the query Preprints (www.preprints.org) | NOT PEER-REVIEWED | Posted: 25 October 2018 doi:10.20944/preprints201810.0587.v1 Peer-reviewed version available at Genes 2018, 9, 595; doi:10.3390/genes9120595 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 sequences and local reference databases consisting of sequences from NCBI GenBank and reference collections from fieldwork and unpublished data from collaborators [33]. Species delimitation was performed using the Poisson Tree Processes (PTP) model, as it has been shown to outperform the Generalized Mixed Yule Coalescent (GMYC) approach as well as OTU-picking methods when evolutionary distances between species are short enough [34]. For all alignments a maximum likelihood (ML) search for the best-scoring tree was performed using the RAxML web server [35] to generate input trees for species delimitation analysis using bPTP [34]. GTRCATI was used to implement the CAT approximation, and the final tree was evaluated using the traditional GTR model. The bPTP.py script settings was 100,000 MCMC chain iterations for all trees; sampling interval thinning value of 100; burn-in of 25%; and a random seed of 1234. No outgroup was defined. The generated convergence curve was visually confirmed. The description of each voucher is built up as following: 'sample #_BLAST search result_identification %_reported city/region of the origin_the country,'; for example: ‘SJK04.09_S.buchananii_97.893_Mwinilunga_Zambia’ (Figure S13, Supplementary Material). For the reference sequences, accession number and species names are described for Disa and Satyrium. The reference Habenaria species and others are described as species name and voucher (sample) number. 198 3. Results 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 3.1. Market surveys and interviews We visited 25 markets in ten Zambian cities: seven cities in the Copperbelt Province, two in the Central Province, and one in Lusaka (Figure 1). Eighty-two persons involved in chikanda trade were interviewed of which 8 were harvesters, 44 middlemen, 29 vendors and one informant was both harvester and middleman. The term ‘harvesters’ refers to people collecting chikanda in the field, ‘middlemen’ to people selling either dried or fresh chikanda orchid tubers on the market, and ‘vendors’ to informal street vendors selling ready-made chikanda cakes contained in a basket carried on the head. Since one of the harvesters also acted as middleman, in total 83 individual interviews were conducted: 72 participants were female and 10 male. The ages of the respondents varied between 18 and 78, with an average of 41 years. The majority of the respondents belonged to the Bemba tribe (62%), while the other the respondents (38%) belonged to smaller ethnic groups. Most interviews were conducted in the Copperbelt Province (55), eight in the Central Province and 19 in Lusaka. 3.2. Local classification system Fifty-three different vernacular names for the various chikanda types were recorded during the interviews. The most common way to distinguish between chikanda tubers was by using the terms original (myala) and fake (mbwelenge or msekelele). In some cases it was the shape of the tuber that was used to differentiate between the different tubers: mshilamshila means root-like in Bemba and referred to the elongated, root-shaped tubers whereas mampanda referred to the heart-shaped tubers. It also appeared common to use the origin of the tuber as a trade name: mwinilunga, chozi, luwingu and kasama are for example all Zambian city names, sumbawanga and iringe are referring to Tanzanian cities (Sumbawanga and Iringa) and angola refers to one of the countries bordering Zambia. In some cases the chikanda tubers were sold pre-mixed, whereas other vendors marketed the different types of tubers separately. The morphology and size of the tubers varied both within and between collections of a certain chikanda type. Tubers could be heart-shaped, rounded, egg-shaped to elongated and almost root-like. The largest tubers were the heart-shaped ones, which could be up to 5 cm long and 6 cm wide. The elongated tubers were up to 9 cm long and maximally 2 cm wide. Harvesters, middlemen and vendors themselves indicated that they distinguish the tubers based on the size of the granules inside the tubers, which can be large or small and in some tubers a concentric ring was said to be present. Figure 2 illustrates the different tuber types and ready-made chikanda that were encountered on the local markets, as well as some of the orchid species producing these tubers. Preprints (www.preprints.org) | NOT PEER-REVIEWED | Posted: 25 October 2018 doi:10.20944/preprints201810.0587.v1 Peer-reviewed version available at Genes 2018, 9, 595; doi:10.3390/genes9120595 233 234 235 236 237 238 239 Figure 2. Chikanda tubers, cake and orchids. (a) Myala – real chikanda; (b) Mbwelenge – fake chikanda; (c) Mshilamshila – supposedly Brachycorytis sp.; (d) Mampanda. (e-g); Chikanda cake; (h) Disa robusta; (i) Disa welwitschii; (j) Platycoryne crocea; (k) Satyrium carsonii; (l) Satyrium buchananii; (m) Satyrium kitimboense; (n) Brachycorytis cf. friesii; Photographs (a-g) by Seol-Jong Kim, (h) by Robert v. Blittersdorff, (i,k and l) by Nicholas Wightman, (j) by Warren McCleland, (m) by Ruth E. Bone and (n) by Sarina Veldman. Preprints (www.preprints.org) | NOT PEER-REVIEWED | Posted: 25 October 2018 doi:10.20944/preprints201810.0587.v1 Peer-reviewed version available at Genes 2018, 9, 595; doi:10.3390/genes9120595 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 3.3. Chikanda trade and availability Many participants had a relatively long experience (average 13.5 year) in the chikanda trade. They indicated that they were asked by family, friends or neighbors to get involved in the chikanda business, or simply because it seemed a profitable industry. Some of the participants, especially harvesters, stated that they started collecting chikanda because the plants were easily accessible and were found growing close to their areas of residence. In general, people involved in chikanda trade indicated that this work alone did not suffice to fully support themselves and their families, and these traders therefore supplemented their income by trading additional natural products such as fruits, maize, groundnuts, beans, mushrooms, snacks, herbs, and kapenta (a dried fish from Lake Tanganyika). In Lusaka, however, income generated with chikanda trade seemed to be sufficient for subsistence, and this suggests that there are significant local differences in profits generated with chikanda business. The chikanda trade was structured in several different ways. Some of the middlemen loaded the orchid tubers in trucks directly from the harvesting areas and brought them to larger cities such as Lusaka and Kitwe, whereas others relied on agents to gather a certain amount of chikanda tubers, which they paid for through the East-African mobile payment system M-Pesa and received as cargo from one of the local buses. In addition to the Tunduma and Nakonde markets that are the tradehubs between Tanzania and Zambia, other centers of trade were identified on the border with the Democratic Republic of Congo (Chililabombwe and Kasumbalesa) and Angola (Mwinilunga), where both chikanda tubers and ready-made chikanda cake was sold. An overview of all interview localities and the reported provenance of the chikanda tubers is given in Figure 1. Most of the participants indicated that chikanda plants are becoming locally rare. Middlemen emphasized the decrease of quantity, whereas harvesters were concerned about the decline in both quantity and quality (size and preferred chikanda type), which may also depend on the species. The participants from urban areas stated that access to chikanda tubers was managed by the chief of each tribe, who seasonally designated the harvestable dambo (wet meadow) within the chiefdom so that the collectors could maintain the quality of the harvests. Nevertheless, interviewed harvesters encountered in the dambo areas of Serenje (Central Province) claimed to be free to harvest tubers whenever available. 3.4. Molecular identification of traded orchids During the DNA extraction many samples formed a thick jelly-like layer in the extraction tubes, despite repeated washing steps with STE buffer. This resulted in a very small water phase and likely negatively influenced the downstream steps of the extraction process. The average DNA concentration of the chikanda tuber samples was 4.96 ng/μl, while the average DNA concentration from leaf samples from chikanda orchids was 28.5 ng/μl. Out of the 304 samples selected for DNA extraction, 232 samples produced detectable DNA. Amplification was attempted for each barcoding marker for all of the samples. A nrITS sequence was obtained for 159 samples, rbcL for 117 samples and matK only for 45 samples. Sequences from all three markers were obtained for 40 samples, sequences from two markers for 58 samples, and 55 samples only yielded sequences for a single marker. Analysis of the inter- and intraspecific variation of nrITS was performed for 141 sequences representing 124 Disa species, 73 sequences representing 59 Habenaria species and 110 sequences representing 67 Satyrium species. In the case of matK 135 sequences belonging to 122 Disa species were included, 507 Habenaria sequences belonging to 239 species and 116 sequences belonging to 60 Satyrium species. For rbcL the available reference material was quite limited: 45 sequences for 37 Habenaria species, eight Satyrium sequences for two species and four Disa sequences for four species. A graphical overview of the inter- and intraspecific variation per genus and per marker can be found in Figure 3. The interspecific variation for nrITS was significantly higher as compared with matK and rbcL in all three genera: on average 10.2% in Disa, 9.51% in Satyrium and 8.75% in Habenaria. For matK the interspecific variation was 2.39% for Disa, 2.79% for Habenaria and 1.39% for Satyrium. The intraspecific distances for nrITS and matK were respectively 1.37% and 0.61% for Disa, 1.98% and 0.36% for Habenaria and 1.45% and 0.48% for Satyrium, The limited reference sequences that were available for rbcL showed little pairwise interspecific distances (1.44% for Disa, 0.82% for Habenaria and 1.13% for Satyrium) and only allowed for the intraspecifc distance calculation of Habenaria Preprints (www.preprints.org) | NOT PEER-REVIEWED | Posted: 25 October 2018 doi:10.20944/preprints201810.0587.v1 Peer-reviewed version available at Genes 2018, 9, 595; doi:10.3390/genes9120595 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 (0.38%), indicating that rbcL is unsuitable as barcode for species level identification of chikanda orchids. The calculated thresholds were subsequently used to evaluate the identifications made with blastn. Figure 3. Boxplots showing the inter- and intraspecific variation for Disa (a), Habenaria (b) and Satyrium (c) based on genetic diversity. In total 15 orchid species were identified using DNA barcoding: Brachycorytis sp. SJK7, Disa caffra Bolus, Disa celata Summerh., Disa robusta N.E.Br., Disa satyriopsis Kraenzl., Disa welwitschii Rchb. f., Disa sp. SJK4.1., Habenaria sp. SJK31.15, Habenaria aff. helicoplectrum Summerh., Habenaria cf. sp. DO112, Platycoryne crocea Rolfe, Satyrium buchananii Schltr., Satyrium carsonii Rolfe, Satyrium kitimboense Kraenzl. as well as one species in an unidentified genus, which seems to be closely related to Habenaria based on the similarity-based BLAST identification. Additionally, one orchid that flowered from a sprouting tuber was identified based on morphology as Brachycorythis cf. friesii (Schltr.) Summerh. bPTP analysis for matK and rbcL showed lumping of several supposedly different species within one clade on several occasions (Figure S1 and S2, Supplementary Material). In case of nrITS the bPTP outcome tree often reflected the expected species boundaries, although the posterior probabilities on the nodes were often too low to determine with confidence if species delimitation had been performed correctly (Figure S3, Supplementary Material). Something that could be observed from both the nrITS BLAST identification, as well as the bPTP analysis, was that some of the identifications showed ambiguity: in collection SJK41, SJK44 and SJK46 some samples were identified unambiguously as Platycoryne crocea, whereas others showed a mix of Habenaria and Platycoryne top hits. The matK BLAST identification showed an unambiguous Platycoryne crocea identification for these samples, despite the presence of several Habenaria species in the reference database and despite the fact that matK shows a lower level of interspecific variation. A similar observation could be made for several Disa samples, but here a geographic pattern could be observed. The samples that could Preprints (www.preprints.org) | NOT PEER-REVIEWED | Posted: 25 October 2018 doi:10.20944/preprints201810.0587.v1 Peer-reviewed version available at Genes 2018, 9, 595; doi:10.3390/genes9120595 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 unambiguously be identified as Disa robusta were all collected from Tanzania, where the reference sequence originated from. Some of the Zambian samples also showed Disa robusta as the closest relative, but not with a high enough percentage identity match to confirm this identification. The bPTP result in this case shows a lumped and mixed clade with several Disa species and several other samples included, which confirms that sequence divergence is too limited to resolve the relationship. These samples were identified as Disa sp. 1, to reflect the fact that they all grouped together in the same clade. In other cases, such as for certain Habenaria species, we named the closely related sequences in the identification, whereas in this case the clade was too large to allow for this, since it contained the following sequences: Disa engleriana Kraenzl., D. erubescens Rendle., D. miniate Summerh., D. ochrostachya Rchb.f., D. satyriopsis, D. ukingensis Schltr., D. verdickii De Wild., D. welwitschii, D. zombiaca N.E.Br., an unidentified Disa species and several samples that showed a highest percentage identity match with D. robusta, D. satyriopsis and D. welwitschii. The posterior probability for this clade is only 0.25, and combined with the posterior probabilities present on the within-species nodes, indicates a lot of uncertainty for this identification. Overall, the most frequently encountered species were Satyrium buchananii, with 41 samples in eleven different collections, Platycoryne crocea with 19 samples in three different collections and Disa robusta with 4-16 samples in two to seven different collections. Myala or original chikanda seems to contain Disa robusta, D. welwitschii and Satyrium buchananii. Mbwelenge or fake chikanda seems to correspond only to Satyrium buchananii and mshilamshila to one or several Brachycorythis species. Kasebelele and kapapa referred to Habenaria spp., Platycoryne crocea, Satyrium carsonii and Satyrium kitimboense. The mixed collections contained all of the above-mentioned species as well as an unidentified Habenaria species. An overview of the local chikanda classification types, their collection numbers, identifications and number of samples can be found in Table 1. 343 4. Discussion 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 4.1. Species used for chikanda Using DNA barcoding as an identification tool for chikanda tubers sold on local Zambian markets has allowed us to determine for the first time what orchid species are sold on local markets. Previous studies identifying orchids used for chikanda relied on voucher collections made with collectors in the field and their morphological identification, which requires a qualified orchid taxonomist [7,12,14,17,36]. Additionally, relying on local harvesters for details on chikanda collection might not always lead to collection in the areas where actual intensive harvest is taking place, since some harvesters do not like to divulge where the best places to harvest are and in some initial harvesting areas such as the Kitulo Plateau in Tanzania’s Southern Highlands, collection is now prohibited [37]. The current study identified 16 orchid species present on the markets, including at least three previously undocumented ones: Brachycorythis cf. friesii, Platycoryne crocea and an unidentified species in a genus, which appears to be closely related to Habenaria, but is not present in our reference database. Orchids used for chikanda seem to be harvested from several provinces in Zambia, as well as at least two regions in Tanzania. Moreover, three international chikanda trade-hubs in towns on or close to the border with the DRC (Chililabombwe and Kasumbalesa) and Angola (Mwinilunga) were identified, in addition to the already known trade-hub Tunduma-Nakonde on the Tanzanian border [8,12]. However, unlike reported in other studies, no harvest from Malawi was mentioned by the people interviewed in this study, which could mean that trade from this country is currently not taking place. Another explanation is that this information is lost on the way and that only Tanzania, being geographically closer and thus more easily accessible for the Bemba people, is mentioned as a region of origin for chikanda traded in Zambia. 4.2 DNA barcoding performance Since the term DNA barcoding was coined in 2003 [20], a plethora of studies applying DNA (meta)barcoding has been performed ranging from retrieving orchids from paleoenvironments [38], preserved in mammoth dung [39] to the identification of Iranian orchid tubers used for salep [5]. In Preprints (www.preprints.org) | NOT PEER-REVIEWED | Posted: 25 October 2018 doi:10.20944/preprints201810.0587.v1 Peer-reviewed version available at Genes 2018, 9, 595; doi:10.3390/genes9120595 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 this study a combined use of the core plant markers matK and rbcL and the nuclear ribosomal ITS region was used to attempt species level identification of tubers traded on Zambian markets. Although genetic distance calculations showed limited interspecific distances between closely related species for all three barcoding markers, DNA barcoding allowed for species-level identification for several of the frequently sold chikanda species. The data shows that the core land-plant DNA barcoding markers rbcL and matK were not suitable because of limited variability between species (matK and rbcL), amplification problems (matK), and/or a limited sequence reference database (rbcL). nrITS was shown to be more suitable as a barcode marker to distinguish between different chikanda species, but is not discriminative enough to enable reliable species level identification in certain orchid clades, such as the clade with Platycoryne crocea, P. buchananiana and Habenaria buchananii; the clade with H. schimperiana, H. kyimbilae and H. microsaccos and the clade containing Disa sp. 1. Another drawback of nrITS are the multiple ITS paralogs present in the ribosomal genome. Usually these copies would show high similarities due to concerted evolution [40,41], but this is not always the case [42–45]. In our case potentially different nrITS ribotypes became fixed in different orchid populations and having only one of them in our reference database could lead to the unresolved identifications observed. Our bPTP results show that even nrITS has too little resolution to reliably delimit species with high posterior probability support using this method. It does demonstrate, however, how valuable the use of tree-based methods can be, since it shows the relations between the sequences and can be used to determine if some of the unidentified samples are likely to belong to the same species or different species within the same genus. Even if no species-level identification can be made for these samples, it is possible to use the clustering to determine the diversity of species used. Although several samples can only be reliably identified as Habenaria sp., we find that they are likely to belong to at least three different species (Habenaria aff. helicoplectrum, Habenaria cf. sp. DO122 in the clade with H. schimperiana, H. kyimbilae and H. microsaccos and lastly the Platycoryne sp./Habenaria sp., which group together with H. buchananii, P. buchananiana and P. crocea. Expansion of the reference database, by including at least one individual per species, and preferably multiple individuals per species from different populations and countries, could ultimately solve remaining challenges, and this seems the way forward in identification of the traded chikanda tubers, as well as other species unidentifiable based on morphology. Similar studies using DNA barcoding for the identification of unknown samples show comparable results for the employed barcoding markers. In a study on the identification of orchids used for salep, nrITS showed an over three-time higher sequencing success than matK as well as a two-times higher species-level identification success [5]. Moreover, the similarity-based approach seemed to outperform the tree-based identification method (ML) in this study as well with 57% and 39% species-level identifications, respectively. nrITS also shows the highest identification performance in studies on the identification of medicinal plants [46,47], which supports the idea that it is recommendable to add a more discriminative marker to the two core landplant barcodes in studies where it is needed to distinguish between closely-related species. Moreover, our results stress the need for a phylogenetically underpinned taxonomic framework, which is currently available for Disa [48] and Satyrium [49], but not yet for Habenaria and related genera. 4.3. Local versus scientific classification of chikanda The results of our identifications of chikanda orchids traded in the Zambia show that the local classification systems for chikanda are not in congruence with the botanical classification of orchid species. The orchids sold on the markets were grouped according to area of origin, tuber consistency preference, or shape of the tuber, but often the tubers offered for sale were mixtures. The different local types of chikanda sometimes show a variation in orchid species that are identified within these local grouping. When we look at the chikanda type known as kapapa for example, SJK44 contains Platycoryne crocea, whereas SJK11, which is supposed to be a mix of kasebelela and kapapa, only seems to contain Satyrium species. In case of other chikanda types there seems to be more consistency: myala or real chikanda referred to Disa robusta, D. welwitschii and Satyrium buchananii, mshilamshila samples Preprints (www.preprints.org) | NOT PEER-REVIEWED | Posted: 25 October 2018 doi:10.20944/preprints201810.0587.v1 Peer-reviewed version available at Genes 2018, 9, 595; doi:10.3390/genes9120595 Table 1. Overview of the different local chikanda classification types, their collections and the identified scientific species. # Vernacular name Collections Reported origin Fungulwe SJK16 unknown Iringe SJK17 Tanzania John White SJK39 Mporokoso, Zambia Kabula seke SJK46 Serenje, Zambia Kapapa SJK44 Mporokoso, Zambia Kasebelela, John White and Myala Kasebulela and Kapapa Mbwelenge Mshilamshila SJK41 Chinsali and Mporokoso, Zambia and Tanzania SJK11 Luwingu, Zambia SJK5 Luwingu, Zambia SJK32 Serenje, Zambia SJK7 Luwingu, Zambia SJK12 Kawamba, Zambia SJK4 Mwinilunga, Zambia; Myala SJK18 Sumbawanga, Tanzania barcoding IDs samples Disa robusta 1 Satyrium buchananii 1 Satyrium carsonii 1 Satyrium buchananii 1 Habenaria sp. (Clade H. schimperiana, H. kyimbilae, H. microsaccos 1 Platycoryne crocea Rolfe 7 Platycoryne crocea Rolfe 6 Habenaria cf sp. DO122 (Clade H. schimperiana, H. kyimbilae, H. microsaccos 4 Platycoryne crocea Rolfe 4 Platycoryne sp./Habenaria sp. 2 Satyrium kitimboense 6 Satyrium carsonii 5 Satyrium buchananii 11 Satyrium sp. 1 Satyrium buchananii 6 Brachycorythis sp. 1 Brachycorythis sp. 1 Brachycorythis cf. friesii 1 Disa robusta 4 Disa welwitschii 1 Satyrium buchananii 4 Disa robusta 4 Satyrium buchananii 1 Preprints (www.preprints.org) | NOT PEER-REVIEWED | Posted: 25 October 2018 doi:10.20944/preprints201810.0587.v1 Peer-reviewed version available at Genes 2018, 9, 595; doi:10.3390/genes9120595 Myala SJK37 Myala and nampanda SJK21 Luapula, Zambia Ntonkonshi SJK25 Democratic Republic of Congo Sumbawanga SJK20 mixed unknown-mixed SJK31 SJK19 SJK8 unknown SJK9 SJK13 421 Kawambwa, Zambia Sumbawanga, Tanzania Serenje, Zambia Luwingu, Zambia Mwinilunga, Zambia; Luwingu, Zambia Kawamba, Zambia Satyrium buchananii 1 Disa welwitschii 2 Satyrium buchananii 1 Disa robusta 1 Disa satyriopsis 1 Disa caffra 1 Disa robusta 1 Habenaria cf sp. DO122 (Clade H. schimperiana, H. kyimbilae, H. microsaccos 1 Satyrium buchananii 6 Satyrium carsonii 1 Habenaria aff. helicoplectrum (BB3151) 1 Disa miniata 1 Disa robusta 2 Disa welwitschii 1 Satyrium buchananii 2 Satyrium carsonii 1 Disa celata 1 Disa welwitschii 1 Satyrium buchananii 3 Preprints (www.preprints.org) | NOT PEER-REVIEWED | Posted: 25 October 2018 doi:10.20944/preprints201810.0587.v1 Peer-reviewed version available at Genes 2018, 9, 595; doi:10.3390/genes9120595 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 were identified as Brachycorythis species and mbwelenge or fake chikanda was made of Satyrium buchananii. However, our previous study on the analysis of Tanzanian chikanda cakes showed that the cake made with fake chikanda tubers also contained Disa miniata, Satyrium anomalum, S. comptum, S. elongatum, S. riparium, S. shirense and S. volkensii, indicating that there might be some differences between fake chikanda samples as well [19]. It is well-known from literature that local species concepts are not necessarily congruent with scientific classifications and that species might be subject to overor underdifferentiation [50,51]. In this case, the grouping of the orchids according to the area of origin, shape or consistency preference or plainly under the general term chikanda is clearly a case of underdifferentiation as a much higher diversity was retrieved when employing DNA barcoding. In order to more reliably identify the orchid species used for a particular chikanda type more samples per local classification need to be analyzed. 4.4. Orchid availability and conservation Througout recent decades chikanda has made a remarkable leap in popularity. The first record of chikanda use made by Audrey Richards [9], described the relish as a poor man’s food, eaten in times of famine. Recent studies, from 2002 onwards, show that chikanda has emerged as a Zambian snack popular throughout the country. Studies on chikanda report that with this rise in popularity, the orchid harvest has escalated and is pressuring local Zambian orchid populations as well as those in neighboring countries [52,12,10,53]. Many people involved in chikanda trade indicated that chikanda plants were becoming scarce, and many were concerned about both the quality as well as the quantity of the orchids available. Our study also confirmed a significant international trade network for chikanda sources in several regions in Tanzania, as well as in the Democratic Republic of Congo and Angola. In our current study, we found at least 16 different orchid species sold as chikanda on the Zambian markets and an overview of previous studies contains 46 species reported to be used for chikanda (Table S7). This brings us from the use of an initial two orchid species reported for chikanda [13] to a total of 49 species in eight different genera. The increased harvesting pressure in combination with the indiscriminate harvesting and use of many more species than earlier assumed pose a threat to nearly half of the terrestrial orchids occurring in these regions. Despite the establishment of Kitulo National Park in Tanzania, with orchid conservation as a prime concern, it seems that harvesting continues even there, since iringe tubers found in this study come specifically from this region [12,53]. Currently there are only seven Disa species from Zambia and surrounding countries registered on the global IUCN Red List and no species from other genera used for chikanda [54]. Most of the orchid species used for chikanda seem to have a widespread distribution, but local populations as well as endemic species could be at risk of overharvesting, and we urge to add the most frequently traded chikanda species, such as Disa robusta and Satyrium buchananii to the IUCN Red List. Although there seems no stopping to commerce, people involved in chikanda trade seem genuinely concerned about welfare of local orchids and interested in exploring other options. Since especially the chikanda harvesters seem to be in a vulnerable position, where they have to rely on surrounding natural resources to secure their livelihoods [16], it is essential that when trying to protect orchids used for chikanda, the situation of the people dependent on the trade is taken into account as well. Currently the development of sustainable cultivation or chikanda orchids is attempted in collaboration with the Cape Institute of Micropropagation (Barrydale, South Africa) and possible alternative sources of income for the people involved in chikanda trade, such as honey production, are being explored [55]. Alternatively, since the purpose of the chikanda orchids mainly is to bind and create an elastic structure to the cake, it might be possible to encourage the use of an alternative source of starch to replace the tubers. 5. Conclusions DNA barcoding using the nuclear ribosomal ITS marker proved to be useful in identifying terrestrial orchid species traded as chikanda on local Zambian markets and outperformed identification using the core land-plant barcoding markers matK and rbcL. Sixteen orchid species, of which three previously undocumented, were identified from marketed chikanda tubers, bringing the Preprints (www.preprints.org) | NOT PEER-REVIEWED | Posted: 25 October 2018 doi:10.20944/preprints201810.0587.v1 Peer-reviewed version available at Genes 2018, 9, 595; doi:10.3390/genes9120595 475 476 477 478 479 480 481 482 483 484 485 486 total number of orchid species used for chikanda to at least 49. The species most frequently found on the markets were Disa robusta, Satyrium buchananii and Platycoryne crocea. However, the results are only as good at the reference material is and an increased reference database in combination with an underpinned phylogenetic framework for Habenaria and related genera would likely ameliorate the reliability of the identifications. Tubers are harvested from various regions in Zambia and Tanzania, and additional international chikanda trade-hubs have been identified on the border with the Democratic Republic of Congo and Angola. People involved in chikanda trade indicate that both orchid quality as well as quantity are decreasing and are willing to consider alternatives to chikanda trade to secure their income. Authors should discuss the results and how they can be interpreted in perspective of previous studies and of the working hypotheses. The findings and their implications should be discussed in the broadest context possible. Future research directions may also be highlighted. 487 488 Supplementary Materials: The following supplementary figures and tables are available online at www.mdpi.com/xxx/s1 489 490 491 492 493 494 495 496 497 498 499 500 Figure S1. bPTP analysis of all chikanda matK query and reference samples. Figure S2. bPTP analysis of all chikanda rbcL query and reference samples. Figure S3. bPTP analysis of all chikanda nrITS query and reference samples. Table S1. Brahms RDE file of novel voucher specimens of orchid taxa samples in this study. Table S2. Genbank accession numbers of the chikanda tuber sequences. Table S3. Hit table with the first 5 BLAST top hits per sample for nrITS. Table S4. Hit table with the first 5 BLAST top hits per sample for matK. Table S5. Hit table with the first 5 BLAST top hits per sample for rbcL. Table S6. Successfully sequenced samples with their vernacular name, reported origin, identification based on sequence-similarity for nrITS and matK, identification based on the tree-based bPTP analysis and the consensus ID. Table S7. Orchid species used for chikanda according to literature. 501 502 503 504 505 506 507 508 509 Author Contributions: S.V., R.B. and H.d.B. conceived and designed the study in collaboration with S.J.K., D.C., R.V. and N.W.; S.J.K. performed the chikanda collection and interviews in collaboration with D.C., G.M., N.W. and R.V; R.B., N.W. and K.Y. collected reference vouchers in the field, which were identified in collaboration with B.B.; B.B, G.N and F.M. provided the reference sequence database for Habenaria species and some species in closely related genera; S.J.K. performed the labwork for the chikanda collections and M.B.F. for the reference collections; S.J.K. performed the data analysis under supervision of S.V. and H.d.B.; S.V. and S.J.K. wrote the manuscript in consultation with the other authors and the manuscript was edited and reviewed in detail by B.G., T.v.A. and H.d.B. 510 511 Funding: This research was funded by the Darwin Initiative (UK Government) Grant No. 23034 and NWO-SIDA-COSTECH TASENE Grant W 02.29.102. 512 513 514 Acknowledgments: The authors would like to thank the participants in the study for sharing their knowledge and Simon Hultby from the Uppsala Botanical Garden for growing the sprouted chikanda orchid tubers. 515 516 517 518 Conflicts of Interest: The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results. 519 References 520 521 522 1. Hinsley, A.; de Boer, H. J.; Fay, M. F.; Gale, S. W.; Gardiner, L. M.; Gunasekara, R. S.; Kumar, P.; Masters, S.; Metusala, D.; Roberts, D. L. A review of the trade in orchids and its implications for conservation. Bot. J. Linn. Soc. 2017. Preprints (www.preprints.org) | NOT PEER-REVIEWED | Posted: 25 October 2018 doi:10.20944/preprints201810.0587.v1 Peer-reviewed version available at Genes 2018, 9, 595; doi:10.3390/genes9120595 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. Kasparek, M.; Grimm, U. European trade in Turkish Salep with special reference to Germany. Econ. Bot. 1999, 53, 396–406. Ece Tamer, C.; Karaman, B.; Utku Copur, O. A traditional Turkish beverage: Salep. Food Rev. Int. 2006, 22, 43–50. Kreziou, A.; de Boer, H.; Gravendeel, B. Harvesting of salep orchids in north-western Greece continues to threaten natural populations. Oryx 2015, 1–4. Ghorbani, A.; Gravendeel, B.; Selliah, S.; Zarré, S.; Boer, H. DNA barcoding of tuberous Orchidoideae: a resource for identification of orchids used in Salep. Mol. Ecol. Resour. 2017, 17, 342–352. Davenport, T. R. B.; Ndangalasi, H. J. Orchid harvest - An assessment of the harvesting and trade of orchid tubers across Tanzania’s Southern Highlands; Wildlife Conservation Society, 2001; p. 23;. Bingham, M. G. Chikanda trade in Zambia. Orchid Conserv. News 2004, 4, 22–25. Veldman, S.; Otieno, J. N.; Andel, T. van; Gravendeel, B.; de Boer, H. J. Efforts urged to tackle thriving illegal orchid trade in Tanzania and Zambia for chikanda production. TRAFFIC Bull. 2014, 26, 47–50. Richards, A. I. Land, labour and diet in Northern Rhodesia: an economic study of the Bemba tribe; Oxford Univ Press: Oxford, UK, 1939; Bingham, M. G. The Lowdown Magazine. 2007,. Temzie Bites Zambian Food | Chikanda African Ham (Polony) Recipe. Retrieved from: https://www.youtube.com/watch?v=mDGKx-GOLoA; 2017; Davenport, T. R. B.; Ndangalasi, H. J. An escalating trade in orchid tubers across Tanzania’s Southern Highlands: assessment, dynamics and conservation implications. Oryx 2003, 37, 55–61. Cribb, P. J.; Leedal, G. P. The mountain flowers of southern Tanzania: a field guide to the common flowers; AA Balkema: Rotterdam, NL, 1982; Nyomora, A. M. S. Distribution and abundance of the edible orchids of the Southern Highlands of Tanzania. Tanzanian J. Sci. 2005, 31, 45–54. Mapunda, L. N. D. Edible orchids in Makete district, the Southern Highlands of Tanzania: distribution, population and status, Master thesis: Swedish Biodiversity Centre, Uppsala Universitet (Sweden), 2007. Challe, J. F.; Price, L. L. Endangered edible orchids and vulnerable gatherers in the context of HIV/AIDS in the Southern Highlands of Tanzania. J. Ethnobiol. Ethnomedicine 2009, 5, 41, doi:10.1186/1746-4269-5-41. Kasulo, V.; Mwabumba, L.; Munthali, C. A review of edible orchids in Malawi. J. Hortic. For. 2009, 1, 133–139. Hamisy, C. W. Development of conservation strategies for the wild edible orchid in Tanzania. Prog. Rep. Rufford Small Grants Found. 2010. Veldman, S.; Gravendeel, B.; Otieno, J. N.; Lammers, Y.; Duijm, E.; Nieman, A.; Bytebier, B.; Ngugi, G.; Martos, F.; van Andel, T. R.; others High-throughput sequencing of African chikanda cake highlights conservation challenges in orchids. Biodivers. Conserv. 2017, 1–18. Hebert, P. D. N.; Cywinska, A.; Ball, S.; de Waard, J. Biological identifications through DNA barcodes. Proc. R. Soc. B 2003, 270, 313–322. Preprints (www.preprints.org) | NOT PEER-REVIEWED | Posted: 25 October 2018 doi:10.20944/preprints201810.0587.v1 Peer-reviewed version available at Genes 2018, 9, 595; doi:10.3390/genes9120595 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 21. Raclariu, A. C.; Heinrich, M.; Ichim, M. C.; Boer, H. Benefits and limitations of DNA barcoding and metabarcoding in herbal product authentication. Phytochem. Anal. 2017. 22. ISE International Society of Ethnobiology Code of Ethics (with 2008 additions). Available online: http://ethnobiology.net/code-of-ethics/. 23. Goodman, L. A. Snowball sampling. Ann. Math. Stat. 1961, 148–170. 24. Martos, F.; Le Péchon, T.; Ngugi, G.; Pailler, T.; Bellstedt, D. U.; Linder, H. P.; Bytebier, B. Phylogenetic relationships amongst the African genera of the tribe Orchideae (Orchidaceae). Manuscript in prep. 2018. 25. Yoon, C. S.; Glawe, A.; Shaw, P. D. A method for rapid small-scale preparation of fungal DNA. Mycologia 1991, 83, 835–838. 26. Shepherd, L. D.; McLay, T. G. Two micro-scale protocols for the isolation of DNA from polysaccharide-rich plant tissue. J. Plant Res. 2011, 124, 311–314. 27. Ghorbani, A.; Gravendeel, B.; Selliah, S.; Zarre, S.; de Boer, H. J. DNA barcoding of tuberous Orchidoideae: A resource for identification of orchids used in Salep. Mol Ecol Resour 2016, doi:10.1111/1755-0998.12615. 28. Kress, W. J.; Wurdack, K. J.; Zimmer, E. A.; Weigt, L. A.; Janzen, D. H. Use of DNA barcodes to identify flowering plants. Proc. Natl. Acad. Sci. 2005, 102, 8369–8374. 29. Dunning, L. T.; Savolainen, V. Broad-scale amplification of matK for DNA barcoding plants, a technical note: AMPLIFICATION OF matK FOR DNA BARCODING PLANTS. Bot. J. Linn. Soc. 2010, 164, 1–9, doi:10.1111/j.1095-8339.2010.01071.x. 30. Sun, Y.; Skinner, D. Z.; Liang, G. H.; Hulbert, S. H. Phylogenetic analysis of Sorghum and related taxa using internal transcribed spacers of nuclear ribosomal DNA. Theor. Appl. Genet. 1994, 89, doi:10.1007/BF00226978. 31. Bonfield, J. K.; Smith, K. f; Staden, R. A new DNA sequence assembly program. Nucleic Acids Res. 1995, 23, 4992–4999, doi:10.1093/nar/23.24.4992. 32. Staden, R. The Staden sequence analysis package. Mol. Biotechnol. 1996, 5, 233–241. 33. Larsson, A. AliView: a fast and lightweight alignment viewer and editor for large datasets. Bioinformatics 2014, 30, 3276–3278, doi:10.1093/bioinformatics/btu531. 34. Zhang, J.; Kapli, P.; Pavlidis, P.; Stamatakis, A. A general species delimitation method with applications to phylogenetic placements. Bioinformatics 2013, 29, 2869–2876, doi:10.1093/bioinformatics/btt499. 35. Stamatakis, A.; Hoover, P.; Rougemont, J. A rapid bootstrap algorithm for the RAxML web servers. Syst. Biol. 2008, 57, 758–771, doi:10.1080/10635150802429642. 36. Hamisy, C. W. Development of conservation strategies for the wild edible orchid in Tanzania. Prog. Rep. Rufford Small Grants Found. 2010. 37. Davenport, T. R. B.; Bytebier, B. Kitulo Plateau, Tanzania-a first African park for orchids. Orchid Rev. 2004, 112, 161–165. 38. Boessenkool, S.; Mcglynn, G.; Epp, L. S.; Taylor, D.; Pimentel, M.; Gizaw, A.; Nemomissa, S.; Brochmann, C.; Popp, M. Use of Ancient Sedimentary DNA as a Novel Conservation Tool for High-Altitude Tropical Biodiversity. Conserv. Biol. 2014, 28, 446–455, doi:10.1111/cobi.12195. 39. van Geel, B.; Aptroot, A.; Baittinger, C.; Birks, H. H.; Bull, I. D.; Cross, H. B.; Evershed, R. P.; Gravendeel, B.; Kompanje, E. J.; Kuperus, P. The ecological implications of a Yakutian mammoth’s last meal. Quat. Res. 2008, 69, 361–376. Preprints (www.preprints.org) | NOT PEER-REVIEWED | Posted: 25 October 2018 doi:10.20944/preprints201810.0587.v1 Peer-reviewed version available at Genes 2018, 9, 595; doi:10.3390/genes9120595 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 40. Elder, J. F.; Turner, B. J. Concerted Evolution of Repetitive DNA Sequences in Eukaryotes. Q. Rev. Biol. 1995, 70, 297–320, doi:10.1086/419073. 41. Ganley, A. R. D.; Kobayashi, T. Highly efficient concerted evolution in the ribosomal DNA repeats: Total rDNA repeat variation revealed by whole-genome shotgun sequence data. Genome Res. 2007, 17, 184–191, doi:10.1101/gr.5457707. 42. Harpke, D.; Peterson, A. Non-concerted ITS evolution in Mammillaria (Cactaceae). Mol. Phylogenet. Evol. 2006, 41, 579–593, doi:10.1016/j.ympev.2006.05.036. 43. Feliner, G.; Rosselló, J. Better the devil you know? Guidelines for insightful utilization of nrDNA ITS in species-level evolutionary studies in plants. Mol. Phylogenet. Evol. 2007, 44, 911–919. 44. Zheng, X.; Cai, D.; Yao, L.; Teng, Y. Non-concerted ITS evolution, early origin and phylogenetic utility of ITS pseudogenes in Pyrus. Mol. Phylogenet. Evol. 2008, 48, 892– 903, doi:10.1016/j.ympev.2008.05.039. 45. Xu, B.; Zeng, X.-M.; Gao, X.-F.; Jin, D.-P.; Zhang, L.-B. ITS non-concerted evolution and rampant hybridization in the legume genus Lespedeza (Fabaceae). Sci. Rep. 2017, 7, 40057, doi:10.1038/srep40057. 46. Chen, S.; Yao, H.; Han, J.; Liu, C.; Song, J.; Shi, L.; Zhu, Y.; Ma, X.; Gao, T.; Pang, X.; Luo, K.; Li, Y.; Li, X.; Jia, X.; Lin, Y.; Leon, C. Validation of the ITS2 region as a novel DNA barcode for identifying medicinal plant species. PLoS ONE 2010, 5, 1–8, doi:10.1371/journal.pone.0008613. 47. Mezzasalma, V.; Bruni, I.; Fontana, D.; Galimberti, A.; Magoni, C.; Labra, M. A DNA barcoding approach for identifying species in Amazonian traditional medicine: The case of Piri-Piri. Plant Gene 2017, 9, 1–5, doi:10.1016/j.plgene.2016.11.001. 48. Bytebier, B.; Dirk U. Bellstedt; H. Peter Linder A New Phylogeny-Based Sectional Classification for the Large African Orchid Genus Disa. Taxon 2008, 57, 1233–1251. 49. van der Niet, T.; Linder, H. P. Dealing with incongruence in the quest for the species tree: A case study from the orchid genus Satyrium. Mol. Phylogenet. Evol. 2008, 47, 154–174. 50. Cunningham, A. B. Applied ethnobotany: people, wild plant use and conservation; Earthscan: London, 2001; ISBN 1853836974. 51. Martin, G. J. Ethnobotany: a methods manual; Earthscan: London, UK, 2004; 52. Bingham, M. G.; Smith, P. P. Southern African plant red data lists: Zambia; Golding, J. S., Bandeira, S. O., Eds.; SABONET: Pretoria, South Africa, 2002; 53. Veldman, S.; Otieno, J. N.; Andel, T. van; Gravendeel, B.; de Boer, H. J. Efforts urged to tackle thriving illegal orchid trade in Tanzania and Zambia for chikanda production. TRAFFIC Bull. 2014, 26, 47–50. 54. IUCN IUCN Red List of Threatened Species. Version 2017-3.; 2018; 55. Bone, R. E.; Wightman, N.; Vinya, R.; Veldman, S.; Yokoya, K.; Hargreaves, S.; Kendon, J.; Crous, H. Edible wild orchid trade: sustaining livelihoods and biodiversity in Zambia. Darwin Initiative Main Project Annual Report.; 2017;