Wild Plants Used as Vegetables by Transhumant People Around the Georgia–Turkey Border in the Western Lesser Caucasus

Recent ethnobotanical studies in the Caucasus, mainly in Georgia, reveal the significant ethnobotanical knowledge of local people related to wild edible plants. However, few studies have been conducted in the Lesser Caucasus, west Georgia, and Turkish Caucasus. is study aims to represent and evaluate the cultural importance of wild vegetable plants and their patterns of use along the Georgia–Turkey border. During the transhumance period in the summers of 2017 and 2018, 104 participants (65 in Turkey and 39 in Georgia) were interviewed using a semistructured questionnaire. e Cultural Importance Index and Relative Frequency of Citation were used to compare the relative importance of species in each region. e use of 83 wild plant species from 23 plant families as vegetables was documented, with 45 species recorded in Georgia and 72 species in Turkey. One-third of the recorded wild plant species and 52 use instances out of 122 species-use combinations were shared on both sides of the border. Women and men had mentioned almost the same number of species, and there was a nonsignificant correlation between the plant knowledge and age. Although there were no significant differences in the plant parts used, the way people used plants as vegetables varied significantly across the border. Considering the floral similarity across the border, the number of species used in common and shared vegetable plant knowledge was quite low. ere is not a significant difference between the two countries in terms of the most frequently cited and culturally important species (Rumex, Urtica, and Polygonum spp.). However, the recognition of some of the most important shared species (Heracleum, Chaerophyllum, Arctium, and Campanula spp.) diverged significantly in different administrative regions.


Introduction
Human communities worldwide have collected and consumed many wild plants as food throughout history (Turner et al., 2011). However, with the decrease in gathering and use of these plants as a result of lifestyle changes, many observations, practices, and a significant amount of traditional knowledge related to wild edible plants is being forgotten (Łuczaj, 2010;Menendez-Baceta et al., 2017;Turner & Turner, 2008). Concurrently, in industrialized societies, several plant species are being appreciated as healthy delicacies, and plant gathering traditions continue as a recreational activity Reyes-García et al., 2015; with small-scale agriculture and relatively large-scale livestock farming as their main economic activities. Nearly all participants in this study were transhumant, maintaining an agropastoral way of life. Highland pastures, referred to as "yayla" in Turkey, are known as "mta" and/or "ialagi" (iala) in Georgia. People move to their summer pastures at the end of May, where they live mainly in wooden houses for 3 to 5 months, with some people living in dry stone dwellings or even tents.

Ethnobotanical Data Collection
To restrict the focus of the study on the ethnobotanical knowledge of transhumant people, more than two-thirds of the fieldwork was conducted in highland pastures along the Georgian-Turkish border. Firstly, over 150 potential highland pastures were identified between altitudes of 1,600 m and 2,500 m within the study area using Google Earth. Subsequently, possible research locations were selected from among those settlements according to a number of geographical barriers (mountains, rivers, lakes, and passes) that would help identify a high diversity of floral and cultural characters. We attempted to reach people who had maintained their agropastoral transhumance lifestyle, as they would have been in contact with a variety of vegetation types during regular seasonal migrations, thus having a relatively strong living memory of traditional knowledge and practices related to wild plants.
In the summer of 2016, 2 weeks of nonsystematic preliminary fieldwork was conducted, and informal interviews were conducted in 20 highland pastures and villages in Georgia and Turkey (Oruç & Kazancı, 2018). Over the following two summers (2017)(2018), field work was carried out for approximately 90 days during the period of transhumance (approximately June 15-September 15), which involved visits to 102 highland pastures, 65 in Turkey, and 37 in Georgia (Figure 1). During that period, 104 participants were interviewed: 65 in Turkey and 39 in Georgia. e mean ages of the participants were 57 and 58 years in Georgia and Turkey, respectively. e presented results are part of a larger research project involving medicinal plants. e results that refer to the use of medicinal plants were gathered using the same methodology and oen from the same respondents, and were published recently (Kazancı et al., 2020). e initial investigation considered the flora in different vegetation zones (forest, meadow, wetlands, steppe, and rocky areas) en route to and in the vicinity of each selected highland pasture. is process took approximately 2-3 hours for each highland pasture. is reconnaissance involved the identification of wild plants to the species level when possible, in which photographs and a minimum of three samples were collected for each plant. is had a dual purpose: first, to enable the participants to be shown plants to identify and discuss, and second, to prepare herbarium voucher specimens for later detailed identification. e research team was comprised of three or four people. e first (female) and second (male) authors were always involved in the interviews, together with a translator, either male or female. roughout the study, the first author was the principal interviewer. In Turkey, the interviews were conducted in the Turkish language, while in Georgia, interviews were conducted in Georgian, Russian, or Turkish. A majority of the interviews in Georgia were performed with the help of translators who spoke Georgian, Russian, and English, either as a mother tongue or as a second language. e interviews were translated into English. e translators were provided with information and terminology relevant to the research 2 weeks before the fieldwork. Information regarding the purpose of the study was provided to all participants and their free, prior informed consent for interviewing, recording, photographing, and/or publishing their knowledge was obtained orally from each participant at the beginning of their interviews. All interviews conformed to the International Society of Ethnobiology's Code of Ethics (International Society of Ethnobiology, 2006  the person's relatives and neighbors also contributed to the responses in the interviews. e first author took notes directly in a notebook during the interviews. Depending on the preferences of the participants, audio or video recordings were made during the interviews. Information about plants collected from the wild was documented, specifically with data regarding their folk names in different languages and dialects, collection time and place, parts used, processes of preparation, and sources of plant knowledge. In addition, observations were made and photographs were taken in byres, cellars, and other relevant places whenever possible, so as to document unmentioned uses and to observe ethnobotanical practices that were still in use. Initially, the participants were asked to discuss points about wild plants that they immediately thought of (ca. 15 min). ey were then shown fresh plants and asked to identify the vernacular names and usage of the plants (ca. 45 min). Depending on the weather and the participants' willingness, a "walk around the house" was undertaken to observe wild plants in the vicinity (ca. 15 min). To confirm previous information and to gain further learning about various plants, participants were shown an illustrated plant catalog, including 400 plant species from the flora of the region (ca. 45 min). Certain participants were visited a second time to complete the first interview or to confirm the information (Figure 3).

Taxonomic Identification of Plants
e preliminary identification of plant species was carried out in the field by the authors. e plants were photographed together with their coordinates, and voucher herbarium specimens were prepared by the first author for further identification. Relevant flora resources were used for identification (Davis, 1965(Davis, -1985Davis et al., 1988;Güner et al., 2000;K' etskhoveli, 1941-1952K' etskhoveli et al., 1971Komarov, 1968Komarov, -2002. Some of the species were identified through detailed plant and habitat descriptions and previously recorded folk names. Specimens identified in Georgia were stored in the National Herbarium of Georgia (TBI) in Botanical Institute of Ilia State University, while specimens identified in Turkey were stored in the Herbarium of Artvin Çoruh University. Species names were based on the currently accepted names from e Plant List (http://www.theplantlist.org/). Furthermore, plant synonyms were provided based on Güner et al. (2012) and Gagnidze (2005).

Data Analysis
Firstly, all reported plant species and their relevant ethnobotanical data were entered into a Microso Excel spreadsheet in a use-report (UR)-based order. Each use in every use-category was counted as one UR. In this study, ethnobotanical data of only wild (noncultivated) plants consumed as vegetables during the fieldwork from 2017 and 2018 were considered in the analysis. We included all of the green aboveground Acta Societatis Botanicorum Poloniae / 2021 / Volume 90 / Article 906 Publisher: Polish Botanical Society parts of plants as well as their underground storage organs (roots and bulbs) in the category of wild vegetables, regardless of whether they were used raw or processed by cooking or frying. ese include wild green vegetables and root vegetables that were boiled and/or stewed for dishes, soups, pastries, pickles, wraps, and eaten as snack vegetables. ese do not include beverages, fruits, sweets, or spices.
Two indices were used to compare the relative importance of species and the ways in which they were used in each region: • Relative Frequency of Citation (RFC) (Tardío & Pardo-de-Santayana, 2008 e Pearson correlation coefficient was used to measure the correlation between the age of the participants and the number of species, as well as the age of the participants and the number of use reports. Student's t test was used to determine the relationship between gender and the number of plant species known (initially, the normality of data was checked with the Shapiro-Wilk test). e chi-square test was used to compare the plant parts used, and the methods of preparation mentioned between the locations on both sides of the border. e Mann-Whitney U test was used to test the differences between the mean number of species used in both countries.

Overall Results
Altogether, 1,154 citations of 83 wild plant species used as vegetables were obtained in the study area, of which 45 species were recorded from Georgia and 72 species were recorded from Turkey (Table 1). More than a third of the recorded species (34) were shared on both sides of the border ( Figure 4A).
In the comparison of the 122 species-use combinations, participants in both countries shared similar knowledge of 52 use instances (43% of total species-use combinations) ( Figure  Overall, 83 wild plant species from 23 families were recorded in this study. e best represented families were Apiaceae (11 species) and Asteraceae (seven species) in Georgia, and similarly, Apiaceae (14 species) and Asteraceae (12 species) in Turkey. e most frequently cited genera by more than half of the participants were Rumex, Urtica, and Polygonum in Georgia, and Polygonum, Rumex, Urtica, Heracleum, Chaerophyllum, and Arctium in Turkey. Although Chaerophyllum (RFC 0.75), Arctium (RFC 0.45), and Heracleum (RFC 0.30) species had considerable mentions in Samtskhe-Javakheti in Georgia, they were scarcely recognized by participants in Adjara.
In summary, we rejected the first two hypotheses presented in the introduction. e number of species used was significantly higher on the Turkish side, and there  was no significant correlation between the age of the participants and the number of species listed. However, as we predicted, both genders had similar knowledge of wild vegetables.

Comparison Between Four Administrative Regions
e highest number of plants were mentioned in Ardahan (64 species), followed by Artvin (39), Samtskhe-Javakheti (37), and Adjara (21 species). Fourteen plant species were reported throughout the entire study area in each administrative region, 12 of which were also found among the 15 most culturally important species (Figure 7), namely Urtica dioica, Rumex crispus, Chaerophyllum spp., Polygonum bistorta, Rumex acetosa, Rumex acetosella, Heracleum spp., Polygonum cognatum, Anthriscus spp., Campanula lactiflora, Capsella bursa-pastoris, and Crocus spp. e most popular genera used by the majority of the participants (RFC ≥ 0.7) in all regions were Urtica spp. (mostly Urtica dioica) and Rumex spp. Urtica spp. have similar cultural significance, with comparable CI values in each administrative region (Figure 7). While Rumex crispus is used mainly in the Ardahan and Samtskhe-Javakheti regions, R. acetosella and R. acetosa are popular chiefly in the Ardahan and Adjara regions. Polygonum bistorta was popular in Artvin, Ardahan, and Adjara (RFC ≥ 0.7) but scarcely reported in Samtskhe-Javakheti (RFC 0.05). Chaerophyllum spp. have significant and comparable cultural importance in Ardahan, Samtskhe-Javakheti, and Artvin, but a very low CI value in Adjara. ere were also several species that were exclusively mentioned in one administrative region, such as Campanula lactiflora (RFC 0.53) and Aruncus vulgaris (RFC 0.32) in the Adjara region. Sempervivum species (RFC 0.53) were also popular only in Ardahan. ey occur in rocky habitats at high altitudes, have mostly been appreciated by shepherds, and are called "çoban ekmeği" in Turkish, which literally  translates to "shepherd's bread". Similarly, Echinops and Tragopogon species are highly popular in Ardahan and Samtskhe-Javakheti, but not in Artvin or Adjara.

Plant Parts Used and Preparation Methods
Wild vegetables are consumed in various ways in the study area. Young aerial parts are consumed in almost every possible way. e plant parts used were not significantly different across the border (chi-squared test value = 2.96, df = 5, p = 0.71). e cultural importance of use categories and the number of species used for each category in the four administrative regions are compared in Figure 8. A significant difference in plant use categories across the border was found (chi-squared test value = 60.7, df = 5, p < 0.05). e main methods of consumption in the study area were raw, as a snack, and processed. Raw snacks are more frequently used in Turkey than in Georgia. is is especially apparent in Ardahan, the only region where consumption as a raw snack (74% of the reported genera) predominates over processed consumption (63%). e young aerial parts of Rumex and Tragopogon, young stems of Heracleum, young stems and underground parts of Arctium, underground parts of Chaerophyllum, receptacles of Echinops, and leaves of Sempervivum are the preferred snack vegetables in Ardahan. On the other hand, raw consumption is uncommon in Adjara, where 50% of the genera are consumed raw, whereas 83% of them are processed.
e most common preparation methods on both sides of the border were as boiled and/or stewed dishes. Young leaves or aerial parts of Polygonaceae and Urticaceae family members were mostly mentioned in this category. Urtica dioica, Rumex crispus, and Polygonum bistorta were the preferred species in Georgia. Urtica dioica, Polygonum bistorta, P. aviculare, P. cognatum, and Rumex crispus were the most widely used vegetable species in Turkey. "Pkhali" is the main boiled dish with various recipes mentioned by Georgia participants in this category. e two most common recipes in this study included boiling, squeezing, and chopping the plants, then either stewing them with onion in oil or occasionally with eggs, or mixing them together with garlic, coriander, and walnut. It has been reported that pkhali recipes can be prepared with a single plant species, usually with a mixture of several spring plants, such as Aruncus vulgaris, Campanula lactiflora, Capsella bursa-pastoris,  Chaerophyllum spp., Chenopodium album, Convolvulus arvensis, Polygonum aviculare, P. bistorta, Rumex acetosa, R. crispus, and Urtica dioica. Although the above recipes are shared among participants on both sides of the border, dishes in Turkey are also prepared by stewing plants with bulgur or rice aer mixing them with fried onion in Turkey.
Ardahan was distinguished from other regions by its larger number of root vegetables. Six genera were appreciated for their underground parts. e most widely used species based on RFC were Chaerophyllum, Crocus, and Arctium.
Lacto-fermented pickling is another widely applied form of consumption and is represented by 30 species in this study. Young aerial parts of Chaerophyllum, Heracleum, and Anthriscus species are used in all regions, but there were few reports from Adjara. Local people were observed selling pickles prepared from these plants along the roadsides, especially in the Samtskhe-Javakheti region (Figure 9). ere were also several species exclusively used in one region, such as the peeled stems and leaves of Trachystemon orientalis in Artvin, young aerial parts of Campanula lactiflora, Aruncus vulgaris, Polygonum bistorta, and Silene spp. in Adjara. e most common and rapid method for pickling in Turkey is preserving slightly boiled and drained plants in whey "şırat" and salt, and occasionally with garlic. However, whey (liquids le aer cheese making) is rarely used in Georgia, where pickles are produced through lactic fermentation in salty water with other flavorings such as garlic, pepper, dill, coriander, and parsley.
Both fresh and dried (for winter) consumption of young aerial parts of plants in soups were mostly preferred in Turkey. Young aerial parts of Urtica dioica and Polygonum bistorta were widely used in soups in Artvin and Ardahan. e soup of Polygonum bistorta has a particularly special significance in transhumant life in Artvin (Figure 10), where several summer festivals have been named aer the plant, which they call "yayla pancarı. " Although people used to gather this plant and eat the "yayla pancarı" soup together for healing, nowadays the plant is used individually and remains a symbol through the names of the festivals. Several participants transplanted the species and grew them in their highland homegardens. Similarly, in Georgia, it is known for its delicious and healthy leaves, which are used with milk and cream to make a unique soup. is soup is considered as a panacea. Rumex crispus is another popular species used in soups in the Ardahan and Samtskhe-Javakheti regions. Fresh or mainly dried braided leaves of the plant were prepared as a winter soup (Figure 10). Boiled leaves were mixed with dairy products (milk, cream, ayran, yogurt, cheese, or whey) and/or flour or cereal varieties (corn, wheat, and rice). Alternatively, this soup mixture was prepared with eggs.  e wrap (sarma) is also a common method of consumption only in Turkey. irteen species were used in this way in Turkey. e most common species were Rumex spp. (mostly R. crispus) and Plantago major in Ardahan and Artvin. Generally, bulgur or rice were wrapped in the leaves of these species. Arctium spp. were also preferred mostly in Ardahan. On the other hand, only two species (Rumex crispus and Plantago major) were mentioned, and only rarely, as being used to prepare wraps (sarma) in Samtskhe-Javakheti in Georgia. Some recipes reported in the study area include the leaves of Caltha palustris consumed as a wrap (sarma).
Although infrequently, young aerial parts of Chenopodium album and Urtica dioica species were used in stuffing for pastry (katmer, börek) in Turkey. No wild vegetable species for stuffing pastry in Georgia were reported in this study.

Discussion
Regional differences in the number of plant species recorded in each administrative region may be partly due to the differences in the number of interviews carried out as well as habitat types of each region. Communities living in open lands (in Ardahan and Samtskhe-Javakheti) seem to use more diverse wild vegetable plants than communities living close to the forests (in Artvin and Adjara).
Considering the floral similarity across the border, the number of common species used and vegetable plant knowledge shared among communities was quite low. However, there was not much difference in terms of the most frequently cited and culturally important species with several uses. Nevertheless, some of the most important shared species (Heracleum, Chaerophyllum, Arctium, and Campanula spp.) diverge significantly in recognition on a regional scale, between Adjara and  Samtskhe-Javakheti. is might be due to differences in the frequencies of species occurences and cultural differences among communities in these regions. Although two of the most popular vegetable species (Urtica dioica and Rumex acetosa) according to this study were also frequently reported previously from the Caucasus region, the most frequently cited species (Polygonum bistorta) in this study was earlier reported in only three studies in Turkey (Güneş & Özhatay, 2011;Kadıoğlu et al., 2020;Saraç et al., 2013)  eir popularity in the study area, in comparison with that in other studies, could be related to the way of life of this study's specific participants (transhumant communities) as well as their intense contact with diverse habitats such as dense forests, highlands, and high rocky places for their livestock.
Among the species mentioned above, Campanula lactiflora and Aruncus vulgaris appear to be specific to Adjara, even though they are sometimes used in the Artvin region as well. ese species have similar uses and are known by the same local names in the two regions. is convergence might be related to similar forest habitats in these regions, as well as the similarity in cultural background of the communities living there.
e high popularity of Sempervivum in Ardahan and Echinops as snack vegetables in Ardahan and Samtskhe-Javakheti, but not in the other two regions, may be related to the predominance of seminomadic pastoral life in these regions. Since the consumption of raw plants in the field has been associated with mobile pastoralism , the highest number of mentioned species with dominance of consumption of raw plants as a snack in Ardahan could be related to the existence of diverse pastoral communities. In addition, having access to relatively larger agricultural lands may enable the agropastoralists of Ardahan to come into contact with diverse synanthropic plant species such as Atriplex, Bunias, Sisymbrium, and Sinapis. On the other hand, the presence of small-scale, agropastoral-based communities with vegetable home gardens, even in the highlands, might be the reason for the lower dependency on wild vegetables in Adjara. Wild plant snacks are an oen an overlooked type of wild food, but more detailed wild food studies oen feature them (e.g., Kaliszewska & Kołodziejska-Degórska, 2015). ey are oen associated with childhood, especially in areas where children play an important role in guarding grazing animals (Łuczaj & Kujawska, 2012).
Caltha palustris is usually regarded as a toxic plant and has an acrid taste when eaten unprocessed. It has been reported to be used in food several times in the relevant literature (Çakır, 2017;Kadıoğlu et al., 2020;Kang et al., 2014;Pieroni et al., 2020). In the case of Caltha, the toxic agent is protoanemonine, which is broken down by prolonged boiling or drying (Aslam & Ijaz, 2012). e use of Colchicum (a genus that is also regarded as very toxic) for pickles should be investigated, as this genus is usually regarded as highly toxic.
e results of this study are yet another example of the high cultural importance of lacto-fermented foods in the Caucasus (Nanagulyan et al., 2020). e number of wild vegetables pickled in this way is higher than that recorded in the review of fermented foods of Eastern Europe (Sõukand et al., 2015), according to which only cultivated vegetables are pickled in those areas.
Crocus species were the most common root vegetable species in the highlands of all four regions. In our study area, they are historically known as a signature of seasonal change and are considered indicators of oncoming cold days or snow. eir blooming calls for "the end of yayla season" or "the time to migrate down. " Before going down from the highlands to the villages, the bulbs are collected and mostly boiled or cooked in the stove, but are also sometimes eaten raw (Figure 11).
A very interesting feature of the studied food culture is a relatively long list of species whose underground organs are used as snacks. is is similar to the observations made by  among Kurds. e wide use of underground organs can be seen as a very archaic cultural feature that can be practiced more easily in mountainous areas with low population density. Knowledge of edible bulbs and roots is probably of high importance to the survival of shepherds, giving them access to plant calories in seasons during which grassland geophytes are dormant.

Conclusion
is study indicates a significantly diverse ethnobotanical knowledge heritage regarding wild vegetable plants among transhumant communities in the Western Lesser Caucasus. Having been in contact with various environments and with other communities for centuries has enabled these communities to create diverse knowledge. However, despite the richness of species used as vegetables, very few species are actually used by the majority of people in these communities. Knowledge of most of the species, as well as a rich local cuisine, seem to be at a risk of disappearing due to industrialization and disrupted intergenerational bonds. Publications, such as books and articles on wild edible plants and practical workshops (elders and children plant gathering and preparation tours) on wild food plants during local festivals are vital for the transmission of this knowledge to future generations. More importantly, supporting the local products of transhumant communities could help maintain this way of life.