Abstract

The checklist includes 270 species that belong to 80 families and 183 genera. Each species has been annotated with the following information: life form, wetland indicator status, and location. In this checklist, Angiosperms are represented by 252 species (93.3%), Bryophytes – 10 species (3.7%), Pteridophytes – 8 (3%), Gymnosperms – 1 (0.4%). The largest families by the number of species are Cyperaceae – 39 (14.4%), Poaceae – 29 (10.7%), Rosaceae – 19 (7.1%), Asteraceae – 17 (6.3%), Fabaceae – 11 (4.1%) and Juncaceae – 11 (4.1%). The checklist is dominated by 55 Palaearctic species (20.4%), followed by 46 Holarctic (17.1%), 31 Euro-Mediterranean (11.5%), 31 Cosmopolitan (11.5%), and 27 Euro-Siberian (10.03%) species. The endemism rate is 4.8%, and the proportion of invasive and naturalized plants is 8.5%. Obligate wetland plants, mainly belonging to the families Cyperaceae and Juncaceae, make up 34.2% of the floristic composition. This is the first comprehensive published checklist of the flora of Georgian wetlands, annotated with wetland indicator values.

Chorotypes, Floristic composition, Life Forms, Wetlands, Wetland Indicator Status

Wetland ecosystems are considered a natural resource of global importance. They provide many valuable ecosystem services, among them flood protection, water quality improvement, and carbon dioxide sequestration. Wetlands are home to thousands of plant and animal species (Verhoeven et al. 2006; Ramsar Convention Secretariat 2013), including endemic, relict, and rare species (Suppl. material 1, figs 1–3). Furthermore, the aesthetic and recreational benefits of wetlands are also important (Suppl. material 2, figs 1–3). Hence, the preservation of this ecosystem is of the utmost importance (Hazeu et al. 2010; Tedoradze et al. 2020). Many wetlands are threatened by eutrophication (Suppl. material 2: fig. 4) driven by both global and local factors, as well as habitat fragmentation and degradation (Verhoeven et al. 2006). Wetlands harbor many characteristic species, such as hydrophytes (Suppl. material 3, figs 1–2), which are often used as specific indicators for assessing the current condition of wetlands (Reed and Peters 1988; Lyon and Lyon 2011; Eggers and Reed 2015). Therefore, having a comprehensive checklist of wetland flora is the first necessary step for planning and executing the monitoring and conservation of wetlands in Georgia.

Georgia is a mountainous country adjacent to the sea, and such settings create steep climatic gradients, in which geographical location plays an important role (Kikvidze 2020). Therefore, we also performed a gradient analysis of the species composition of wetland communities across Georgia using geographical coordinates and altitude as environmental variables.

Here we present the recorded and assessed 270 species as a list of plants found in different wetlands in most regions of Georgia. It should be noted that number of studies have been published on wetlands in Georgia, relating to different periods (Kimeridze 1966, 1985; Melia 1985; Kikvidze and Oshawa 2001; Denk et al. 2001; BSSGW 2008; Matchutadze 2008; Krebs et al. 2009; Matchutadze et al. 2015). However, there are practically no studies covering all the major wetlands, nor do we have any studies on the indicator status of typical wetland species recorded. Therefore, our study addresses these two needs.

Georgia is a climatically diverse country owing to the adjacent Black Sea and high mountain chains of various orientations. Through our sampling regions, the mean annual temperature varies from 14 °C in Colchis to 3 °C in Kazbegi; annual precipitation varies roughly between 600 and 900mm and its distribution changes from almost ideally uniform in Imereti (760mm ± 15SD) to strongly seasonal in Sioni (812mm ± 45SD). Therefore, the study regions represent quite a diverse climatic character (Fick and Hijmans 2017).

Study regions are located at different hypsometric heights and in different climatic zones. In total, we sampled 218 plots in seven regions across Georgia (Fig. 1): Kolkheti Lowland (121 plots), Javakheti Plateau (53 plots), mountainous Adjara – the highlands of Adjara (8 plots), Kazbegi region (7 plots), Iori River Gorge – the northwestern part of Kakheti province (9 plots), Upper Imereti (6 plots), Shida Kartli Lowland (6 plots), and Alazani Valley (8 plots).

Figure 1. 

Map showing study sites/plots in Georgia.

Fieldwork campaigns were conducted during the vegetation seasons of 2015-2017. The sampling locations were selected prior to our fieldwork using topographic maps of the Soviet period (1950-60s), which depicted mires, swamps, and other wetlands in a rather detailed manner. Before sampling, we determined the location and morphology of wetland bodies with 5m accuracy using RapidEye satellite multispectral imagery.

Sampling points on the pre-selected sites were located using navigation instruments − GPS and satellite images. First, we described the hydrological conditions of a given wetland by assessing its water surface: water not visible (0), well visible (1), and partly visible (2); we also assessed observable human impacts through three categories (low, intermediate, and high impacts).

Vegetation was sampled as follows: once a wetland was recognized and inspected, the team chose visually the most representative area(s) for sampling. A central point of circular plots was placed at random in a selected area, and radii for plots were delineated: a 9 meter-radius circular plot around a sampling point was used to sample trees; a 6 meter-radius plot was used for sampling shrubs, and a 3 meter-radius circle was used for sampling herbaceous plants (USACE 1987). Within these plots, we identified and recorded all detectable species and assessed their abundance by the modified scale of Braun-Blanquet (r, +, 1, 2m, 2a, 2b, 3, 4, 5) (Braun-Blanquet 1964; van der Maarel 2007). Plant species were identified using the Key to Plants of Georgia (Ketskhoveli et al. 1964, 1969) and the Excursion Flora of Germany (Exkursionsflora von Deutschland) (Jäger et al. 2013). The plant species that were not easily identifiable in the field were collected and later accurately identified by comparing them to the specimens in the National Herbarium (TBI) of the Institute of Botany, Ilia State University, Georgia.

In all, we collected 216 herbarium specimens of 132 plant species (herbarium specimens will be available at the TBI). Latin names of taxa (families, genera, and species) are based on nomenclatural checklists of ‘Flora of Georgia’ (Gagnidze 2005; Davlianidze et al. 2018) and are harmonized with international plant databases (Euro+Med 2006; GBIF 2022; IPNI 2022; POWO 2022; WFO 2022). Species that are alien to Georgia are listed according to Kikodze et al. (2010).

Each species was ranked according to its wetland indicator status on the following scale (USACE 1987; Reed and Peters 1988; Lyon and Lyon 2011; Lichvar et al. 2013; Eggers and Reed 2015):

• Obligate wetland plants (OBL) – almost always occur in wetlands (estimated probability > 99%) under natural conditions.
• Facultative wetland plants (FACW) – usually occur in wetlands (estimated probability 67% – 99%), but is occasionally found in non-wetlands.
• Facultative plants (FAC) – equally likely to occur in wetlands (estimated probability 34% – 66%) or non-wetlands.
• Facultative upland plants (FACU) – usually occur in non-wetlands (estimated probability 67% – 99%), but are occasionally found in wetlands (estimated probability 1% – 33%).
• Obligate upland plants (UPL) – almost always occur (estimated probability > 99%) in non-wetlands under natural conditions.

Each species was assessed for its phytogeographical distribution (Komarov 1934-1960; Takhtajan 1954-2011, 2003-2012; Tutin et al. 1964-1993; Davis 1965-2001; Ketskhoveli et al. 1971-2011) and online international plant databases (GBIF 2022; Euro+Med 2006; POWO 2022). In addition, we used different regional works in this field to determine the chorotypes of the species (Gagnidze and Kemularia-Nathadze 1985; Portenier 1993, 2000; Gagnidze 2004; Shetekauri and Chelidze 2016; Lachashvili et al. 2020; Lachashvili et al. 2021; Tedoradze 2022). Plant life forms were determined using the LEDA Traitbase (Klotz et al. 2002). Endemics of the Caucasus are given after Solomon et al. (2013).

To analyze the variation of species composition on geographical gradients, we performed Canonical Correspondence Analysis (CCA). Statistical significance was computed with 999 permutation tests. We analyzed species with a frequency of occurrence in the community matrix equal to or greater than 5. For this purpose, we used PAST (v. 4.12) software (Hammer et al. 2001).

We sampled wetlands across the diverse landscapes of Georgia: the Black Sea coastal zone with a maritime climate, the inland lowlands with a dry and continental climate, high dry plateaus, and subalpine environments. Our results permitted us to construct an annotated checklist of the wetland flora of Georgia, which includes 270 species sampled from 218 pre-selected sites. Each species on the resulting checklist (Table 1) is annotated with the following data: life form, wetland indicator status, chorotype, and location (study regions). The coordinates of each plot can be found in the supplementary information, where the species presence-absence matrix is also included as Suppl. material 4.

The checklist includes 270 species belonging to 80 families and 183 genera. Angiosperms are the core of the floristic composition – 251 species (93%). Bryophytes are represented by 10 species (3.7%), Pteridophytes – 8 (3%), Gymnosperms – 1 (0.4%). Leading families by the number of species are Cyperaceae – 39 (14.4%), Poaceae – 29 (10.7%), Rosaceae – 19 (7.1%), Asteraceae – 17 (6.3%), Fabaceae – 11 (4.1%) and Juncaceae – 11 (4.1%). The floristic composition of Georgian wetlands is dominated by Palaearctic species 55 (20.4%), followed by Holarctic 46 (17.1%), Euro-Mediterranean 31 (11.5%), Cosmopolitan 31 (11.5%) and Euro-Siberian 27 (10.03%) species.

Out of 270 species, half are represented by only one or two sites. The most frequent species are Alnus glutinosa subsp. barbata – (69 sites), Molinia caerulea – (56 sites), Juncus effusus – (42 sites), Carex vesicaria – (40 sites), Phragmites australis – (37 sites), Persicaria thunbergii – (36 sites), Sphagnum palustre – (35 sites), Lythrum salicaria – (32 sites), Cladium mariscus – (29 sites), Frangula alnus – (29 sites). Among the species listed above, five are OBL in wetlands, three – FACW, and of the remaining two, one – FACU and the other – FAC. The presence of the latter, among the ten most frequent species, indicates the current state of the wetland, which is often expressed by fragmentation and the spread of species uncharacteristic for wetlands.

The percent share of endemic species in the checklist is 4.8 %; the rate of endemism is low compared to the flora of Georgia, where the endemism rate is approximately 31% (Solomon et al. 2013). There are 13 endemic species, four of which are endemic to Georgia and nine to the Caucasus. At the same time, we recorded 23 non-native species, which gives a rate (8.5%) higher than that of endemics. The low endemism rate in wetland vegetation is not surprising since the typical wetland species usually have a very wide distribution; wetland vegetation is often considered to be azonal (sensu Sieben 2019), with very little variable composition across large gradients within subcontinents such as Mexico (Ramos-Tapia et al. 2022) or South Africa (Mucina et al. 2006) or even between so distant places such as the western and eastern extremities of Eurasia (Kikvidze and Ohsawa 2001; see also Denk et al. 2001).

In the checklist, species that almost always occur in wetlands (OBL) and species that are commonly (FACW) found in wetlands together made up 52.1% of the total number of species (OBL – 92 and FACW – 48 species). Species that are equally likely to occur in wetlands or non-wetlands (FAC) made up 68 species (25.3%). The plants that usually occur in non-wetlands but are occasionally found in wetlands (FACU) consisted of 62 species (23.1%).

Out of the 218 plots, 65 were taken in seasonally dry areas (water not visible), 37 were taken in fully water-covered habitats (water well visible), 113 – partially water-covered habitats (water hardly visible) and for three plots no data. It should be noted that we canceled a certain number of plots since we found different types of agricultural fields in places where, according to the old Soviet maps, there should have been wetlands.

According to the degree of human impact on the habitat, 46 plots were assessed as low (insignificant or not observed at all), 82 plots were assessed as intermediate (low impact rate), 89 plots were assessed as high (high anthropogenic impact), and for one plot there was no data.

During the fieldwork, we recorded Drosera rotundifolia for the first time in the Samegrelo region, where, together with the Rhynchospora alba, they make a unique plant community in the middle of the urbanized area (village Namikolaevo). This location is not mentioned for this species in the ‘Flora of Georgia’ (Ketskhoveli et al. 1971-2011), nor in the other reviewed literature. Regretfully, this unique habitat is in a very poor state owing to grazing by domestic animals and fragmentation, and it is possible that without proper protective measures it will disappear soon (see Suppl. material 2, fig. 5).

The presence of FAC and FACU species among the ten most frequently occurring species, the relatively low level of endemism against a high percentage of non-native species, and the assessment of 89 out of 218 sites as being under heavy human influence are indicative of the current challenging status of wetlands. The remaining undamaged wetlands with their highly diverse vegetation, distinctive relics, unique flora, and picturesque landscapes are impressive and can attract many visitors, provided such recreational and touristic activities are duly organized.

CCA axis 1 explained 63% of the variation in species distribution along geographical gradients, this axis was statistically highly significant (p = 0.005). The second axis explained 37% of this variation, yet it was insignificant (p = 0.34). Axis 3 could be dismissed as its explanatory power was very low (less than 1%, p = 0.53). The triplot (Fig. 2) shows that axis 1 represents variation in altitude and longitude. Variation in latitude apparently does not play an important role as it is associated with the insignificant axis 2. Altitude in Georgia is tightly correlated with the mean annual temperature and can serve as its precise proxy, whils longitude can be a proxy of precipitation (Kikvidze 2020). Therefore, perhaps the main climatic driver of species composition in the wetlands of Georgia is temperature, while precipitation might play a secondary role since the main feature of the wetlands is waterlogging, which can only weakly depend on precipitation.

Figure 2. 

Wetland vegetation species composition along geographical gradients as analyzed with the CCA ordination; all data are well within the 95% confidence interval (encircled). “Eastern lowlands” include the wetlands in Shida Kartli Lowland and Alazani Valley; Iori Valley stands for the entire Iori River gorge.