Diatom diversity, distribution and ecology in Mediterranean ecosystems of Abrau Peninsula, north-western Caucasus

Abstract Background The North Caucasus is an extensive region with a multitude of landscapes and high biological diversity. Amongst various ecosystems, the xerophytic sub-Mediterranean forests of the Abrau Peninsula (Utrish State Nature Reserve) and its vicinity are unique but have been poorly studied. The diversity of diatoms in North Caucasian ecosystems have been studied partially and only little information is available about their presence and distribution on the Abrau Peninsula. Here, we present a comprehensive check-list of diatoms sampled during a July 2021 field campaign. Samples were collected in 67 sites, including 39 permanent streams, 21 temporal (puddles) and seven permanent waterbodies. Results of the current study contribute to improving the knowledge about diatoms in the north-western Caucasus and its sub-Mediterranean ecosystems in particular. New information Here, we provide a detailed dataset that contains 215 freshwater and brackish diatom occurrences collected during a field campaign in July 2021. A total of 88 diatom (Bacillariophyta) taxa which belong to 12 orders, 25 families and 39 genera were collected. The genera with the highest number of occurrences per site were Gomphonema (26), Nitzschia (22), Navicula (20), Cocconeis (14), Amphora (14), Achnanthidium (14) and Planothidium (11). The genera with the highest number of infrageneric taxa were Nitzschia (8), Navicula (7), Gomphonema (6) and Mastogloia (5). Naviculablazencicae, known as the endemic of the Lake Prespa (Levkov 2007) is found from two sites in our study. Three specimens of the genus Mastogloia could not be assigned to a known species and may represent new diatom species. Distribution and ecology data are provided for each taxa. Occurrence data are given. Statistical analysis of diatom communities showed a significant dependence on habitat type and their ecological conditions.


Introduction
Diatoms are a widely distributed group of algae whose representatives populate both aquatic (marine and freshwater) and terrestrial ecosystems, such as soils, mosses, wet walls and rocks (Round et al. 1990, Smol andStoermer 2010) and play a key role in the nutrient cycle and energy flux (Benoiston et al. 2017). In seas and oceans, organic carbon produced by diatoms is consumed rapidly and serves as a base for marine food webs. In coastal waters, diatoms support most productive fisheries. In the open ocean, a relatively large proportion of diatom organic matter sinks rapidly from the surface, becoming food for deep-water organisms (Armbrust 2009). Soils and other terrestrial ecosystems have more severe effects for diatoms and differ from aquatic ecosystems in diatom species composition, although diatoms can be the dominant algal group at periods of the year with high soil moisture (Foets et al. 2020).
Diatoms are regularly used as biological indicators for the water quality environmental assessment (Patrick 1973, Reid et al. 1995, Kelly et al. 1998, Battarbee et al. 2002. The analysis of diatom communities and their biodiversity is a useful tool to secure an ecological and sustainable use of the water resources and the correct elaboration of guidelines for their preservation, in particular, in specially protected natural areas. Some recent studies have shown that natural springs in protected areas may act as biodiversity hotspots Bona 2011, Falasco et al. 2012).
Different ecological groups of Black Sea diatoms have been actively studied, especially from the perspective of water quality assessment (Petrov and Nevrova 2007, Ryabushko et al. 2017, Ryabushko et al. 2019. Additionally, there is ongoing research on diatom diversity of specially protected natural sea areas of the Black Sea (Nevrova 2015, Ryabushko et al. 2018, Polyakova and Davidovich 2019, Davidovich and Polyakova 2022. In a recent study of diatom communities in the water area surrounding Bolshoi Utrish (Anapa District, Russia), it was found that 77% of the biomass and 25% of the total phytoplankton abundance was composed of Bacillariophyta species (Yasakova and Kolesnikov 2021). Inland research of diatom communities have mainly been focused on Abrau Lake, the largest freshwater lake in the Abrau Peninsula (Kovaleva 2005, Kovaleva 2018). There are, however, a number of freshwater waterbodies and streams in the Abrau Peninsula and the nearby Black Sea coastal zone that are still understudied in terms of diatom diversity and distribution. We assume that some sampling locations (freshwater streams) on the territory of Utrish State Nature Reserve, especially without anthropogenic disturbance, potentially might be hidden hotspots of diatom biodiversity.
This study presents a taxonomical characterisation and occurrence dataset of the diatoms found in Mediterranean ecosystems of the Abrau Peninsula, north-western Caucasus, particularly in protected areas of Utrish State Nature Reserve. We aim to contribute to the current knowledge of diatom diversity and distribution in the freshwater and brackish inland water in the north-western Caucasus and its sub-Mediterranean ecosystems in particular.

Project description
Title: Diatom diversity, distribution and ecology in Mediterranean ecosystems of Abrau Peninsula, north-western Caucasus.
Personnel: Samples were collected on 12-20 July 2021 on the Abrau Peninsula by Alisa Neplyukhina and Angelina Paskhina. Identifications were made by Alisa Neplyukhina. Statistical analyses were performed by Daniil Korobushkin and Ruslan Saifutdinov. The text was written by Alisa Neplyukhina, Daniil Korobushkin and Ruslan Saifutdinov.

Study area description:
The Abrau Peninsula is located between the city of Anapa and Abrau-Durso settlement in Novorossiysk District, Krasnodar Krai, Russia. Most of the Abrau Peninsula is under the protection of the Utrish State Nature Reserve (hereinafter referred to as "Utrish") and is not affected or disturbed by human activity, with the exception of the coastal zone and suburbs. The Abrau Peninsula has a humid subtropical (Cfa) and Mediterranean climate (Csa) according to the Koppen climate classification with cool rainy winters without stable snow cover and with hot dry summers (Chen and Chen 2013). The mean annual precipitation ranges from 480 mm (Anapa) to 788 mm (Novorossiysk), the mean July and February temperatures for both localities are 21℃ and 2℃, respectively (weatherbase, CantyMedia 2022).
The study area belongs to the Mediterranean ecoregion (Olson et al. 2001, Ogureeva et al. 2018 and is the only place in Russia covered by Mediterranean forests. The vegetation here forms three major belts (Bocharnikov et al. 2020, Seregin andSuslova 2007): 1. coastal slopes with sub-Mediterranean xerophytic forests and shrublands with pistachio (Pistacia mutica), juniper (Juniperus excelsa, J. oxycedrus, J. foetidissima), oak (Quercus pubescens) and oriental hornbeam (Carpinus orientalis); 2. piedmont and low-mountain area with a combination of mesophilic and xerophilic forests and a predominance of two oak species (Q. pubescens, Q. petrea), oriental hornbeam and junipers; 3. low mountains with mesophilic deciduous forests with a domination of oak (Q. petrea), hornbeam (Carpinus caucasica), lime (Tilia begoniifolia), maple (Acer laetum), ash (Fraxinus excelsior ) and beech (Fagus orientalis). A distinctive feature of the Utrish flora is unique pronounced Mediterranean core tertiary relict elements. It is inhabited by numerous rare, endemic species of flora and fauna.
Freshwater habitats are represented by permanent and temporary streams flowing to the Black Sea, as well as temporary small waterbodies (hereinafter referred as "puddles") scattered across the Abrau Peninsula. Brackish habitats are represented by small permanent lagoons located along the coastline.

Sampling methods
Study extent: Diatoms were collected from 67 sampling sites, including 39 permanent streams, 21 temporal waterbodies (puddles) and seven permanent waterbodies (lakes and lagoons) collected on the Abrau Peninsula, north-western Caucasus, Russia (Fig. 1 Sampling description: Material for this research was collected in July 2021. Sampling was carried out after the annual peak of summer precipitation in June (CantyMedia 2022) and performed after a week of strong rains (CantyMedia 2022). This made it possible to collect material from both permanent and temporary waterbodies. Diatom samples were collected from 67 sites on the Abrau Peninsula (Table 1). The sampling sites differed in salinity from brackish to freshwater. Sample types include 36 rock scrap samples, 24 sediments samples, six soil samples and one moss squeeze sample. Diatom samples were collected in 50 ml plastic containers and immediately fixed with Lugol's solution (2 ml to 50 ml of sample) in order to keep other algae groups in their best condition for futher research (Sadchikov 2003). Material was cleaned from the organics in accordance with the hot peroxide method following Kelly et al. (2001). Light microscopical investigations were performed in bright-field optics using a Leica DM 750 microscope, equipped with a Leica ICC50 HD digital camera. Permanent slides were prepared with Naphrax®. For the scanning electron microscopy investigation, drops of cleaned material were air-dried on pieces of aluminium foil, mounted on brass stubs with double-sided carbon tape and coated with Au in a S150A Sputter Coater (Edwards, UK) ion coater. Scanning electron microscopic investigations were conducted using TESCAN MIRA 3 LMH (TESCAN, Czech Republic) in the Joint Usage Center «Instrumental methods in ecology» at the IEE RAS. All prepared LM slides and SEM stubs are stored in the collection of the Laboratory for Ecology of Aquatic Communities and Invasions, IEE RAS.
Step description: The data have been published as a Darwin Core Archive (DwC-A), which is a standardised format for sharing biodiversity data as a set of one or more data tables. The core data table contains 215 occurrences (Neplyukhina et al. 2022). Statistical analysis: Similarity between diatom communities of Abrau Peninsula was evaluated using hierarchical cluster analysis. Before analysis, the data were prepared via the dplyr 1.0.8. package (Wickham et al. 2018) into species x communities matrix with presence/absence data. Data on diatoms were pooled into communities according to their presence in the habitat type (stream, waterbody or puddle) and according to the sampling method (scrap, sediment, moss and soil). A detailed description of habitat type selection and sampling methods is given in Table 1. Distances between communities were calculated using a binary method and the Ward.D2 method was selected for the hierarchical clustering procedure. Additionally for each cluster, bootstrap probability value (BP) and approximately unbiased (AU) probability values (p-values) were calculated via multi-scale bootstrap on 10000 resamplings using the package pvclust 2.2-0 (Suzuki and Shimodaira 2006). To define our clusters, we used a significance level of p < 0.05, i.e. the AU value equal or higher than 95. The obtained dendrogram was customised with the dendextend 1.15.2 package (Galili 2015

Taxonomic coverage
Description: All diatoms were identified to genus or species/intraspecific level. In total, 88 infrageneric taxa were identified belonging to two classes, 12 orders, 25 families and 39 genera distributed in the subphylum Bacillariophytina, 11 of them being identified only to genus level. The taxonomic coverage of the diatoms found in studied material is given in Table 2. The diatom species list with their ecological preferences, distribution and occurrence is given in Table 3.  Table 2.
Taxonomic coverage of diatoms from studied samples. Table 3.
List of diatom species found in samples with notes on their ecology, distribution and occurrence (number of samples). Data on ecology and distribution are given according to Kulikovskiy et al. 2016, Cantonati et al. (2017, and Guiry and Guiry 2022 .  Description: This dataset presents the first data on the distribution of freshwater and brackish diatoms on Abrau Peninsula and especially in the territory of the Utrish State Nature Reserve. The data in this occurrence resource have been published as a Darwin Core Archive (DwC-A), which is a standardised format for sharing biodiversity data as a set of one or more data tables. The core data table contains 215 occurrences. This IPT archives the data and, thus, serves as the data repository.

Column label Column description
id The ID of the record.

type
The nature of the resource. basisOfRecord The specific nature of the data record. identificationQualifier Contains commentaries about taxon identification (marks sp., sensu lato etc.)
The UT-2021-67 site is a quite unique sampling site, where freshwater from the Zhemchuzhnyj Waterfall stream mixes with seawater and rocks with water from the stream being covered with moss. From this site, we sampled both rock scrap and moss squeeze and found the highest diversity of diatom taxa (Fig. 3).
Light microscope (LM) and scanning electron microscope (SEM) images of the most frequently occurring species and some others are represented in Fig. 2.
Navicula blazencicae Levkov (Fig. 2, 1-3) was originally described by Levkov and colleagues (Levkov et al. 2007) from North Macedonia and, until now, has been known as the endemic of Lake Ohrid. In the study, it was found in two sampling locations represented by two freshwater temporal waterbodies (UT-2021-05 and UT-2021-09).  One of Mastogloia species, referred to as Mastogloia sp.2 (Fig. 2, 8-10), held a unique combination of morphometric characteristics (paratecta and raphe structural features) which we were unable to identify as a known species. Probably the same species was also found by A. Kaleli (Kaleli 2019) in similar habitats of coastline brackish waterbodies. According to the published illustration (see Mastogloia sp.1 in Kaleli 2019), the valves collected by A. Kaleli are quite similar to Mastogloia sp.2 in the current study and supposedly belonged to the same species, although it has also not been identified and needs additional verification. Beside that, two other Mastogloia species (Mastogloia sp.1 and Mastogloia sp.3) which were found in the current research are also likely to be new species and require further study.

Data analysis
The cluster analysis revealed a considerable modulation effect of habitat type on the floristic composition of diatom communities of the Abrau Peninsula (Fig. 4). Diatom communities collected from streams, regardless of the sampling method, were significantly dissimilar to the communities collected from waterbodies and communities collected from scraps and soils of puddles (p < 0.05). In turn, communities collected from waterbodies Hierarchical cluster analysis using the presence/absence matrix of diatom communities collected by different sampling methods from various biotopes of Abrau Peninsula (binary method, Ward.D2 clustering). Different symbols at the nodes of clusters illustrates biotope types: triangles -streams, squares -puddles (temporary waterbodies) and circlespermanent waterbodies. Right part of labels illustrates the type of sampling method: Mossmoss squeeze, Sedim -sediment from the bottom of waterbody or stream, Scrap -scrapping from the stones and Soil -soil in the littoral zone of waterbodies. Values at branches are approximately unbiased p-values (red colour) and bootstrap probabilities (green colour) in percentage. Clusters that are framed by red dashed line are supported by a p-value < 0.05.
were combined with communities sampled from scraps and soils of puddles and formed significant clusters (p < 0.01).
The results of cluster analysis suggest that the floristic composition of diatom communities from streams is quite different from that in small ephemeral water objects (puddles) and stagnant water bodies (such as ponds, lakes and lagoons). Although some of the species living in streams might sometimes be present in puddles (see Fig. 4), the floristic composition of streams is most likely conservative and does not mix with other types of water objects.
The ecological conditions of marine and brackish waterbodies were obviously antagonistic to freshwater, thus the PCA by factor 1 clearly and predictably separated the frequency of freshwater and marine and brackish species (Fig. 5). The frequency of occurrence of oligotrophic and mesotrophic species strongly and positively correlated with freshwater streams of Abrau Peninsula. Only here were collected freshwater species such as Amphora inariensis, A. pediculus, Cymbella affinis, C. hantzschiana, Encyonopsis microcephala, Gomphonema pygmaeum, Navicula tripunctata and Reimeria uniseriata.
The majority of collected eutraphentic species tended to be from freshwater puddles (e.g. Craticula dissociata, C. molestiformis, Gomphonema parvulum) and, to a lesser extent, freshwater waterbodies of the study area. The latter were positively correlated with the occurrence of species that prefer polluted water and, conversely, were antagonistic to the habitats of oligotrophic species and stream habitats. This may be due to the location of this type of waterbodies mainly near recreational areas and settlements. Aerophilic species did not show any strong correlation with the studied habitat types. Relationship between frequency of species occurrences with different separability (italic, active variables) and environment preferences (normal, supplementary variables) and in various investigated habitats (bold and italic, active variables) determined using the principal component analysis (PCA).