=Paper=
{{Paper
|id=Vol-3006/51_short_paper
|storemode=property
|title=Study of the diversity and mapping of xerophytic vegetation of the southeastern coast of Crimea peninsula using remote sensing data
|pdfUrl=https://ceur-ws.org/Vol-3006/51_short_paper.pdf
|volume=Vol-3006
|authors=Nikolai B. Ermakov,Igor A. Pestunov,Vladislav V. Korzhenevskiy,Elena V. Ermakova,Sergey A. Rylov,Natalia A. Trusheva
}}
==Study of the diversity and mapping of xerophytic vegetation of the southeastern coast of Crimea peninsula using remote sensing data==
https://ceur-ws.org/Vol-3006/51_short_paper.pdf
Study of the diversity and mapping of xerophytic
vegetation of the southeastern coast of Crimea
peninsula using remote sensing data
Nikolai B. Ermakov1,2,3 , Igor A. Pestunov1,4 , Vladislav V. Korzhenevskiy1 ,
Elena V. Ermakova1 , Sergey A. Rylov1,4 and Natalia A. Trusheva5
1
Nikita Botanical Gardens — National Scientific Center of the RAS, Yalta, Crimea, Russia
2
Central Siberian Botanical Garden SB RAS, Novosibirsk, Russia
3
Khakassia State University, Abakan, Khakassia Republic, Russia
4
Federal Research Center for Information and Computational Technologies, Novosibirsk, Russia
5
Maykop State Technological University, Maykop, Adygeya Republic, Russia
Abstract
The study of diversity and ecological-phytocenotic mapping of the vegetation cover was carried out at
the key area located in the eastern part of the Echki-Dag mountain range in the territory of the Lisya Bay
Reserve (Eastern Crimea). A generalization of the classification and ecological patterns of vegetation was
carried out to create a legend for a large-scale ecological-phytocenotic cartographic model (at the scale
of 1:10000). The ecological-geomorphological series and combinations of xerophytic and mesoxerophytic
plant communities indicating erosion-denudation processes on different substrates of the underlying
parent rocks were the thematic basis of the cartographic model. The developed legend is based on the
units of vegetation of the association rank obtained using the Braun – Blanquet method combined into
ecological series in accordance with their position on the gradients of the leading ecological factors as well
as on the hierarchy of phytochories determined by the categories of erosion-denudation relief of coastal
slopes. The resulting vegetation map demonstrates the main regularities of the regional phyto-diversity
and serves as the basis for assessing the resource potential of vegetation, its landscape-stabilizing and
nature conservation value.
Keywords
Vegetation, geobotanical mapping, phytodiversity, ecology, plant resources, relief, Crimea.
1. Introduction
Large-scale mapping of vegetation is currently of great importance for the study of theoretical
problems of phytodiversity, structure and dynamics of vegetation, as well as for solving applied
problems of assessing natural resources and conservation of rare and endangered species. The
mountain systems of the southeastern Crimea are characterized by a high diversity of zonal
natural vegetation of the European-Siberian and Mediterranean types with a high level of floristic
richness. Here, in a small area, ecologically diverse unique xerophytic plant communities form
combinations due to the influence of regional geological-geomorphological, bioclimatic and
floro-genetic processes. At the same time, this territory has been influenced by human activity
SDM-2021: All-Russian conference, August 24–27, 2021, Novosibirsk, Russia
" pestunov@ict.sbras.ru (I. A. Pestunov)
© 2021 Copyright for this paper by its authors. Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0).
CEUR
Workshop
Proceedings
http://ceur-ws.org
ISSN 1613-0073 CEUR Workshop Proceedings (CEUR-WS.org)
437
Nikolai B. Ermakov et al. CEUR Workshop Proceedings 437–445
for a long time. It led to a decrease of natural vegetation and a wider spread of disturbed
communities there. Therefore, large-scale geobotanical mapping forms the basis for regional
assessing the degree of anthropogenic disturbance of vegetation and habitats, as well as the
basis for assessing the conservation value of plant communities, monitoring the state of rare
and endangered plant species. Until now, detailed large-scale mapping of vegetation cover as
an indicator of ecological and anthropogenic changes in the natural environment of Crimea
was carried out only for some geographically local areas. Maps of actual vegetation on the
territory of the Republic of Crimea, made according to modern technological requirements
using remote sensing data and modern computing technologies, are completely absent. There
are only overview maps of potential vegetation, created on a small and medium scale [1, 2],
which can be very limitedly applied to solve modern large-scale economic and environmental
problems.
At present, fundamentally new possibilities for correct and detailed representation of veg-
etation categories on geobotanical maps at different scales have appeared as a result of the
development of GIS technologies and methods for processing remote sensing data. It has been
implemented in various works [3, 4, 5]. N.B. Ermakov [6], M.A. Polyakova and N.B. Ermakov [7]
studied the spatial patterns of steppe vegetation in the Minusinskaya intermountain basin using
GIS technologies.
The aim of this study is to reveal the regularities of the spatial organization of natural
vegetation in the mountain-steppe belt and to conduct a large-scale geobotanical mapping in a
key area located on the territory of the Lisya Bukhta regional landscape reserve (the eastern
part of the Echki-Dag mountain range) using remote sensing data.
2. Materials and methods
The study was carried out on the basis of 80 geobotanical releves of steppe vegetation carried
out in accordance with the standard technique [8] on the plots of 100 m2. A whole series
of releves was put into the database based on the European package TURBO (VEG) [9]. The
methodology of the study was a system analysis that considers the vegetation cover as a complex
hierarchically organized open mega-system with the leading role of ecological and geographical
factors at different levels of its organization [10].
Vegetation classification was carried out using the Braun – Blanquet method [11]. The
releves were quantitatively classified using cluster analysis implemented in the JUICE 7.0
program [12]. The nomenclature of syntaxonomic units was made in accordance with the Code
of Phytosociological Nomenclature [13].
The assessment of the ecological peculiarities of the syntaxa habitats was carried out accord-
ing to the results of the ecological analysis of floristic compositions using ecological species
groups. These groups were distinguished by traditional methods [14, 15] by taking into account
the attitude of plants to the following environmental factors: to the water regime of the habitat,
to soil fertility, to the composition of bedrocks. The study of the vegetation spatial structure
was performed in accordance with Sochava’s approach [10] including the application of a clas-
sification system for phytocoenomeras and phytocoenochoras. The system of spatial categories
of vegetation proposed by S.A. Gribova, T.I. Isachenko [16] was used as well. The regularities of
438
Nikolai B. Ermakov et al. CEUR Workshop Proceedings 437–445
the spatial organization of plant communities in connection with the dynamics of the relief on a
cartographic model were studied on the basis of field interpretation of high-resolution satellite
images (WorldView-2 with a resolution of 1.8 m) using a multispectral image segmentation algo-
rithm, based on computationally efficient hierarchical clustering algorithm HECA [17, 18], and
spectral-texture segmentation algorithm ESEG [19]. These algorithms were developed within
the framework of the grid-based and ensemble approaches and allow to extract hierarchically
nested clusters of complex shapes, different sizes and densities, even in the case of intersecting
clusters (in the space of spectral features). The processing of satellite information was carried
out in the field so that the vegetation structures and individual plant communities revealed
in the image could be found directly in the landscape. The boundaries of plant communities
and phytochoras were identified as clearly as possible because this is very important for the
organization of further long-term monitoring.
3. Results and discussion
A vegetation classification system was developed using the Braun – Blanquet approach. It
includes 5 vegetation classes, 5 orders, 6 alliances and 15 associations. Such a high diversity of
associations and higher categories in a relatively small area is explained by the diverse ecological
conditions and the location of the study area to the plant-geographical border between the
European and Mediterranean regions.
The spatial organization of the diversity of geographic and ecological categories of xerophytic
vegetation is clearly determined by erosion-denudation slope processes in the eastern part of
the Echki-Dag ridge on different types of bedrocks. Based on the analysis of the relations of
plant communities with certain relief elements and the composition of ecological groups of
species, it was revealed that the leading factor determining the formation of the highest spatial
vegetation units is the geological-geomorphological process on the southern macro-slope of
the mountain ridge. In accordance with this feature, the entire diversity of spatial units of
vegetation is combined into two large altitude categories:
1) vegetation of the geomorphological belt of colluvial trails (200–400 m);
2) vegetation of the geomorphological belt of gully-ravine relief with a wide development of
fluvial terraces (0–150 m).
Within each geomorphological belt, an assessment of the ecological regimes of certain habitats
was carried out as a result of an analysis of the ratio of ecological groups of species indicating
the leading factors of temperature, humidity, soil fertility and composition of bedrocks. The
results of the analysis demonstrated the presence of a hierarchy of environmental factors, among
which the petrographic composition of the bedrocks was the leading one, which contributes to
the formation of large geomorphological belts in the eastern part of the Echki-Dag ridge. The
factors of temperature and humidity determine the local regularities of certain vegetation types
distributions on specific elements of relief within each geomorphological belt. Interpretation of
WorldView-2 satellite images and recognition of vegetation categories directly in the field made
it possible to reveal the basic ecological and topographic combinations of communities (series)
and individual associations within each of the highest rank phytochories at a larger scale of
mapping.
439
Nikolai B. Ermakov et al. CEUR Workshop Proceedings 437–445
The vegetation of the upper geomorphological belt of colluvial plumes at altitudes of
200–400 m is represented by four types of phytochories.
On weakly stony habitats of colluvial trails and in the upper parts of slopes, an ecological
series of communities associated with a moisture factor has been identified. It is represented by
associations of moderately drought-resistant communities Artemisio alpini-Festucetum callierii
and the most xerophytic communities Artemisio tauricii-Koelerietum cristatii.
The vegetation of bedrock outcrops on the walls of the breakdown and drainage depressions
of the erosion-landslide relief according to the results of interpretation of the satellite image is
presented as a single phytochoria — a combination of a petrophytic series of associations that
occupy habitats with poorly developed and fragmented soils — from communities of bedrock
outcrops Euphorbio petrophyreti-Cotone communities of slopes with well-developed soil —
Artemisio tauricii-Koelerietum cristatii. The vegetation of bedrock outcrops on the walls of the
breakdown and drainage depressions of the erosion-landslide relief according to the results of
interpretation of the satellite image is presented as a single phytochoria — a combination of a
petrophytic series of associations occupying habitats with poorly developed and fragmented
soils — from communities of bedrock outcrops of Euphorbio petrophylli-Cotoneasteretum to
weakly on trails of slopes with well-developed soil — Artemisio tauricii-Koelerietum cristatii.
Widespread erosional relief elements — landslides are ecologically homogeneous and occupied
by an association of moderately petrophytic meso-xerophilous shrub communities of the Cotino-
Paliuretum. The most eroded habitats of the detachment walls of ancient landslides are occupied
by petrophytic xerophytic communities of the association Onosmo rigidi-Asphodeletum tauricii.
A similar leading role of humidity and petrophytic factors is observed for the distribution
of communities in the geomorphological belt of gully-gully relief with a wide development
of fluvial terraces (0–150 m). Nevertheless, occupying lower hypsometric levels, this geomor-
phological belt is characterized by a generally flatter relief with a predominance of clay or
loamy diluvium. Here, the most eroded slopes of ravines crossing the deposits of Maikop clays,
fleeshes, sandstones and conglomerates create conditions for the mosaic distribution of small-
sized communities of the xero-petrophytic ecological series: Ceratoido papposae-Seselietum
dichotomii, Astrodauco orientali-Ptilostemonietum, Astragalo-Tragacantietum, which are closely
related to the type of parent bedrocks and the corresponding moisture regimes.
The shaded parts of the ravine slopes are occupied by weakly petrophylous xero-mesophilous
communities of juniper-oak forests and woodlands of the association Junipero-Quercetum
pubescentis.
A well-defined combination of low, moderately gentle proluvial terraces is characterized
by relatively stable, slow erosion processes. A predominance of one xerophytic association
Botriochloe-Paliurietum on well-developed soils is observed there.
Older wide high proluvial terraces are characterized by weakly pronounced erosion processes
and are covered with a layer of loam, on which well-developed chernozem-type soils are formed.
They are dominated by weakly petrophylous meso-xerophylous steppes the association Poterio-
Jurinetum stoechadifolii. The higher ancient slopes with well-developed loamy soils resting
on high proluvial terraces are occupied by xerophytic steppes of the association Medicago
romanici-Bromopsietum cappadocii.
Ultra-xerophytic habitats widespread in the study area on steep eroded southern coastal
slopes from Maikop clays and fleish (badlands) are occupied by the Mediterranean communities
440
Nikolai B. Ermakov et al. CEUR Workshop Proceedings 437–445
of Atrophaco-Capparidetum.
The regularities of the spatial organization of the diversity of plant communities in the
key area were studied on the basis of field interpretation of high-resolution satellite images
(WorldView-2 with a resolution of 1.8 m) using a multispectral image segmentation algorithm,
based on computationally efficient hierarchical clustering algorithm HECA, and spectral-texture
segmentation algorithm ESEG.
They are shown in the legend of the map for the key area in the eastern part of the Echki-
Dag mountain range and are based on taking into account the hierarchy of regional and local
environmental factors.
The substantial part of the large-scale map in the eastern part of the Echki-Dag ridge at a
visible scale is presented in two parts (in Figures 1 and 2) with a single legend. Two parts of
the cartographic model represent ecological and geomorphological series of vegetation and
combinations of xerophytic and mesoxerophytic plant communities within two phytochorias of
the highest rank. The legend presents the regularities of vegetation formation associated with
erosion-denudation processes on different substrates of the bedrocks.
Legend of the large-scale vegetation map of the eastern part of the Echki-Dag ridge.
Vegetation of the geomorphological belt of colluvial plumes (200–400 m).
1. Vegetation of colluvial trails and near-watershed slopes (combinations of associations of
weakly petrophytic xerophytic steppes of associations Artemisio alpini-Festucetum callierii
and Artemisio tauricii-Koelerietum cristatii).
2. The vegetation of the stall walls and drainage depressions of the erosion-landslide relief is a
combination of the petrophytic series of associations Euphorbio petrophylli-Cotoneasteretum,
Alysso obtusifolio-Seselietum dichotomii, Hedysaro taurici-Agropyretum ponticii.
3. Landslide vegetation — moderately petrophytic meso-xerophilous communities of Cotino-
Paliuretum shrubs.
Figure 1: Fragment of a large-scale vegetation map of the eastern part of the Echki-Dag ridge (within
the boundaries of the geomorphological belt of colluvial trails (200–400 m) created on the basis of
high-resolution satellite image WorldView-2. The map legend is in the text.
441
Nikolai B. Ermakov et al. CEUR Workshop Proceedings 437–445
Figure 2: Fragment of a large-scale vegetation map of the eastern part of the Echki-Dag ridge (geomor-
phological belt of gully-gully relief with a wide development of fluvial terraces 0–100 (150) m), created
on the basis of a high-resolution satellite image WorldView-2. The map legend is in the text.
4. The vegetation of the walls of the breakdown of ancient landslides — petrophytic xerophytic
communities of the association Onosmo rigidi-Asphodeletum tauricii.
Vegetation of the geomorphological belt of gully-ravine relief with wide development
of fluvial terraces (0–150 m).
5. Vegetation of eroded slopes of ravines from Maikop clays, fleeshes, sandstones and conglom-
erates (a combination of communities of the ecological xero-petrophytic series of associations
Ceratoido papposae-Seselietum dichotomii, Astrodauco orientali-Ptilostemonietum, Astragalo-
Tragacantietum).
6. The vegetation of the first and second proluvial terraces: weakly petrophytic xerophilous
communities of Botriochloe-Paliurietum on well-developed soils.
7. The vegetation of the third, fourth, and fifth proluvial terraces: weakly petrophytic meso-
xerophytic communities of the association Poterio-Jurinetum stoechadifolii.
8. Non-petrophytic vegetation of ancient slopes resting on high terraces: xerophytic steppes of
the association Medicago romanici-Bromopsietum cappadocii.
9. The vegetation of the badlands of the southern slope is ultra-xerophytic petrophytic commu-
nities of the association Atrophaco-Capparidetum.
10. The vegetation of the shaded slopes of ravines and ravines: weakly petrophytic xero-
mesophilous communities of forests and woodlands of the association Junipero-Quercetum
pubescentis.
11. Sea area.
442
Nikolai B. Ermakov et al. CEUR Workshop Proceedings 437–445
The results of large-scale mapping of vegetation demonstrate the main regularities of the
regional phytodiversity and also serve as the basis for assessing the resource potential of vegeta-
tion and the conservation value of communities. The associations shown on the map and their
combinations are interpreted from the standpoint of resource and environmental significance
according to data from thematic information systems “Rare and endangered plants of Crimea”
and “Plant resources of Crimea”. This made it possible to create a series of cartographic layers,
Figure 3: The cartographic model of the concentration of rare and endangered plant species demon-
strates their highest number in the petrophytic series of associations — Euphorbio petrophylli-
Cotoneasteretum, Alysso obtusifolio-Seselietum dichotomii and Hedysaro taurici-Agropyretum ponticii (1)
as well as in the petrophytic xerophylous association Onosmo rigidi-Asphodeletum tauricii (2). Other
vegetation communities — 3.
Figure 4: Cartographic model of the highest concentration of valuable medicinal plants according
to the ratio of the “number of species – abundance – occurrence” values in the association Medicago
romanici-Bromopsietum cappadocii (1). Other designations: 2 — other plant communities; 3 — sea area.
443
Nikolai B. Ermakov et al. CEUR Workshop Proceedings 437–445
in which not only the traditional quantitative indicators of important plant species, but also
their ecological characteristics are reflected as accurately as possible to show the resource and
nature conservation role of specific plant communities. This made it possible to create a series
of cartographic layers that most accurately demonstrate the traditional quantitative indicators
of important plant species, their ecological characteristics, the resource and nature conservation
role of certain plant communities. The peculiarities of the distribution of groups of plant species
(which are important in environmental and resource aspects) in the vegetation cover and in spe-
cific ecological and geomorphological conditions are shown on cartographic models (Figures 1
and 2). The cartographic model (Figure 3) shows the regularities of the concentration of rare
and endangered plant species of Crimea in specific plant communities and their biotopes. These
thematic models show that the largest number of these species is concentrated in the petro-
phytic series of plant communities of the associations Euphorbio petrophylli-Cotoneasteretum,
Alysso obtusifolio-Seselietum dichotomii, Hedysaro taurici-Agropyretum ponticii, as well as in the
petrophytic association Onosmo rigidi-Asphodeletum tauricii. At the same time, communities
of the non-petrophytic association of xerophytic steppes Medicago romanici-Bromopsietum
cappadocii (Figure 4) are of the greatest resource significance in terms of quantitative indicators
(number of species, abundance, occurrence) of valuable medicinal plants.
4. Conclusion
As a result of the classification and analysis of the ecological composition of plant communities,
ecological series and combinations determined by the factors of temperature, humidity and
petrophyticity have been identified. These series demonstrate the regularities in the formation
of xero-perophytic vegetation in the southeastern Crimea. For the first time for the Black Sea
region, large-scale mapping of vegetation was carried out on the basis of high-resolution satellite
images and the use of modern methods of processing remote sensing data. A series of maps
of biotopes of rare and endangered plant species, resource significance of plant communities
was created on the basis of detailed large scale cartographic models of vegetation. The obtained
results are important for the conservation and rational use of vegetation, as well as for the
development of a scientific and methodological approach to monitoring and predicting changes
in the ecosystems and environment of the Crimea Republic. The implemented methodological
approach allows large-scale mapping of vegetation using remote sensing data and modern
information technologies.
Acknowledgments
The study was supported by budget theme of Nikita Botanical Gardens — National Scientific
Center of the RAS (project No. 1009-2015-0018).
References
[1] Rubtsov N.I. Plant world // In: Rubtsov N.I., Makhaeva L.V., Shalyt M.S., Kotova I.N. The
nature of the Crimea. Simferopol: Publishing house “Crimea”, 1964. 123 p.
444
Nikolai B. Ermakov et al. CEUR Workshop Proceedings 437–445
[2] Didukh Ya.P. Vegetation cover of the Crimean mountains (structure, dynamics, evolution
and protection). Kiev: Naukova Dumka, 1992. 256 p.
[3] Danilova I.V., Ryzhkova V.A., Korets M.A. Algorithm for automated mapping of the current
state and dynamics of forests based on GIS // Vestnik NSU. Series: Information Technology.
2010. Vol. 8. Is. 4. P. 15–24.
[4] Neshataev M.V., Neshataev V.Yu. Vegetation mapping for the cadastral assessment of
Nature protected areas (on the example of the Lapland Nature Reserve) // Bulletin of the
Samara Scientific Center of the RAS. 2012. Vol. 14. No. 1(6). P. 1630–1633.
[5] Kashirina E.S., Bondarev L.V. Mapping the vegetation of the landscape reserve of national
importance “Cape Fiolent” and the zoological reserve of national importance “Bukhta
Kazachya” (Big Sevastopol, Crimea) // Ecosystems, Their Optimization and Protection.
2013. Is. 8. P. 130–140.
[6] Ermakov N., Larionov A., Polyakova M., Pestunov I., Didukh Ya. Diversity and spatial
structure of cryophytic steppe of the Minusinskaya basin in Southern Siberia (Russia) //
Tuexenia. 2014. Vol. 34. P. 431–446.
[7] Polyakova M.A., Ermakov N.B. The study of steppe vegetation spatial structure in Khakassia
using satellite images of different resolution // Ekosistemy. 2019. Is. 18. P. 3–13.
[8] Field geobotany. Methodical guide / E.M. Lavrenko (Ed.). Novosibirsk: Publishing House
of the Academy of Sciences of the USSR, 1964. Vol. 3. 530 p.
[9] Hennekens S.M., Shaminée J.H.J. TURBOVEG, a comprehensive data base management
system for vegetation data // Journal of Vegetation Science. 2001. Vol. 12. P. 589–591.
[10] Sochava V.B. Vegetation cover on thematic maps. Novosibirsk: Nauka, Siberian Branch,
1980. 190 p.
[11] Westhoff V., van der Maarel E. The Braun – Blanquet approach. // Handbook of Vegetation
Science. 1973. Vol. 5. P. 617–726.
[12] Tichy L. JUICE, software for vegetation classification // Journal of Vegetation Science. 2002.
Vol. 13. P. 451–453.
[13] Theurillat J.P., Willner W., Fernández-González F, Bültmann H, Čarni A, Gigante D,
Mucina L, Weber H. International code of phytosociological nomenclature. 4th edition //
Applied Vegetation Science. 2020. Vol. 24. No. 2. 62 p. DOI:10.1111/avsc.12491.
[14] Bel’gard A.L. Steppe forestry. Moscow: Lesnaya Promyshlennost, 1971. 336 p.
[15] Tsyganov D.N. Ecomorphes of the flora of coniferous-deciduous forests. Moscow: Nauka,
1976. 60 p.
[16] Gribova S.A., Isachenko T.I. Field geobotany. Methodical guide. Novosibirsk: Publishing
House of the Academy of Sciences of the USSR, 1972. Vol. 4. P. 137–330.
[17] Pestunov I.A., Rylov S.A., Berikov V.B. Hierarchical clustering algorithms for segmentation
of multispectral images // Optoelectronics, Instrumentation and Data Processing. 2015.
Vol. 51. P. 329–338.
[18] Rylov S.A., Pestunov I.A. Fast hierarchical clustering of multispectral images and its
implementation on NVIDIA GPU // Journal of Physics: Conference Series. 2018. Vol. 1096.
P. 012039.
[19] Pestunov I.A., Rylov S.A. Spectral-textural segmentation algorithms for satellite images
with high spatial resolution // Bulletin of Kemerovo State University. 2012. No. 4/2(52).
P. 104–110. (In Russ.)
445