17:15 〜 19:15
[MIS05-P12] Methods to access Eastern Mediterranean ecosystem changes at Caucasus Black Sea coast
キーワード：Mediterranean ecosystems, Black Sea coast, Remote sensing data, Ecosystem changes
The Caucasus region is characterized by high biodiversity, number of endemic species and protected areas. However, its ecosystems are under significant anthropogenic impact, especially in the coastal area due to the large number of tourists and local residents, expansion of resort infrastructure and agricultural lands. In addition to anthropogenic impact, there is a high risk of dangerous climate events: extreme values of air temperature, wind speed, precipitation, droughts, floods, mudflows etc. According to the modern data aridization and desertification trend is also observed in the region: surface air temperature growth rate is 0.9-1.2°C per 10 years, but average annual precipitation is nearpermanent.
Studied ecosystems are located at Caucasus Black Sea coast on the low mountain range (434 m) west of Gelendzhik (Krasnodar region). Based on field research data (about 300 points of landscape and geobotanical descriptions), literature sources and remote sensing data, we identified main ecosystems and assessed their changes.
First, we constructed digital elevation model (DEM) using an area of the topographic map (1:50 000 scale). Based on the DEM, maps of the exposure of the main relief elements and their steepness were constructed. The combined data on steepness and exposure were superimposed on the altitude map (50 m step). As a result, a map of the main ecotopes was obtained. It includes 72 types in the research area. Each polygon on the map was assigned its own number (Id) and corresponding ecotope type. Thus, initial database was formed.
Using high-resolution (1 m) Esri satellite image, Sentinel-2 satellite images (10 m resolution) in visible and shortwave infrared (SWIR) range we divided forest and non-forest ecosystems and coniferous, mixed and broadleaf forests, as well as areas that have been terraced for planting Pinus pityusa and Pinus pallasiana.
We also used NDVI calculated from Sentinel-2 satellite images for the active phase of vegetation period to compare it with field data and clarify information about ecosystems.
Then all the data (field descriptions, literature information and results of satellite images interpretation) were compared with ecotopes types. As a result, we selected a corresponding low-disturbed ecosystem for each ecotope type and constructed its map.
Based on field data and remote sensing data we identified four gradations of ecosystem disturbance:
1) undisturbed or low-disturbed ecosystems – more than 50% of principal species dominate;
2) short-derivative ecosystems – associate species makes up 50% or more, principal species are in the first layer;
3) long-derivative ecosystems – undergrowth of principal species, mature trees are absent;
4) highly disturbed ecosystems – open forests and meadows where forests used to be with single surviving trees and shrubs.
There are three main ecosystems in the region: Pinus pityusa, Quercus pubescens and Q. petraea. Juniper sparse forests and petrophytic vegetation on steep slopes occupy small areas and are not revealed at the map scale. Most of the territory is occupied by Pinus pityusa ecosystems (35%). They have an area almost 7 times larger than its natural distribution. However, most of these ecosystems were planted in oak natural area distribution. Quercus pubescens ecosystems occupy 18% of the territory, which is about 56% of their natural distribution. Q. petraea ecosystems occupy 9% of the territory, which is about 36% of their natural distribution.
Currently low-disturbed ecosystems are the least widespread. Most Pinus pityusa ecosystems are long-derivative (51%). Quercus pubescens ecosystems are mainly short-derivative (70%), as are Q. petraea (42%).
Studied ecosystems are located at Caucasus Black Sea coast on the low mountain range (434 m) west of Gelendzhik (Krasnodar region). Based on field research data (about 300 points of landscape and geobotanical descriptions), literature sources and remote sensing data, we identified main ecosystems and assessed their changes.
First, we constructed digital elevation model (DEM) using an area of the topographic map (1:50 000 scale). Based on the DEM, maps of the exposure of the main relief elements and their steepness were constructed. The combined data on steepness and exposure were superimposed on the altitude map (50 m step). As a result, a map of the main ecotopes was obtained. It includes 72 types in the research area. Each polygon on the map was assigned its own number (Id) and corresponding ecotope type. Thus, initial database was formed.
Using high-resolution (1 m) Esri satellite image, Sentinel-2 satellite images (10 m resolution) in visible and shortwave infrared (SWIR) range we divided forest and non-forest ecosystems and coniferous, mixed and broadleaf forests, as well as areas that have been terraced for planting Pinus pityusa and Pinus pallasiana.
We also used NDVI calculated from Sentinel-2 satellite images for the active phase of vegetation period to compare it with field data and clarify information about ecosystems.
Then all the data (field descriptions, literature information and results of satellite images interpretation) were compared with ecotopes types. As a result, we selected a corresponding low-disturbed ecosystem for each ecotope type and constructed its map.
Based on field data and remote sensing data we identified four gradations of ecosystem disturbance:
1) undisturbed or low-disturbed ecosystems – more than 50% of principal species dominate;
2) short-derivative ecosystems – associate species makes up 50% or more, principal species are in the first layer;
3) long-derivative ecosystems – undergrowth of principal species, mature trees are absent;
4) highly disturbed ecosystems – open forests and meadows where forests used to be with single surviving trees and shrubs.
There are three main ecosystems in the region: Pinus pityusa, Quercus pubescens and Q. petraea. Juniper sparse forests and petrophytic vegetation on steep slopes occupy small areas and are not revealed at the map scale. Most of the territory is occupied by Pinus pityusa ecosystems (35%). They have an area almost 7 times larger than its natural distribution. However, most of these ecosystems were planted in oak natural area distribution. Quercus pubescens ecosystems occupy 18% of the territory, which is about 56% of their natural distribution. Q. petraea ecosystems occupy 9% of the territory, which is about 36% of their natural distribution.
Currently low-disturbed ecosystems are the least widespread. Most Pinus pityusa ecosystems are long-derivative (51%). Quercus pubescens ecosystems are mainly short-derivative (70%), as are Q. petraea (42%).