The use of regression tree method for Sentinel-2 satellite data to mapping percent tree cover in different forest types.

Quantifying forest systems is of importance for ecological services and economic benefits in ecosystem models. This study aims to map the percent tree cover (PTC) of various forest stands in the Buyuk Menderes Basin, located in the western part of Turkey with different characteristics in the Mediterranean and Terrestrial transition regions Sentinel-2 data with 10-m spatial resolution. In recent years, some researches have been carried out in different fields to show the capabilities and potential of Sentinel-2 satellite sensors. However, the limited number of PTC researches conducted with Sentinel-2 images reveals the importance of this study. This study aimed to demonstrate reliable PTC data in landscape planning or ecosystem modeling by introducing an advanced approach with high spatial, spectral, and temporal resolution and more cost-effective. In this study, a regression tree algorithm, one of the popular machine learning techniques for ecological modeling, was used to estimate the tree cover's dependent variable based on high-resolution monthly metrics' spectral signatures. Six frames of TripleSat images were used as training data in the regression tree. Monthly Sentinel-2 bands and produced metrics including NDVI, LAI, fCOVER, MSAVI2, and MCARI were almost the first time used as predictor variables. Stepwise linear regression (SLR) was applied to select these predictor bands in the regression tree and a correlation coefficient of 0.83 was obtained. Result PTC maps were produced and the results were evaluated based on coniferous and broadleaf. The results were tested using high spatial resolution TripleSat images and higher model accuracy was determined in both forest types. The high correlation is due to the Sentinel 2 satellite's band characteristics and the metrics are directly related to the tree cover. As a result, the high-accuracy availability of the Sentinel2 satellite is seen to map the PTC on a regional scale, including complex forest types between the Mediterranean and terrestrial transition climates.

Modelling climate change impacts on regional net primary productivity in Turkey.

This study projects and models the terrestrial net primary productivity (NPP) considering the representative concentration pathways (RCPs) scenarios of Turkey using remote-sensing-based biogeochemical modelling techniques. Changes in annual NPP between 2000-2010 and 2070-2080 were projected with the biogeochemical ecosystem model NASA-Carnegie Ames Stanford Approach (CASA). A multi-temporal data set, including 16-day MODIS composites with a spatial resolution of 250 m, was used within the CASA model. The 5th Assessment Report (AR5) of the IPCC presented several scenarios for RCPs named RCP 2.6, RCP 4.5, RCP 6.0, and RCP 8.5 that laid the foundation for the future climate projections. The futuristic NPP modelling was based on the assumptions of maintaining CO2 level in the range of 421 to 936 ppm and a rise in temperature from 1.1 to 2.6 °C. The NPP in Turkey averaged 1232 g C m2 year-1 as per the model results. Considering 2000-2010 as the baseline period, the NPP was modelled within the range of 9.6 and 316 g C m2 year-1. Modelled average NPP was 1332.4 g C m2 year-1 per year between 2061 and 2080. The forest productivity was also estimated to be increased up to 113 g C m-2 year-1 under the climate change scenarios. However, there were minor differences in the projected average NPP under the baseline period covering years from 2000 to 2080 from those under RCPs. It appeared that variation in temperature and precipitation as a result of climate change affected the terrestrial NPP. The regional environmental and socio-economic consequences of climate change on diverse landscapes such as Turkey were properly modelled and analysed to understand the spatial variation of climate change impacts on vegetation. Changes in NPP imply that forests in Turkey could be carbon sinks in the future as their current potential that would profile Turkey's climate mitigation. This is one of the pioneering studies to estimate the future changes of regional NPP in Turkey by integrating various spatial inputs and a biogeochemical model.

Spatial and temporal evaluation of soil erosion in Turkey under climate change scenarios using the Pan-European Soil Erosion Risk Assessment (PESERA) model.

The impacts of climate change on soil erosion are mainly caused by the changes in the amount and intensity of rainfall and rising temperature. The combination of rainfall and temperature change is likely to be accompanied by negative or positive variations in agricultural and forest management. Turkey contains vast fertile plains, high mountain chains and semi-arid lands, with a climate that ranges from marine to continental and therefore is susceptible to soil erosion under climate change, particularly on high gradients and in semi-arid areas. This study aims to model the soil erosion risk under climate change scenarios in Turkey using the Pan-European Soil Erosion Assessment (PESERA) model, predicting the likely effects of land use/cover and climate change on sediment transport and soil erosion in the country. For this purpose, PESERA was applied to estimate the monthly and annual soil loss for 12 land use/cover types in Turkey. The model inputs included 128 variables derived from soil, climate, land use/cover and topography data. The total soil loss from the land surface is speculated to be approximately 285.5 million tonnes per year. According to the IPCC 5th Assessment Report of four climate change scenarios, the total soil losses were predicted as 308.9, 323.5, 320.3 and 355.3 million tonnes for RCP2.6, RCP4.5, RCP6.0 and RCP8.5 scenarios respectively from 2060 to 2080. The predicted amounts of fertile soil loss from agricultural land in a year were predicted to be 55.5 million tonnes at present, and 62.7, 59.9, 61.7 and 58.1 under RCP2.6, RCP4.5, RCP6.0 and RCP8.5 respectively. This confirms that approximately 30% of the total erosion occurs over the agricultural lands. In this respect, degraded forests, scrub and arable lands were subjected to the highest erosion rate (68%) of the total, whereas, fruit trees and berry plantations reflected the lowest erosion rates. Low soil organic carbon, sparse vegetation cover and variable climatic conditions significantly enhanced the erosion of the cultivated lands by primarily removing the potential food for organisms. Finally, process-based models offer a valuable resource for decision-makers when improving environmental management schemes and also decrease uncertainty when considering risks.

Modeling land use/land cover change using remote sensing and geographic information systems: case study of the Seyhan Basin, Turkey.

Land use and land cover (LULC) changes affect several natural environmental factors, including soil erosion, hydrological balance, biodiversity, and the climate, which ultimately impact societal well-being. Therefore, LULC changes are an important aspect of land management. One method used to analyze LULC changes is the mathematical modeling approach. In this study, Cellular Automata and Markov Chain (CA-MC) models were used to predict the LULC changes in the Seyhan Basin in Turkey that are likely to occur by 2036. Satellite multispectral imagery acquired in the years 1995, 2006, and 2016 were classified using the object-based classification method and used as the input data for the CA-MC model. Subsequently, the post-classification comparison technique was used to determine the parameters of the model to be simulated. The Markov Chain analyses and the multi-criteria evaluation (MCE) method were used to produce a transition probability matrix and land suitability maps, respectively. The model was validated using the Kappa index, which reached an overall level of 77%. Finally, the LULC changes were mapped for the year 2036 based on transition rules and a transition area matrix. The LULC prediction for the year 2036 showed a 50% increase in the built-up area class and a 7% decrease in the open spaces class compared to the LULC status of the reference year 2016. About an 8% increase in agricultural land is also likely to occur in 2036. About a 4% increase in shrub land and a 5% decrease in forest areas are also predicted.

Urban change analysis and future growth of Istanbul.

This study is aimed at analyzing urban change within Istanbul and assessing the city's future growth potential using appropriate approach modeling for the year 2040. Urban growth is a major driving force of land-use change, and spatial and temporal components of urbanization can be identified through accurate spatial modeling. In this context, widely used urban modeling approaches, such as the Markov chain and logistic regression based on cellular automata (CA), were used to simulate urban growth within Istanbul. The distance from each pixel to the urban and road classes, elevation, and slope, together with municipality and land use maps (as an excluded layer), were identified as factors. Calibration data were obtained from remotely sensed data recorded in 1972, 1986, and 2013. Validation was performed by overlaying the simulated and actual 2013 urban maps, and a kappa index of agreement was derived. The results indicate that urban expansion will influence mainly forest areas during the time period of 2013-2040. The urban expansion was predicted as 429 and 327 km(2) with the Markov chain and logistic regression models, respectively.

Response of the regression tree model to high resolution remote sensing data for predicting percent tree cover in a Mediterranean ecosystem.

Percent tree cover is the percentage of the ground surface area covered by a vertical projection of the outermost perimeter of the plants. It is an important indicator to reveal the condition of forest systems and has a significant importance for ecosystem models as a main input. The aim of this study is to estimate the percent tree cover of various forest stands in a Mediterranean environment based on an empirical relationship between tree coverage and remotely sensed data in Goksu Watershed located at the Eastern Mediterranean coast of Turkey. A regression tree algorithm was used to simulate spatial fractions of Pinus nigra, Cedrus libani, Pinus brutia, Juniperus excelsa and Quercus cerris using multi-temporal LANDSAT TM/ETM data as predictor variables and land cover information. Two scenes of high resolution GeoEye-1 images were employed for training and testing the model. The predictor variables were incorporated in addition to biophysical variables estimated from the LANDSAT TM/ETM data. Additionally, normalised difference vegetation index (NDVI) was incorporated to LANDSAT TM/ETM band settings as a biophysical variable. Stepwise linear regression (SLR) was applied for selecting the relevant bands to employ in regression tree process. SLR-selected variables produced accurate results in the model with a high correlation coefficient of 0.80. The output values ranged from 0 to 100 %. The different tree species were mapped in 30 m resolution in respect to elevation. Percent tree cover map as a final output was derived using LANDSAT TM/ETM image over Goksu Watershed and the biophysical variables. The results were tested using high spatial resolution GeoEye-1 images. Thus, the combination of the RT algorithm and higher resolution data for percent tree cover mapping were tested and examined in a complex Mediterranean environment.

Quantifying coastal inundation vulnerability of Turkey to sea-level rise.

The vulnerability of low-lying coastal areas in Turkey to inundation was quantified based on the sea-level rise scenarios of 1, 2, and 3 m by 2205. Through digital elevation model (DEM) acquired by the shuttle radar topography mission (SRTM), the extent and distribution of the high to low-risk coastal plains were identified. The spatio-temporal analysis revealed the inundated coastal areas of 545, 1,286, and 2,125 km2 at average rates of 5, 10, and 15 mm yr(-1) for 200 years, respectively. This is equivalent to minimum and maximum land losses by 2205 of 0.1-0.3% of the total area and of 1.3-5.2% of the coastal areas with elevations of less than 100 m in the country, respectively. This study provides an initial assessment of vulnerability to sea-level rise to help decision-makers, and other concerned stakeholders to develop appropriate public policies and land-use planning measures.

Modeling Forest Productivity Using Envisat MERIS Data.

The aim of this study was to derive land cover products with a 300-m pixelresolution of Envisat MERIS (Medium Resolution Imaging Spectrometer) to quantify netprimary productivity (NPP) of conifer forests of Taurus Mountain range along the EasternMediterranean coast of Turkey. The Carnegie-Ames-Stanford approach (CASA) was usedto predict annual and monthly regional NPP as modified by temperature, precipitation,solar radiation, soil texture, fractional tree cover, land cover type, and normalizeddifference vegetation index (NDVI). Fractional tree cover was estimated using continuoustraining data and multi-temporal metrics of 47 Envisat MERIS images of March 2003 toSeptember 2005 and was derived by aggregating tree cover estimates made from high-resolution IKONOS imagery to coarser Landsat ETM imagery. A regression tree algorithmwas used to estimate response variables of fractional tree cover based on the multi-temporal metrics. This study showed that Envisat MERIS data yield a greater spatial detailin the quantification of NPP over a topographically complex terrain at the regional scalethan those used at the global scale such as AVHRR.

Modeling Potential Distribution and Carbon Dynamics of Natural Terrestrial Ecosystems: A Case Study of Turkey.

We derived a simple model that relates the classification of biogeoclimatezones, (co)existence and fractional coverage of plant functional types (PFTs), and patternsof ecosystem carbon (C) stocks to long-term average values of biogeoclimatic indices in atime- and space-varying fashion from climate-vegetation equilibrium models. ProposedDynamic Ecosystem Classification and Productivity (DECP) model is based on the spatialinterpolation of annual biogeoclimatic variables through multiple linear regression (MLR)models and inverse distance weighting (IDW) and was applied to the entire Turkey of780,595 km² on a 500 m x 500 m grid resolution. Estimated total net primary production(TNPP) values of mutually exclusive PFTs ranged from 108 26 to 891 207 Tg C yr-1under the optimal conditions and from 16 7 to 58 23 Tg C yr-1 under the growth-limiting conditions for all the natural ecosystems in Turkey. Total NPP values ofcoexisting PFTs ranged from 178 36 to 1231 253 Tg C yr-1 under the optimalconditions and from 23 8 to 92 31 Tg C yr-1 under the growth-limiting conditions. Thenational steady state soil organic carbon (SOC) storage in the surface one meter of soil wasestimated to range from 7.5 1.8 to 36.7 7.8 Pg C yr-1 under the optimal conditions andfrom 1.3 0.7 to 5.8 2.6 Pg C yr-1 under the limiting conditions, with the national range of 1.3 to 36.7 Pg C elucidating 0.1% and 2.8% of the global SOC value (1272.4 Pg C), respectively. Our comparisons with literature compilations indicate that estimated patterns of biogeoclimate zones, PFTs, TNPP and SOC storage by the DECP model agree reasonably well with measurements from field and remotely sensed data.

Ecological constraints on breeding system evolution: the influence of habitat on brood desertion in Kentish plover.

1. One of the fundamental insights of behavioural ecology is that resources influence breeding systems. For instance, when food resources are plenty, one parent is able to care for the young on its own, so that the other parent can desert and became polygamous. We investigated this hypothesis in the context of classical polyandry when females may have several mates within a single breeding season, and parental duties are carried out largely by the male. 2. We studied a precocial wader, the Kentish plover Charadrius alexandrinus, that exhibits variable brood care such that the chicks may be raised by both parents, only by the female or, more often, only by the male. The timing of female desertion varies: some females desert their brood at hatching of the eggs and lay a clutch for a new mate, whereas other females stay with their brood until the chicks fledge. Kentish plovers are excellent organisms with which to study breeding system evolution, as some of their close relatives exhibit classical polyandry (Eurasian dotterel Eudromias morinellus, mountain plover Charadrius montanus), whereas others are polygynous (northern lapwing Vanellus vanellus). 3. Kentish plovers raised their broods in two habitats in our study site in southern Turkey: saltmarsh and lakeshore. Food intake was higher on the lakeshore than in the saltmarsh as judged from feeding behaviour of chicks and adults. As the season proceeded and the saltmarsh dried out, the broods moved toward the lakeshore. 4. As the density of plovers increased on lakeshore, the parents spent more time defending their young, and female parents stayed with their brood longer on the lakeshore. 5. We conclude that the influence of food abundance on breeding systems is more complex than currently anticipated. Abundant food resources appear to have profound implications on spatial distribution of broods, and the social interactions between broods constrain female desertion and polyandry.

Quantifying carbon budgets of conifer Mediterranean forest ecosystems, Turkey.

Aboveground biomass, aboveground litterfall, and leaf litter decomposition of five indigenous tree stands (pure stands of Pinus brutia, Pinus nigra, Cedrus libani, Juniperus excelsa, and a mixed stand of Abies cilicica, P. nigra, and C. libani) were measured in an eastern Mediterranean evergreen needleleaf forest of Turkey. Measurements were converted to regional scale estimates of carbon (C) stocks and fluxes of forest ecosystems, based on general non-site-specific allometric relationships. Mean C stock of the conifer forests was estimated as 97.8 +/- 79 Mg C ha(-1) consisting of 83.0 +/- 67 Mg C ha(-1) in the aboveground and 14.8 +/- 12 Mg C ha(-1) in the belowground biomass. The forest stands had mean soil organic carbon (SOC) and nitrogen (SON) stocks of 172.0 +/- 25.7 Mg C ha(-1) and 9.2 +/- 1.2 Mg N ha(-1), respectively. Mean total monthly litterfall was 376.2 +/- 191.3 kg C ha(-1), ranging from 641 +/- 385 kg C ha(-1) for Pinus brutia to 286 +/- 82 kg C ha(-1) for Cedrus libani. Decomposition rate constants (k) for pine needles were 0.0016 for Cedrus libani, 0.0009 for Pinus nigra, 0.0006 for the mixed stand, and 0.0005 day(-1) for Pinus brutia and Juniperus excelsa. Estimation of components of the C budgets revealed that the forest ecosystems were net C sinks, with a mean sequestration rate of 2.0 +/- 1.1 Mg C ha(-1) yr(-1) ranging from 3.2 +/- 2 Mg C ha(-1) for Pinus brutia to 1.6 +/- 0.6 Mg C ha(-1) for Cedrus libani. Mean net ecosystem productivity (NEP) resulted in sequestration of 98.4 +/- 54.1 Gg CO2 yr(-1) from the atmosphere when extrapolated for the entire study area of 134.2 km2 (Gg = 10(9) g). The quantitative C data from the study revealed the significance of the conifer Mediterranean forests as C sinks.

Sustainable management for the eastern Mediterranean coast of Turkey.

The objective of this article is to propose a program for the integrated coastal zone management that is required to stimulate and guide sustainable development of the Mediterranean coastal zone of Turkey. Improved data collection, quality control, analysis, and data management will provide a firm basis for future scientific understanding of the East Mediterranean coast of Turkey and will support long-term management. Various innovative procedures were proposed for a promising ecosystem-based approach to manage coastal wetlands in the Mediterranean: remote data acquisition with new technologies; environmental quality monitoring program that will provide a baseline for monitoring; linking a Geographic Information System (GIS) with natural resource management decision routines in the context of operational wetlands, fisheries, tourism management system; environmental sensitivity analysis to ensure that permitted developments are environmentally sustainable; and use of natural species to restore the wetlands and coastal dunes and sustain the system processes. The proposed management scheme will benefit the scientific community in the Mediterranean and the management/planning community in Eastern Turkey.