Open Access Published by De Gruyter Open Access August 20, 2019

Forestry Aridity Index in Vojvodina, North Serbia

Milivoj B. Gavrilov, Tin Lukić, Natalija Janc, Biljana Basarin and Slobodan B. Marković
From the journal Open Geosciences

Abstract

In investigating aridity in Vojvodina (a region in the northern part of Serbia), the Forestry Aridity Index (FAI) was used. This index was chosen due to being one of the most suitable indices for the analysis of the interaction of climate and vegetative processes, especially in forestry. The spatial distribution of the FAI for annual and decennial periods, as well as its annual trend, is analysed. Satisfactory compatibility between the low (forest) and high (steppe) FAI values with the forest and steppe vegetation on the Vojvodina terrains was obtained. The calculated values of the FAI showed that there was no particular annual trend. These results correspond to the earlier calculated values of the De Martonne aridity index and the Pinna combinative index. Therefore, it can be concluded that there were no recent changes in aridity during the observed period. Results of the correlation indicate weak linearity between the FAI, and the North Atlantic Oscillation and El-Niño South Oscillation.

1 Introduction

In recent decades, the issues of climate variability and climate change have been at the center of many scientific studies. Global climate change, caused by natural processes as well as anthropogenic factors, is a major 21st century environmental issue [1, 2, 3, 4]. Governments, the scientific community, as well as the media and its consumers all over the world, have been paying increasingly

more attention to the recent trends regarding global climate change [5]. Temperature change [6], precipitation change [7], and the rate of these changes are the most important variables in determining the possible effects of climate change. However, sometimes it is more useful to analyze temperature and precipitation together. Parameters in which there is a mathematical quotient/ratio between precipitation and some air humidity and temperature values are known as the aridity indices and are used to measure aridity. According to the American Meteorological Society [8], aridity is the degree to which a climate lacks effective, life-promoting moisture. It is the opposite of humidity, in the climate studies usage of the term. Furthermore, the higher the values of the aridity indices in a region, the greater the water resource variability [9]. Aridity indices are not only important indicators for the potential growth and development of plants, but could be good indicators of climate changes, be used to estimate change in runoff [10], and also be a factor in explaining the meteorological/climatological characteristics of a landscape. Also, it is the aridity indices that can help in the recognition and classification of slow climatological natural hazards [11], such as droughts.

Famous aridity parameters that include both temperature and precipitation are the De Martonne [12] aridity index and the Pinna [13] combinative index. The first index can be calculated for different time scales, such as years, seasons and months, while the second of them is calculated only for years. Both aridity indices are used worldwide in order to identify the dry/humid climate conditions of different regions e.g. [9, 14, 15, 16, 17, 18]. They describe the climate in specified regions as well as any climate changes, on the territory of Serbia from 1949 to 2015 [19, 20, 21]. However, some scientists prefer to characterize aridity conditions by using indices based on (potential or reference) evapotranspiration, calculated by using different formulas [22, 23, 24, 25, 26, 27]. Then, for the classification of continental and oceanic climates, the Johansson [28] continentality index and the Kerner [29] oceanity index are used, e.g. [9, 14, 17]. Also, there is class of the aridity indices that are based on relevant data associated with the vegetation process. One of the oldest is the Emberger [30] index, obtained from the mean annual precipitation and the mean temperature of both the coldest and hottest months. The Food and Agriculture Organization (FAO) [31] aridity index is the ratio between total annual precipitation and potential evapotranspiration.

In this paper, the focus will be to research the recently defined Forestry Aridity Index (FAI) published by Führer [32]. The calculation of the FAI takes into account the ratio of the average temperature of the critical months (July to August), the precipitation sums in the main growth cycle of the forest (May to July), and the precipitation sums in the critical months (from July to August). This index can be classified as vegetative because the FAI climatic categories are defined according to the types of forest and steppe vegetation. In order to forecast the future impact of climate variability on the forest vegetation in Vojvodina (North Serbia) usage of climate projections together with forest aridity index would be appropriate approach. Raising questions regarding variability of aridity in the region due to climate change can be focused on obtaining an insight about how these conditions will influence European forest vegetation and how to prepare adequate adaptation and mitigation strategies. Therefore, the FAI is defined for the conditions that exist in Hungary, but due to climatic similarities of the regions, it appears to be applicable to the region of Vojvodina in Northern Serbia as well. Both regions belong to the Carpathian (Pannonian) Basin, making them very similar in their climatic, environmental, and vegetative conditions. The FAI results will also be useful in comparing the results of both the De Martonne and Pinna combinative indices [19], as all three indices were calculated for the same period, from 1949 to 2006, yielding data containing nearly two 30-year climatic cycles. Therefore, the results could act as efficient indicators for recent interpretations of climate. To investigate a possible aridity connection with the North Atlantic Oscillation (NAO) [33] and the El-Niño Southern Oscillation (ENSO) [34], the correlations between the NAO and ENSO indices as well as the FAI, are calculated. Finally, the characteristics of recent aridity are interpreted in the context of paleoclimatic and paleoenvironmental phenomena during the Late Pleistocene.

2 Data and methods

2.1 Study Area

Vojvodina is a region in northern Serbia, located in the southeastern part of the Carpathian (Pannonian) Basin, encompassing the confluence area of the Danube, Sava and Tisa rivers with an area of 21,506 km2 (Figure 1). More than 60% of this lowland area is covered by loess and loess-like sediments [35, 36]. The loess–paleosol sequences, situated in the Vojvodina region, represent the most detailed archive of climatic and environmental fluctuations during the Middle and Late Pleistocene on the European continent [37, 38, 39]. The most distinctive landforms of the Vojvodina region are two mountains: Fruška Gora Mountain,which is situated between the Danube and Sava rivers, and Vršac Mountain, which is located in southeastern part of the region [40]. Sandy areas (Deliblato Sands and Subotica-Horgoš Sand), lower terrains, and alluvial plains are also located in the Vojvodina region [41]. Forests comprise only about 7% of Vojvodina and mainly occur in the mountains and terrains along the banks of rivers, while the agriculture land occupies about 84% [41].

Figure 1 Geographical location of meteorological stations in Vojvodina region ([19], modified)

Figure 1

Geographical location of meteorological stations in Vojvodina region ([19], modified)

The climate of Vojvodina is moderate continental, with cold winters, and hot and humid summers, with huge range of extreme temperatures and a very irregular distribution of rainfall per month, which lead to different values in calculating aridity types. The diversity of the climate of Vojvodina is influenced by surface wind, which blows from two prevailing opposite directions; from NW when it is cold and humid, and from SE when it is warm and dry [41, 42]. The mean annual surface air temperature is about 11C [43, 44] and the annual amount of precipitation is approximately 600 mm [45].

2.2 Data

The aridity indices were calculated from 1949 to 2006 for 10 meteorological stations across the territory of Vojvodina (Table 1). These stations are operated by the Republic Hydrometeorological Service of Serbia [46]. Data sets for each of the stations were analysed and processed for the calculation of the mean monthly values of surface air temperature and monthly amount of precipitation. Thus, a database was created with a time series of temperatures and precipitation values, for the purpose of creating a calculation of aridity indices. Before this analysis, the homogeneity of the precipitation and temperature series was examined, according to the Alexandersson test [47]. This test is based on the assumption that the difference/ratio between temperature and precipitation amounts at the station being tested and the reference series is fairly constant over time. The correlation coefficients between the candidate stations and the reference stations were above 0.7 because of the relatively low and uniformly flat terrain of Vojvodina. The homogeneity analysis showed that the time series of the data for all the stations are homogeneous. It should be noted that the same set of data was used in [19, 48] for

Table 1

List of meteorological stations and their geographical coordinates and altitudes

Number Meteorological station Latitude (N) Longitude (E) Altitude (m)
1 Bački Petrovac 452212′′ 193412′′ 85
2 Bela Crkva 445400′′ 212512′′ 90
3 Jaša Tomić 452700′′ 205100′′ 80
4 Kikinda 455100′′ 202812′′ 81
5 Palić 460600′′ 194612′′ 102
6 Rimski Šančevi 451948′′ 195100′′ 86
7 Senta 455548′′ 200448′′ 80
8 Sombor 454612′′ 190900′′ 87
9 Sremska Mitrovica 450300′′ 193300′′ 82
10 Vršac 450900′′ 211912′′ 83

calculating the De Martonne aridity index and Pinna combinative index.

In order to examine the relationship between the FAI and potential large scale climate drivers NAO and ENSO, linear correlations were utilized. The station based NAO time series for calculating correlation from 1949 to 2006 [49] was downloaded from Climatic Research Unit [50], while the ENSO record was represented by the NINO3 time series for the same time period National Oceanic and Atmospheric Administration [51].

2.2.1 Forestry Aridity Index

The original Forestry Aridity Index [32] is defined as,

(1) F A I = C g × T V I I V I I I ( P V V I I + P V I I V I I I ) ,

where TVIIVIII is the average temperature in July and August in C, PVVII is the precipitation sum in the period from May to July, PVIIVIII is the precipitation sum in the period from July to August, all in mm, and Cg=100 is constant. By introducing the constant with dimensions, Cg=100 mm/C, instead of the previous constant without dimensions, but with the same numerical value, FAI lost its dimensions and became a "true" index, that is, the nondimensional number. With this modification, the FAI has got a better physical basis, and the interpretation of results will not change.

Thus, the FAI can be a very good tool for presenting climate conditions during the yearly forests’ growth,which is very important for forestry and agriculture. With the FAI, the average weather conditions of different climate categories applied in forestry practice are shown in Table 2.

Table 2

Meteorological features of forestry climate categories [32]

FAI value Forestry climate category
less than 4.75 beech climate
4.75-6.00 hornbeam-oak climate
6.00-7.25 the sessile oak / Turkey oak climate
more than 7.25 forest-steppe climate

2.2.2 Aridity Trend

Two statistical approaches are used to investigate the aridity trend. The first approach is to calculate the tendency (trend) equation of the aridity by fitting the linear trend of the forestry aridity index’s mean values, FAI, per year for the entire territory of Vojvodina. Tendency equation (linear trend) is graphically shown with the mean values of the aridity indices per year. In the second statistical approach, the Mann-Kendall (MK) test [52, 53, 54, 55] is used for estimating the significance of the aridity trend which was obtained from the tendency (trend) equation. According to the MK test, two hypotheses were tested: the null hypothesis, H0, that there is no trend in the FAI aridity time series; and the alternative hypothesis, Ha, that there is a significant trend in the series, for a given α significance level [56]. Probability, p, in percent was calculated [43, 44, 57, 58] to determine the level of confidence in the hypothesis. If the computed value p is lower than the chosen significance level α (e.g., α=5%), H0 (there is no trend) should be rejected, and Ha (there is a significant trend) should be accepted. In case p is greater than the significance level α, the H0 (there is no trend) cannot be rejected. MK tests are widely used in environmental sciences, for example: temperature, precipitation, sunshine hours, cloud cover, relative humidity and wind speed [59]; temperature and precipitation [57]; precipitation [60]; extreme temperatures [43, 44, 58, 61]; hail [62, 63, 64]; aridity [19, 21, 65]; evapotranspiration [66]; and atmospheric deposition [67]; because it is simple, robust, and can cope with missing values.

2.2.3 Software

In order to calculate the forestry aridity index, FAI, special software using C Sharp programming language was created [68]. This paper uses numerical approaches for interpolating aridity indices, for the spatial representation and visualization by applying ArcMap software [69]. For calculating the probability, p, and hypothesis testing, XL-STAT’s [70], statistical analysis software was employed.

3 Results and discussion

3.1 Distribution of aridity

The spatial distribution of FAI from 1949 to 2006 is shown in Figure 2. On the basis of the mean of the 58-year long measurement record, it can be concluded that there are only two meteorological stations (Bela Crkva and Vršac) in the sessile oak (Quercus petraea) / Turkey oak (Quercus cerris) climate zone and that the remaining eight stations belong to a forest-steppe climate (Table 2). Aridity is inversely

Figure 2 The spatial distribution of the mean forestry aridity index, FAI, from 1949 to 2006 in Vojvodina

Figure 2

The spatial distribution of the mean forestry aridity index, FAI, from 1949 to 2006 in Vojvodina

proportional to the values of FAI, i.e., if the values of FAI are higher, aridity is lower. Also, lower values of FAI correspond to higher values of De Martonne index and Pinna combinative index [19] and vice versa, which can be concluded on the basis of climate forestry categories. This is contrary to the statement that the higher the values of the aridity indices in a region, the greater the water resource variability [9], which is the consequence of the FAI definition (1),where is sought the ratio between the temperature and the precipitation, while for De Martonne and Pinna combinative indices the reverse is required, the ratio between precipitation and temperature. Since this contradiction has no effect on the results obtained here, it will be ignored, and potential harmonization of the FAI with the previous statement can be left for future research.

The lowest values of FAI, from 6.95 to 7.25, were found in the southeast Vojvodina region, while in the rest of the region values were higher than 7.25. This means that, according to FAI values, a sessile oak / Turkey oak climate is present in the southeast part of Vojvodina. These results correspond to the values of Pinna combinative index and De Martonne index, characterized by humid climate, according to De Martonne climate classification. It is interesting to note that low values of FAI (6.95–7.25) in the southeast part of Vojvodina are more favorable to forest vegetation, which occupies about 20% of the southeast part of the region [71].

Values of the FAI increase to the north, where its values vary from 7.40 to 8.56, and where a forest-steppe climate can be found. It can also be correlated with lower values of the De Martonne and Pinna indices, which indicate semi-humid conditions or semi-dry Mediterranean climate, respectively. Thus, higher values of FAI or forest-steppe climate in the north of Vojvodina correspond to lower values of Pinna combinative index, which is caused by a lower level of precipitation.

The spatial distribution of the FAI for decennial periods is shown in Figure 3 on six maps.

Figure 3 The spatial distribution of the mean forestry aridity index, FAI, per decennial periods from 1949 to 2006 in Vojvodina

Figure 3

The spatial distribution of the mean forestry aridity index, FAI, per decennial periods from 1949 to 2006 in Vojvodina

For the first eight-year period from 1949 to 1956, conditions were very similar to those conditions specific to the annual values in the entire period, shown in Figure 2, and they are characterized by a sessile oak / Turkey oak climate and forest-steppe climate.

The decennial period from 1957 to 1966 is mostly dominated by forest-steppe climate; only small parts of southeastern Vojvodina and northwestern Vojvodina had sessile oak / Turkey oak climate recording values of FAI lower than 7.25.

For the decennial period from 1967 to 1976 values of FAI are significantly lower, indicating the hornbeam-oak (Carpinus betulus) climate in southeastern part of Vojvodina, and sessile oak / Turkey oak climate in the rest of the territory.

Values of the FAI, from 1977 to 1986, show that about 75% of the territory of Vojvodina is dominated by a sessile oak / Turkey oak climate, while a forest-steppe climate is present at only two stations (Kikinda and Senta) in the northeastern region of investigated area, with the highest FAI value being 8.06.

From 1987 to 1996 the forest-steppe climate dominated at nine stations, but the FAI values were very high with a value of 10.75, while only the station at Vršac, in the southeast, had recorded a sessile oak / Turkey oak climate.

Compared to the previous period, from 1997 to 2006, the forest-steppe climate again dominates, now at 8 stations, with the FAI recording the greatest value at 9.21, while a sessile oak / Turkey oak climate is present at only two stations (Jaša Tomić and Rimski Šančevi).

It appears that the obtained results show a significant diversity, since three types of climate aridity have been identified out of a total of four types. This can be considered a significant result, as Vojvodina is a relatively small territory with, at first glance, uniform geophysical characteristics [35, 41], therefore the climate would be expected to be uniform. Also, decennial results show significant mutual differences that quasi periodically alternate, which is very similar to climate variability. For decennial periods, the situation is not so distinctive when compared to the entire period from 1949 until 2006, except for two periods: 1967–1976 and 1977–1986, where calculated FAI values are lower.

3.2 Trend

By calculating the trend from the available mean annual values of the forestry aridity index, the following tendency equation (linear trend curve) is obtained:

(2) y = 0.002 x + 7.672 , + p = 77.8 ,

where y represents the mean annual values of the aridity index, x is the time in years, p is the probability in percent to determine level of confidence in the hypothesis, and the significance level α=5%.

Figure 4 as well as Equation (2) show that there is no trend. Testing the hypothesis will prove whether this statement

Figure 4 The mean annual values of the forestry aridity index, FAI, and linear trend curve of aridity from 1949 to 2006 in Vojvodina

Figure 4

The mean annual values of the forestry aridity index, FAI, and linear trend curve of aridity from 1949 to 2006 in Vojvodina

is true. As the computed p-value in (2) is greater than the significance level (α=5%), one cannot reject the null hypothesis, H0 (there is no trend). The risk to reject H0 while it is true is 77.8%.

Taking into account Equation (2), its coefficients of linear regression, and the MK test result, it can be concluded that the trend of the FAI is negligibly small, recording almost the value of zero, showcasing that there are basically no trends. Viewing the FAI in this way, it seems that there is no change in aridity in Vojvodina from 1949 to 2006. In [19] very similar results of the aridity trends for De Martonne aridity and Pinna combinative indices for the territory of Vojvodina in the same period from 1949 to 2006 were obtained. It seems that aridity trend in Vojvodina really did not change in the long interval of 58 years, which is opposite to trends in temperature [43, 44], precipitation [45], and hail [63], which are very variable. Very similar results of aridity trends have been obtained for the rest of Serbia’s territory; for Kosovo and Metohija [20] and Central Serbia [21], where it is also found that there is no trend.

Thus, it seems that there is no aridity change in Serbia in this recent period, and perhaps this holds even wider in the Carpathian Basin. It is likely to be the consequence of the conservative nature of aridity, when compared to temperature and/or precipitation, that yields the mathematical definition of aridity as being the quotient/ratio between precipitation and some values of air humidity and temperature. As there is plenty of water on Earth, any increase in the temperature of the atmosphere, generally, causes humidity to rise in the atmosphere and vice versa. That is why aridity is a less variable parameter of climate than the others. In general, it seems that greater conservative nature of aridity provides a greater reliability in its calculation, in the assessment of climate and environmental change.

3.3 Correlation

Correlation coefficients between the NAO and ENSO indices, and the FAI were determined in order to investigate the possible relationships between forest aridity and atmospheric variability. Results are shown in Tables 3 and 4. The correlation between the FAI and the NAO index is significant (at a 90% and 99% level of significance) only in winter. Stations like Bela Crkva, Kikinda, and Rimski Šančevi have significant correlation with values varying between 0.21 and 0.25 at the 90% level of significance, while meteorological station Senta has a significant correlation of 0.33 at the 99% level of significance. On the other hand, no significant correlations were observed between

Table 3

Seasonal and annual correlation coefficients (Cors.) with p-values (pvs) between FAI and NAO

Meteorological Winter NAO Spring NAO Summer NAO Autumn NAO Annual NAO
station and region pvs Cors. pvs Cors. pvs Cors. pvs Cors. pvs Cors.
Bački Petrovac 0.26 0.14 0.48 −0.04 0.69 −0.05 0.50 0.08 0.71 0.05
Bela Crkva 0.09* 0.23* 0.77 −0.04 0.66 −0.06 0.12 0.2 0.24 0.16
Jaša Tomić 0.24 0.16 0.43 −0.11 0.78 0.04 0.87 −0.02 0.74 0.04
Kikinda 0.06* 0.25* 0.90 −0.02 0.54 0.08 0.98 0.00 0.43 0.11
Palić 0.48 0.26 0.75 −0.04 0.95 0.01 0.97 0.00 0.52 0.09
Rimski Šančevi 0.10* 0.21* 0.83 −0.03 0.72 −0.05 0.66 −0.06 0.86 0.02
Senta 0.01** 0.33** 0.90 0.07 0.80 0.03 0.78 −0.04 0.28 0.15
Sombor 0.18 0.18 0.42 0.11 0.87 −0.02 0.65 −0.06 0.90 0.00
Sremska Mitrovica 0.49 0.09 0.35 −0.12 0.28 −0.14 0.96 0.007 0.70 −0.05
Vršac 0.60 0.07 0.50 −0.10 0.87 0.02 0.61 0.07 0.73 0.04
Vojvodina 0.16 0.18 0.60 −0.07 0.08 −0.02 0.84 0.03 0.59 0.07

    *90%significance; **99%significance

Table 4

Seasonal and annual correlation coefficients (Cors.) with p-values (pvs) between FAI and ENSO

Meteorological Winter NINO3 Spring NINO3 Summer NINO3 Autumn NINO3 Annual NINO3
Station and region pvs Cors. pvs Cors. pvs Cors. pvs Cors. pvs Cors.
Bački Petrovac 0.79 −0.35 0.90 −0.02 0.32 −0.13 0.50 −0.09 0.47 −0.09
Bela Crkva 0.87 −0.02 0.65 0.06 0.23 −0.15 0.41 −0.11 0.48 −0.09
Jaša Tomić 0.75 0.04 0.70 0.04 0.32 −0.13 0.54 −0.08 0.66 −0.06
Kikinda 0.4 0.11 0.19 0.17 0.65 −0.06 0.50 −0.08 0.96 0.01
Palić 0.46 0.10 0.57 0.07 0.30 −0.14 0.35 −0.12 0.63 −0.06
Rimski Šančevi 0.90 0.02 0.80 −0.03 0.22 −0.16 0.31 −0.13 0.36 −0.12
Senta 0.20 0.17 0.14 0.19 0.33 −0.13 0.32 −0.13 0.86 −0.02
Sombor 0.92 −0.01 0.70 −0.05 0.19 −0.17 0.39 −0.11 0.34 −0.13
Sremska Mitrovica 0.25 −0.15 0.48 −0.09 0.40 −0.11 0.54 −0.08 0.31 −0.13
Vršac 0.83 0.03 0.81 −0.03 0.16 −0.19 0.22 −0.16 0.27 −0.14
Vojvodina 0.86 0.02 0.78 0.04 0.23 −0.15 0.35 −0.12 0.47 −0.10

the ENSO and FAI, thus generally indicating the absence of strong linearity between the FAI, and the NAO and ENSO.

4 Conclusion

With the help of the FAI, we are able to classify the average climate of a particular small area, or even a region, from a forestry viewpoint. For the territory of Vojvodina, which is situated in the southeastern part of the Pannonian Basin, with its rich and widespread loess and loess-like sediments, aridity was investigated from 1949 to 2006. For this entire period, all the temperature and precipitation data was available for 10 meteorological stations that were evenly distributed throughout Vojvodina. Meteorological data was used to determine the forestry aridity index, FAI, as the most suitable for the description of aridity and climate in Vojvodina as a forest area. The values of FAI have been calculated for the entire recent period and six deccenial periods.

Of the four types of climate that the FAI recognizes, there are two types: the sessile oak (Quercus petraea) / Turkey oak climate and the forest-steppe climate, over the course of 58 years; and in the decades periods there are three types of climate, which all together can be considered to be significant climate diversity for a relatively small territory such as Vojvodina. FAI climatic diversity might have shown more varied values if there were meteorological data for mountainous locations in Fruška Gora Mountain and Vršac Mountain, both of which are very rich in forests. On the other hand, there is a significant overlap of the southeast part of Vojvodina as the area with the richest forests (20%) and low FAI values. This seems to indicate that the FAI can be considered as a good indicator of climate interaction and forest vegetation.

The FAI aridity trend was calculated and tested recently, over a period of 58 years. The conclusion is clear: there is no detectable aridity trend, confirming the results of the aridity trends for the De Martonne and Pinna aridity indices, for the same territory and time period. The similar behavior of aridity trends were present in the rest of Serbia; in the territories of Central Serbia as well as Kosovo and Metohija, between 1949–2015. Unlike the temperature and the precipitation, which show variability of trends, aridity does not show any change in trend. Therefore, aridity can be considered a stable climate parameter and indicator, which should be used more frequently in the interpretation of climate change, in order to avoid any hasty conclusions.

As the correlation coefficients calculated between the FAI, the NAO, and the ENSO very small, one can see that these two large-scale processes of climate variability do not have any significant influence on the FAI.

Similar stable semi-arid conditions have been discovered to have occurred in the Late Pleistocene paleoclimatic and paleoenvironmental records in the region of Vojvodina [72]. The last glacial loess snail fauna in the investigated area, which is characterized by a complete absence of any cold-resistant species, suggests a stable, dry, and relatively warm glacial summer climate, when compared to other European loess provinces [35, 73, 74, 75]. Two independent, environmental pieces of evidence from this loess-paleosol section, novel n-alkane biomarkers and traditional land snail assemblages, independently confirm a relatively stable paleoenvironment during the past 150 kyr. One characterized by the continuous dominance of grassland vegetation, with periodic intervals involving forest elements [72]. Using stable carbon isotopes (δ13C) of organic matter in reconstructing Quaternary climate variability based on loess are widely used methods and the most important factors controlling the isotopic signal in soils and sediments implying the changes in the photosynthetic pathway (shifts from C3 to C4 vegetation), physiological water stress, and changes in the atmospheric CO2 concentration. On the other hand, the isotopic signature of vegetation provides information on photosynthetic pathways (C3 versus C4) and, thus, on environmental changes that are a prelude to the replacement of one vegetation type by another. Nevertheless, and in contrast to all European loess sequences that were recorded along the last climatic cycle with large C3 plant dominance, the organic δ13C record of the Surduk section was, until now, only a glacial record with evidence of several arid periods associated with the presence of C4 plants [76, 77].

Acknowledgement

This research paper was supported by Project 176020 of the Serbian Ministry of Education and Science and by Project 142-451-2511/2017- 02 of Provincial Secretariat for Science and Technological Development, Vojvodina Province. Part of the research was supported by the HUSRB/1602/11/0057 – WATER@RISK – Improvement of drought and excess water monitoring for supporting water management and mitigation of risks related to extreme weather conditions. The authors are grateful to Ivana Hrnjak and Valentina Janc for the support.

References

[1] Hulme M., Barrow E.M., Arnell N.W., Harrisson P.A., Johns T.C., Downing T.E., Relative impacts of human-induced climate change and natural climate variability. Nature, 1999, 397, 688–691 Search in Google Scholar

[2] Kenny G.J., Ye W., Flux T., Warrick R.A., Climate variations and New Zealand agriculture: the CLIMPACTS system and issues of spatial and temporal scale. Environmental International, 2001, 27, 189–194 Search in Google Scholar

[3] Baltas E, Mimikou M., Climate change impacts on the water supply of Thessaloniki. International Journal of Water Resources Development, 2005, 21, 341–353 Search in Google Scholar

[4] The Intergovernmental Panel on Climate Change, Global warming of 1.5°C. An IPCC Special Report on the impacts of global warming of 1.5°C above pre-industrial levels and related global greenhouse gas emission pathways, in the context of strengthening the global response to the threat of climate change, sustainable development, and efforts to eradicate poverty. World Meteorological Organization, Geneva, Switzerland, 2018 Search in Google Scholar

[5] Gavrilov M.B., Marković S.M., Mladjan D., Zarić M., Pešić A., Janc N., Todorović N., GLOBAL WARMING – BETWEEN THE MYTH AND REALITY. In: Kolarić D (Ed.), Proceedings of the International Scientific Conference “Archibald Reiss Days“, Zemun (Belgrade), Serbia. Academy of Criminalistic And Police Studies, Belgrade, 2016, 2, 304-313 Search in Google Scholar

[6] Climatic Research Unit, Global average temperature change 1856–2003, 2003, http://www.cru.uea.ac.uk/cru/data/temperature/ Search in Google Scholar

[7] Diaz H.F., Bradley R., Eischeid J.K., Precipitation fluctuation over global land areas since the late 1800s. J. Geophys. Res. 1989, 94, 1195–1210 Search in Google Scholar

[8] American Meteorological Society, Glossary of Meteorology. In: Huschke, R.E. (Ed.), American Meteorological Society, Boston, 2000 Search in Google Scholar

[9] Deniz A., Toros H., Incecik S., Spatial variations of climate indices in Turkey. International Journal of Climatology, 2011, 3, 394–403 Search in Google Scholar

[10] Arora V.K., The use of the aridity index to assess climate change effect on annual runoff. Journal of Hydrology, 2002, 265, 164–177 Search in Google Scholar

[11] Lukić T., Gavrilov M.B., Marković S.B., Zorn M., Komac B., Mladjan D., Djordjević J., Milanović M., Vasiljević Dj.A., Vujićič M.D., Kuzmanović B., Prentović R., Classification of the natural disasters between the legislation and application: experience of the Republic of Serbia. Acta Geographica Slovenica, 2013, 53-1, 149–164, 10.3986/AGS53301 Search in Google Scholar

[12] De Martonne E., Traité de Géographie Physique: 3 tomes. Ch. Dunod et P.Vick. Editeurs, Paris, 1925 Search in Google Scholar

[13] Zambakas J., General climatology. Department of Geology, National & Kapodistrian University of Athens, Athens, 1992 Search in Google Scholar

[14] Baltas E, Spatial distribution of climatic indices in northern Greece. Meteorological Applications, 2007, 14, 69–78 Search in Google Scholar

[15] Croitoru A. E., Piticar A., Imbroane A. M., Burada D.C., Spatiotemporal distribution of aridity indices based on temperature and precipitation in the extra-Carpathian regions of Romania. Theoretical and Applied Climatology, 2013, 112, 597–607, doi:10.1007/s00704-012-0755-2 Search in Google Scholar

[16] Moral F.J., Rebollo F.J., Paniagua L.L., García-Martín A., Honorio F., Spatial distribution and comparison of aridity indices in Extremadura, southwestern Spain. Theoretical and Applied Climatology, 2016, 126, 801–814, https://doi.org/10.1007/s00704-015-1615-7 Search in Google Scholar

[17] Nistor M.M, Spatial distribution of climate indices in the Emilia-Romagna region. Meteorological Applications, 2016, 23, 304– 313 Search in Google Scholar

[18] Gebremedhin M.A., Hailu Kahsay G.H., Fanta H.G., Assessment of spatial distribution of aridity indices in Raya valley, northern Ethiopia. Applied Water Science, 2018, 217–224 Search in Google Scholar

[19] Hrnjak I., Lukić T., Gavrilov M.B., Marković S.B., Unkašević M., Tošić I., Aridity in Vojvodina, Serbia. Theoretical and Applied Climatology, 2014, 115, 323–332, 10.1007/s00704-013-0893-1 Search in Google Scholar

[20] Bačević N., Vukoičić D., Nikolić M., Janc N., Milentijević N., Gavrilov M.G., Aridity in Kosovo and Metohija, Serbia. Carpathian Journal of Earth and Environmental Sciences, 2017, 12, 563–570 Search in Google Scholar

[21] Radaković M., Tošić I., Bačević N., Mladjan D., Gavrilov M.B., Marković, S.B., The analysis of aridity in central Serbia from 1949 to 2015, Theoretical and Applied Climatology, 2018, 133, 887–898, 10.1007/s00704-017-2220-8 Search in Google Scholar

[22] Thornthwaite C.W., An approach toward a Rational classification of climate. Geogr. Rev., 1948, 38, 55–94 Search in Google Scholar

[23] Palmer W., Meteorological Drought. U.S. Department of Commerce Weather Bureau, Washington DC, 1965 Search in Google Scholar

[24] Ben-Gai T., Bitan A., Manes A., Alpert P., Long-term changes in annual rainfall patterns in southern Israel. Theoretical and Applied Climatology, 1994, 49, 59–67, doi:10.1007/BF00868190 Search in Google Scholar

[25] Cook E.R„ Woodhouse C.A., Mark E.C., Meko D.M., Stahle D.W., Long-term aridity changes in the western United States. Science, 2004, 306, 1015–1018 Search in Google Scholar

[26] Wu S., Yin Y., Zheng D., Yang Q., Moisture conditions and climate trends in China during the period 1971–2000. International Journal of Climatology, 2006, 26, 193–206 Search in Google Scholar

[27] Tabari H., Aghajanloo M-B., Temporal pattern of monthly aridity index in Iran with considering precipitation and evapotranspiration trends. International Journal of Climatology, 2013, 33, 396–409 doi:10.1002/joc.3432 Search in Google Scholar

[28] Johansson O.V., Die Hauptcharakteristika des jahrlichen Temperaturganges. Gerlands Beitr. Z. Journal of Geophysics, 1931, 33, 406–428 Search in Google Scholar

[29] Kerner F.V., Thermoisodromen, Versuch einer kartographischen Darstellung des jahrlichen Ganges der Lufttemperatur (Wien). Geographische Gesellschaft, 1905, 6 Search in Google Scholar

[30] Emberger E., Sur une formule climatique et ses applications en botanique. La Météorologie, 1932, 423–432 Search in Google Scholar

[31] Food and Agriculture Organization of the United Nations, World Map of Desertification. United Nations, Rome, 1977 Search in Google Scholar

[32] Führer E., Horvath L., Jagodics A., Machon A., Szabados I., Application of new aridity index in Hungarian forestry practice. Quarterly Journal of the Hungarian Meteorological Service, 2011, 115, 205–216 Search in Google Scholar

[33] Hurrell J.W., Decadal trends in the North Atlantic Oscillation regional temperatures and precipitation. Science, 1995, 269, 676– 679 Search in Google Scholar

[34] Bjerknes J., Atmospheric teleconnections from the equatorial Pacific. Monthly Weather Review, 1969, 97, 163-172 Search in Google Scholar

[35] Marković S.B., Bokhorst M., Vandenberghe J., Oches E.A., Zöller L., McCoy W.D., Gaudenyi T., Jovanović M., Hambach U., Machalett B., Late Pleistocene loess-paleosol sequences in the Vojvodina region, North Serbia. Journal of Quaternary Science, 2008, 23, 73–84 Search in Google Scholar

[36] Vasiljević Dj.A., Marković S.B., Hose T.A., Smalley I., O’Hara Dhand K., Basarin B., Lukić T., Vujičić M.D., Loess Towards (Geo) Tourism-Proposed Application on Loess in Vojvodina Region (North Serbia). Acta Geographica Slovenica, 2011, 51, 391–406 Search in Google Scholar

[37] Marković S.B., Hambach U., Catto N., Jovanović M., Buggle B., Machalett B., Zöller L., Glaser B., Frechen M., The middle and late Pleistocene loess-paleosol sequences at Batajanica, Vojvodina, Serbia. Quaternary International, 2009, 198, 255–266 Search in Google Scholar

[38] Marković S.B., Hambach U., Stevens T., Kukla G.J., Heller F., William D., McCoy W.D., Oches E.A., Buggle B., Zöller L., The last million years recorded at the Stari Slankamen loess-palaeosol sequence: revised chronostratigraphy and long-term environmental trends. Quaternary Science Reviews, 2011, 30, 1142–1154 Search in Google Scholar

[39] Marković S.B., Stevens T., Kukla G.J., Hambach U., Fitzsimmons K.E., Gibbard P., Buggle B., Zech M., Guo Z.T., Hao Q.Z., Wu H., O’Hara-Dhand K., Smalley I.J., Ujvari G., Sümegi P., Timar-Gabor A., Veres D., Sirocko F., Vasiljević Dj.A., Jari Z., Svensson A., Jović V., Kovács J., Svirčev Z., The Danube loess stratigraphy - new steps towards the development of a pan-European loess stratigraphic model. Earth Science Reviews, 2015, 148, 228–258 Search in Google Scholar

[40] Vujičić M.,D, Vasiljević Dj,A„ Marković S,B„ Hose T,A., Lukić T, Hadžić O., Janićević S., Preliminary Geosite Assessment Model (GAM) and its Application on Fruška Gora Mountain, Potential Geotourism Destination of Serbia. Acta Geographica Slovenica, 2011, 51, 361–376 Search in Google Scholar

[41] Gavrilov M.B., Marković S.B., Schaetzl R.J., Tošić I,A., Zeeden C., Obreht I., Sipos Gy., Ruman A., Putniković S., Emunds K., Perić Z., Hambach U., Lehmkuhl F., Prevailing surface winds in Northern Serbia in the recent and past time periods; modern-and past dust deposition. Aeolian Research, 2018, 31, 117–129, DOI:10.1016/j.aeolia.2017.07.008 Search in Google Scholar

[42] Tošić I., Gavrilov M.B., Marković S.B., Ruman A., Putniković S., Seasonal prevailing surface winds in Northern Serbia. Theoretical and Applied Climatology, 2018, 131, 1273–1284, 10.1007/s00704-017-2044-6 Search in Google Scholar

[43] Gavrilov M.B., Marković S.B., Korać V., Jarad A., The analysis of temperature trends in Vojvodina (Serbia) from 1949 to 2006. Thermal Science, 2015, 19, 339–350 Search in Google Scholar

[44] Gavrilov M.B., Tošić I., Marković S.B., Unkašević M., Petrović P., The analysis of annual and seasonal temperature trends using the Mann-Kendall test in Vojvodina, Serbia. Időjárás, 2016, 120, 183–198 Search in Google Scholar

[45] Tošić I., Hrnjak I, Gavrilov M.B., Unkašević M., Marković S.B., Lukić T., Annual and seasonal variability of precipitation in Vojvodina, Serbia. Theoretical and Applied Climatology, 2014, 117, 331–341, 10.1007/s00704-013-1007-9 Search in Google Scholar

[46] Hydrometeorological Service of Serbia, 2012 http://www.hidmet.gov.rs/ Search in Google Scholar

[47] Alexandersson H., A homogeneity test applied to precipitation data. J. Climatol., 1986, 6, 661–675 Search in Google Scholar

[48] Hrnjak I., Gavrilov M.B., Marković S.B., Lukić T., Tošić I., Unkašević M., Special software for aridity indices calculation (AICS); Vojvodina, Serbia case study. Journal of the Geographical Institute Jovan Cvijic, 2013, 63, 83–94, doi:10.2298/IJGI1303083H Search in Google Scholar

[49] Jones P.D., Jónsson T., Wheeler D., Extension to the North Atlantic Oscillation using early instrumental pressure observations from Gibraltar and South-West Iceland. International Journal of Climatology, 1997, 17, 1433–1450 Search in Google Scholar

[50] Climatic Research Unit, 2018 https://crudata.uea.ac.uk/cru/data/nao/ Search in Google Scholar

[51] National Oceanic and Atmospheric Administration, 2018 http://www.cpc.noaa.gov/data/indices/sstoi.indices Search in Google Scholar

[52] Kendall M.G, New Measure of Rank Correlation. Biometrika, 1938, 30, 81–93 Search in Google Scholar

[53] Kendall M.G., Rank correlation methods, 4th edn. Charles Griffin, London, 1975 Search in Google Scholar

[54] Mann H.B., Non-parametric tests against trend. Econometrica, 1945, 13, 245–259 Search in Google Scholar

[55] Gilbert R.O., Statistical Methods for Environmental Pollution Monitoring. Van Nostrand Reinhold, New York, 1987 Search in Google Scholar

[56] Onoz B., Bayazit M., The Power of Statistical Tests for Trend Detection. Turkish Journal of Engineering and Environmental Sciences, 2003, 27, 247–251. Search in Google Scholar

[57] Karmeshu N., Trend detection in annual temperature & precipitation using the Mann Kendall test – a case study to assess climate change on select states in the northeastern United States. Master’s thesis, University of Pennsylvania, Philadelphia, 2012 Search in Google Scholar

[58] Gavrilov M.B., Marković S.B., Janc N., Nikolić M., Valjarević A., Komac B., Zorn M., Punišić M., Bačević N., Assessing average annual air temperature trends using the Mann–Kendall test in Kosovo. Acta Geographica Slovenica, 2018, 58, 7–25, DOI:10.3986/AGS.1309 Search in Google Scholar

[59] Marin L., Birsan M.V., Bojariu R., Dumitrescu A., Micu D.M., Manea A., An Overview of annual climatic changes in Romania: trends in air temperature, precipitation, sunshine hours, cloud cover, relative humidity and wind speed duration the 1961–2013period. Carpathian Journal of Earth and Environmental Sciences, 2014, 9, 253–258. Search in Google Scholar

[60] Çiçek İ., Duman N., Seasonal and annual precipitation trends in Turkey. Carpathian Journal of Earth and Environmental Sciences, 2015. 10, 77-84 Search in Google Scholar

[61] Serra C., Burgueño A., Lana X., Analysis of maximum and minimum daily temperatures recorded at Fabra Observatory (Barcelona, NE Spain) in the period 1917-1998. International Journal of Climatology, 2001, 21, 617–636, doi: 10.1002/joc.633 Search in Google Scholar

[62] Gavrilov M.B, Lazić L., Pešić A., Milutinović M., Marković D., Stanković A., Gavrilov M.M., Influence of Hail Suppression on the Hail Trend in Serbia. Physical Geography, 2010, 31, 441–454 Search in Google Scholar

[63] Gavrilov M.B., Lazić L., Milutinović M., Gavrilov M.M., Influence of Hail Suppression on the Hail Trend in Vojvodina, Serbia. Geographica Pannonica, 2011, 15, 36–41 Search in Google Scholar

[64] Gavrilov M.B.,Marković S.B., Zorn M., Komac B., Lukić T., Milošević M., Janićević S., Is hail suppression useful in Serbia? – General review and new results. Acta Geographica Slovenica, 2013, 53, 165–179, 10.3986/AGS53302 Search in Google Scholar

[65] Ahmed K., Shahid S., Wang X.-J., Nawaz, N., Khan N., Spatiotemporal Changes in Aridity of Pakistan during 1901-2016. Hydrology and Earth System Science Discusion, 2019, (in press), https://doi.org/10.5194/hess-2018-642 Search in Google Scholar

[66] Tabari H., Marofi S., Aeini A., Talaee P.H., Mohammadi K., Trend analysis of reference evapotranspirationin the western half of Iran. Agricultural and Forest Meteorology, 2011, 151, 128–136, doi: 10.1016/j.agrformet.2010.09.009 Search in Google Scholar

[67] Drapela K., Drapelova I., Application of Mann-Kendall test and the Sen’s slope estimates for trend detection in deposition data from Bílý Kříž (Beskydy Mts., the Czech Republic) 1997–2010. Beskydy, 2011, 4, 133–146. Search in Google Scholar

[68] Sharp J., Microsoft Visual C# 2010 Step By Step, Microsoft Press, US, 2010 Search in Google Scholar

[69] Johnston K., Ver Hoef J.M., Krivoruchko K., Lucas N., Using ArcGIS Geostatistical Analyst. ESRI, Redlands, CA, 2001 Search in Google Scholar

[70] XLSTAT’s, 2013 http://www.xlstat.com/en/ Search in Google Scholar

[71] Public Company "VOJVODINAŠUME": The Forest Holding "Banat" 2019 http://www.banatsume.rs/ Search in Google Scholar

[72] Marković S.B., Sümegi P., Stevens T., Schaetzl R.J., Obreht I., Chu W., Buggle B., Zech M., Zech R., Zeeden C., Gavrilov M.B., Perić Z., Svirčev Z., Lehmkuhl F., The Crvenka loess-paleosol sequence: A record of continuous grassland domination in the southern Carpathian Basin during the Late Pleistocene. Palaeogeography, Palaeoclimatology, Palaeoecology, 2018, 509, 33–46 Search in Google Scholar

[73] Marković S.B., McCoy W.D., Oches E.A., Savić S., Gaudenyi T., Jovanović M., Stevens T., Walther R., Ivanišević P., Galić Z., Paleoclimate record in the Late Pleistocene loess-paleosol sequence at Petrovaradin Brickyard (Vojvodina, Serbia). Geologica Carpathica, 2005, 56, 483-491 Search in Google Scholar

[74] Marković S.B., Oches E., Sümegi P., Jovanović M., Gaudenyi T., An introduction to the Upper and Middle Pleistocene loess-paleosol sequences of Ruma section (Vojvodina, Serbia). Quaternary International, 2006, 149, 80–86 Search in Google Scholar

[75] Marković S.B., Oches E.A., McCoy W.D., Gaudenyi T., Frechen M., Malacological and sedimentological evidence for “warm” climate from the Irig loess sequence (Vojvodina, Serbia). Geochemistry, Geophysics, Geosystems, 2007, 8, Q09008. http://dx.doi.org/10.1029/2006GC001565 Search in Google Scholar

[76] Hatté C., Gauthier C., Rousseau D.D., Antoine P., Fuchs M., Lagroix F., Marković S.B., Moine O., Sima A., Excursions to C4 vegetation recorded in the Upper Pleistocene loess of Surduk (Northern Serbia): an organic isotope geochemistry study. Climate of the Past, 2013, 9, 1001–1014. Search in Google Scholar

[77] Obreht I., Zeeden C., Hambach U., Veres D., Marković S.B., Lehmkuhl F., A critical reevaluation of palaeoclimate proxy records from loess in the Carpathian Basin. Earth Science Reviews, 2019, 190, 498–520 Search in Google Scholar

Received: 2019-03-27
Accepted: 2019-06-26
Published Online: 2019-08-20

© 2019 M. B. Gavrilov et al., published by De Gruyter

This work is licensed under the Creative Commons Attribution 4.0 Public License.