Breeding bird communities of the Talysh mountains (Azerbaijan) and ...

Ernst-Moritz-Arndt-University Greifswald

The breeding bird communities of the Talish mountains (Azerbaijan) and their response to forest degradation

Diploma thesis in the study programme Landscape Ecology and Nature Conservation

Michael Heiß -June 2010Supported by DAAD Michael Succow Foundation

Supervised by Prof. Dr. Michael Succow Dr. Martin Flade

I

Front: Semi-collared Flycatcher (Ficedula semitorquata) in primeval forest. Siov, 27.04.2008. Contact: [email protected]

I

Contents 1

Introduction and research questions .....................................................................................- 1 -

2

Study site ..............................................................................................................................- 4 -

3

4

5

2.1

Location .......................................................................................................................- 4 -

2.2

Climate.........................................................................................................................- 5 -

2.3

Landscape types ...........................................................................................................- 6 -

Methods ..............................................................................................................................- 15 3.1

Bird sampling.............................................................................................................- 15 -

3.2

Habitat sampling ........................................................................................................- 17 -

3.3

Data analysis ..............................................................................................................- 17 -

3.3.1

Data treatment ...................................................................................................- 17 -

3.3.2

Species richness ................................................................................................- 18 -

3.3.3

Calculation of relative abundances ...................................................................- 18 -

3.3.4

Nesting guilds....................................................................................................- 19 -

3.3.5

Statistical analysis .............................................................................................- 19 -

Results ................................................................................................................................- 21 4.1

Species number and species richness.........................................................................- 21 -

4.2

Breeding bird communities........................................................................................- 24 -

4.3

Parameters influencing breeding bird communities...................................................- 32 -

4.4

Relative abundances of bird species ..........................................................................- 35 -

4.5

Response to forest degradation on species-level........................................................- 40 -

4.6

Response to forest degradation on guild-level...........................................................- 42 -

Discussion...........................................................................................................................- 44 5.1

Breeding bird communities........................................................................................- 44 -

5.2

Parameters influencing bird communities..................................................................- 46 -

5.3

Species and species richness......................................................................................- 50 -

5.4

Conclusion and conservation implications ................................................................- 52 -

5.5

Scope and limitations.................................................................................................- 54 -

6

Summary.............................................................................................................................- 55 -

7

Zusammenfassung ..............................................................................................................- 56 -

8

Acknowledgements ............................................................................................................- 58 -

9

References ..........................................................................................................................- 59 -

10

Appendices.....................................................................................................................- 67 II

Figures Figure 1: Location of Azerbaijan and the study site………………………………………………4 Figure 2: Climate diagram of Lenkoran…………………………………………………………...5 Figure 3: Climate diagram of Ardebil……………………………………………………………..5 Figure 4: Schematic of the vertical and horizontal distribution of the landscape types…………...7 Figure 5: The Caspian lowland…………………………………………………………………….8 Figure 6: Drainage channel………………………………………………………………………...8 Figure 7: Widespread natural forest stage of the Caspian forest…………………………………10 Figure 8: Grazing is common in the park-like forest stage……………………………………….11 Figure 9: Intense grazing and lopping ……………………………………………………………11 Figure 10: Hay making in the montane meadow belt…………………………………………….12 Figure 11: Montane semi-desert of the Zuvand…………………………………………………..13 Figure 12: Thorny cushion-forming tragacanthic vegetation……………………………………..13 Figure 13: Like oases meander riparian forests through the montane semi-desert……………….14 Figure 14: Lush meadows of the riparian forest………………………………………………….14 Figure 15: Rocky habitats with cushion-forming plant species…………………………………..14 Figure 16: Rocky habitats above 2000 m a.s.l……………………………………………………14 Figure 17: Data distribution per date……………………………………………………………..15 Figure 18: Overview of the study site…………………………………………………………….22 Figure 19: Species-sampling effort relationship illustrated by the landscape types……………...23 Figure 20: The nine breeding bird communities of the Talish mountains………………………..24 Figure 21: Ordination of breeding bird communities…………………………………………….33 Figure 22: Ordination of breeding bird communities within the forest degradation stages………34 Figure 23: Boxplots of height and cover of each vegetation layer for every community………...36 Figure 24: Boxplots of the altitudinal distribution of breeding bird species……………………..37 Figure 25: Relative abundances values of the 15 most common bird species per landscape type (forest degradation stages)………………………………………………………………………...38 Figure 26: Relative abundances values of the 15 most common bird species per landscape type (outside the forest belt)…………………………………………………………………………….39 Figure 27: Negatively and strongly negatively response of selected forest bird species…………41 Figure 28: Positive response of selected forest bird species to forest degradation……………….42 Figure 29: Response of nesting guilds to forest degradation……………………………………..43 Figure 30: Breeding bird response to forest degradation…………………………………............47 III

Tables Table 1: Totals of surveyed transect length and totals of found breeding bird species per landscape types…………………………………………………………………………………………….…22 Table 2: Frequency table of the breeding bird communities……………………………………...25

Appendices Annex 1: Commented list of the observed bird species of the Talish mountains region…………67 Annex 2: Nesting guilds of selected bird species…………………………………………………74 Annex 3: Bird community table of the nine breeding bird communities………...……………….75 Annex 41: All indicator species of each community and every combination of communities…….76 Annex 5: Statistical analysis of the site parameters of the NMDS including all transects………..78 Annex 6: Statistical analysis of the site parameters of the NMDS including all forest transects...78 Annex 7: Relative abundance values (territory/km) of each bird species per landscape type……79 Annex 8: Relative abundance values (individuals/km) of each bird species per landscape type…81

IV

Introduction


1 Introduction and research questions Deciduous broadleaf forests of the northern hemisphere are one of the most threatened ecosystems on earth. They occur in three major, disjunct expressions (western Eurasia, eastern Asia and eastern North America). An increasing human population and their increasing demands for food and biofuels cause a rapid decline and fragmentation of these forests due to agriculture, livestock farming, fuel wood gathering and timber exploitation. Large areas of deciduous broadleaf forests are already converted into coniferous woodlands or have been cleared for agriculture and pastures. Primeval deciduous broadleaf forests, which never suffered a human impact, are nowadays rare and isolated. A primeval forest, once cut, is irreversible extinct, it can never be reconstructed (Knapp 2005). These forests survived in western Eurasia only in small isolated remnants, mostly in inaccessible areas or within large forested areas, for example Białowieża Forest at the Polish/Belarussian border (Tomiałojć & Wesołowski 2004), Šúr National Nature Reserve in Slovakia (Korňan 2009), Colchic forests in Georgia and partly the Caspian forests in Azerbaijan and Iran (Knapp 2005). The largest remnants of primeval deciduous broadleaf forests on earth can be found in northern Iran and the adjacent Azerbaijan and is called the Caspian or Hirkanian forest. This forest is regarded as tertiary relict forest with a remarkable biodiversity including ca. 90 tree and 211 shrub species with a high proportion of endemic species (Prilipko 1954, Knapp 2005, Seifollahian 2005). 100000 ha (hectare) of pristine forest can be found within the 1.8 million ha sized forest belt at the northern slopes of the Alborz mountains in Iran (Knapp 2005). Globally, the Caspian forest is one of the most important deciduous broadleaf forests, due to its remarkably biodiversity, primeval conditions and large extension. In Iran are 590000 ha of Caspian forest under the protection of 1 national park, 5 nature reserves and 3 natural monuments besides further forest reserves and wildlife refugees (Knapp 2005). Even outside the protected areas, the Caspian forest in Iran appears naturally due to its sustainable use supported by laws for a natural forestry, forestry plenter-use, different aged deciduous forest, indigenous tree

-1-


Introduction

species, resettlement of the wood population out of the forest, reduction of silvopastures (Abdollahpour & Atui 2005, Mohadjer 2005, Saffari 2005). Azerbaijan holds a total of about 100000 ha of the Caspian forest (Michael Succow Stiftung 2009), but the current situation of the Caspian forest in Azerbaijan is different to that in Iran. After the breakdown of the Soviet Union in 1991, no laws for protecting the forest have been reintroduced followed by an absence of a regular forestry or management plans. Timber exploration is for many inhabitants the only monetary income after Russian kolkhozes, as main employer, disappeared. Thus, an ‘open access’ situation arose, which led to a rapid forest degradation caused by overusing the forest (Noack & Hidayatov 2007, Michael Succow Stiftung 2009). Further reasons for the degradation are silvopastures and tree lopping for livestock fodder and fuel wood gathering (Scharnweber et al. 2007). Currently, only 17% are treated as pristine forest, whereas 44% show a heavy degradation (Rietschel 2010). The size of the Hirkan National Park in Azerbaijan was enlarged in 2008 to 40358 ha (MENR 2010). No information is actually available about the impact of the ongoing degradation to the fauna of the Caspian forest. Birds are useful bioindicators. Their ecological behaviour is widely known and they respond fast to a changing environment (Flade 1994). The results and understandings of the bird’s response to forest degradation can be used to derive reactions of the whole fauna of that ecosystem and subsequently guide further nature conservation assignments. Since about 150 years, the Caspian forest of Azerbaijan is of ornithological interest. Its scenically beauty and the uniqueness of its flora and fauna within the country and the former Soviet Union led to several biological studies. Gustav Radde, a German in Russian service, was one of the first who described in detail the avifauna of the region (Radde 1884, Radde 1886a, Radde 1886b) followed by studies of several Russian biologists. The last intense researches of the avifauna were conducted in the 1970s (e.g. Agaeva 1972, Agaeava & Mustafaev 1973, Loskot 1978, Agaeva 1979, Mustafaev & Agaeva 1981). Nevertheless, there are many further gaps of knowledge according to the occurrence and distribution of birds. Several species listed by Radde (1886b), for example Caspian Snow-

-2-

Introduction


cock (Tetraogallus caspius), Crimson-winged Finch (Rhodopechys sanguinea) or Radde’s Accentor (Prunella ocularis), seemed to be disappeared in the last 130 years, as no recent records exist. The Caspian Snowcock was found by Radde (1884, 1886a, 1886b) and Patrikeev (2004) expected this species to be extirpated in the Talish mountains. The occurrence of Shikras (Accipiter badius) in that region was unclear for a long time (Patrikeev 2004, Gregory 2007), but could recently clarified as a regular breeding bird of the Caspian lowland (Heiss & Gauger 2009). The occurrence of further species, like Phasianus colchicus talischensis, Gypaetus barbatus, Apus affinus, Halcyon smyrnensis, Picus canus, Phoenicurus erythrogaster or Tichodroma muraria, are up to day unclear. Species responses to forest degradation, fragmentation and deforestation have garnered much recent interest throughout all continents and all forest types (e.g. Edenius & Elmberg 1996, Poulsen 2002, Sekercioglu 2002, Echeverria et al. 2007, Fuller et al. 2007b, Murakami et al. 2008). The impact of the degradation of the Caspian forest to the fauna is unknown. No detailed studies are published yet. Therefore, the aim of this study is to answer the following questions: •

What are the breeding bird communities of the Talish mountains and which species do they contain?

•

Which parameters are responsible for their species composition?

•

How do the breeding birds respond to the degradation of the Caspian forest in Azerbaijan?

•

What are the derived nature conservation implications?

-3-


Study site

2 Study site 2.1 Location The study site is located in the southernmost part of Azerbaijan at the border to Iran (Figure 1). It includes the districts of Masally, Lenkoran, Astara and Lerik. The total area covers 3960 km² (Statistical Committee Azerbaijan 2009).

Figure 1: Location of Azerbaijan and the study site. Note the dark green band in northern Iran and adjacent southern Azerbaijan indicating the distribution of the Caspian forest. Clearings of the nearby Caspian lowland are light green (image source: NASA 2004). -4-

Study site


The altitude ranges from -24 m a.s.l. (metres above sea level) at the Caspian coast in the East and 2492 m a.s.l. of the Kiumiurkei mountain at the Iranian border in the West (Skvorov 1976a, Skvorov 1976b). The study site contains several protected areas. The three largest are the Hirkan National Park, which was established in 2004 and enlarged in April 2008 to 40358 ha (MENR 2010), the Gizilaghaj state nature reserve at the Caspian coast, which covers 88360 ha and is an important area for wintering and breeding waterbirds (Patrikeev 2004, Sultanov 2008, MENR 2010) and the 15000 ha sized Zuvand state nature sanctuary (MENR 2010).

2.2 Climate The study site can be subdivided into two differing climatic regions. The eastern slopes of the Talish mountains down to the coastal lowland are characterised by a warm-temperate and humid climate. The annual rainfalls exceed 1000 mm per year and peak in February and October. The summer temperatures are warm and the winters are mild (Figure 2). Winter means rarely drop below zero (Mammadov et al. 2007, Mühr 2007).

Figure 2: Climate diagram of Lenkoran after Mühr (2007).

Figure 3: Climate diagram of Ardebil 50 km south of Zuvand upland based on data of IRIMO (2010).

-5-


Study site

The Zuvand area is the second climatic region. Due to mountain ridges, which shield humid air masses coming from the Caspian sea, the Zuvand upland is much drier and the climatic situation is similar to the adjacent Iranian semi-desert. It has an annual mean temperature of 4-8°C. The winters are cold and occasionally the temperature drops below -20°C. The summers show a distinctive drought from June to October and the annual rainfalls are between 200-400 mm per year (Mammadov et al. 2007, cf. Figure 5 in Knapp 2005). No meteorological data was available from the Zuvand region and thus the data of Ardebil (Iran) with a similar climate was taken depicted in Figure 3.

2.3 Landscape types The Talish mountains are subdivided into six landscape or habitat types (Figure 4). The forest belt contains five degradation stages. Each type has its own characteristics:

Caspian lowland From the shore of the Caspian sea to the foothills of the Talish mountains stretches the Caspian lowland, a long but rather narrow stripe including the cities Masally, Lenkoran and Astara (Figure 1, Figure 4). Once it was covered with subtropical broad-leaf forests and impassable wetlands (Radde 1886a). To fight malaria and to gain land for agriculture, forests had been cleared and a widespread system of drainage channels indicate the large range of former wetlands (Patrikeev 2004). Nowadays, the lowland is largely covered with fields, pastures and human settlements (Figure 5, Figure 6). The last remnants of the former lowland forest can be found in a 91 ha sized part of the Hirkan National Park called ‘Moscow forest’ (MENR 2010). The only further forest-like structures in the lowland are afforestations of Quercus spec., orchards or cemeteries. Costal habitats like lagoons, shores or large reedbeds are excluded from this study.

-6-

Study site


Figure 4: Schematic of the vertical and horizontal distribution of the landscape types based on Grossheim (1926), Knapp (2005) and recent Google Earth (2010) satellite images.

-7-


Figure 5: The Caspian lowland is characterised by agriculture and settlements. Sirabil, 21.05.2008.

Study site

Figure 6: Drainage channel. Sirabil, 21.05.2008.

Forest belt and the degradation stages The deciduous broadleaf forest of the Talish mountains ranges from the sea level up to 1800 m a.s.l. (MENR 2004) but the upper tree line is mostly reduced to lower altitudes due to human activities. The number of tree species is high (ca. 90) and includes many endemic or tertiary relict species like Gleditsia caspica, Parrotia persica, Quercus castaneifolia, Albizzia julibrissin, Buxus hyrcana, Ruscus hyrcana or Acer hyrcanum (MENR 2004, Hajiyev 2006). Details about forest types and tree species composition of the Caspian forest are given in Knapp (2005). Scharnweber et al. (2007) identified and described six degradations stages of the Caspian forest in the Talish mountains, on which this study base:

Natural forest stage This stage shows no signs of human activities. It is generally restricted to higher altitudes above 500 m a.s.l., but occurs also elsewhere along steep slopes or apart from human settlements and roads (Figure 7). The natural forest stage can be treated as a primeval forest comparable to the Białowieża forest in Poland. These relicts can be distinguished from other temperate deciduous or mixed forests by their large heights, multistorey profile of stands, a diverse tree community and large amounts of dead wood (Wesołowski 2007). I found this stage near the village Siov and the abandoned village Armudy within the Hirkan National Park.

-8-

Study site


Slightly disturbed forest stage The slightly disturbed forest stage is similar to the natural forest stage. The human impact is low and often not obvious. Thus, this stage was in a few cases difficult to distinguish from the natural forest stage. Hints for a slightly disturbed forest stage are paths, single snags, partly missing dead wood and evidences for grazing like excrements or traces of domestic animals or a reduced understorey. According to Scharnweber et al. (2007), are there no differences to the natural forest stage concerning tree species composition and stand structure. Old-growth trees are a common aspect.

Intermediate disturbed forest stage Logging, lopping and grazing are clearly visible in this stage. This stage can be found near villages (2-4 km) or along roads. Loose cattle or grazing flocks of sheep can regularly be seen in the forest. The tree layer is still good developed, but with a less diverse stand structure and some gaps in the canopy. The intermediate disturbed forest is more suffused with light. This stage was common and widespread, especially along the main roads from Lenkoran to Lerik and from Masally to Iardimli.

Park-like forest stage This degradation stage is characterised by a high logging and grazing activity, which lead to a park-like appearance (Figure 8). Trees are mostly cut down and only single oldgrowth trees are left, which mostly have a chopped appearance due to lopping. Snags are common. The foliage cover of the tree layer is low (ca. 5-10%). The herb layer is short due to grazing. Dense forest fragments can occur. This stage can be found closer to villages (1-2 km), on ridge tops or in plains. Large areas of this stage could be found e.g. near Dashtatuk, Bilarsar, Tankivan or Günesli.

Shrubby woodland stage The shrubby woodland stage is situated close to villages. It is the result of a frequently used former forest. Large trees are almost completely removed and shrubs are dominating -9-


Study site

(Figure 9). Due to lopping and pollarding, trees appear like small bushes, especially Carpinus betulus and Quercus castaneifolia, which rarely exceed heights of more than three metres. Grazing intensity is very high. This stage can also be found at higher altitudes along the upper tree line. Here, the occurrence is probably naturally triggered, as a growth of trees is inhibited by the rougher climate (lower precipitation, lower mean annual temperature, strong winds). I neglected the described treeless bracken fern stage that is dominated by Pteridium aquilinum. It is limited to a few small locations only and could not be adequately sampled. The work of Rietschel (2010) confirmed the classification of Scharnweber et al. (2007).

Figure 7: Widespread natural forest stage of the Caspian forest. Siov, 26.04.2008. - 10 -

Study site


Figure 8: Grazing is common in the park-like forest stage. Lopped trees are visible in the background. Zunguliash, 28.05.2008.

Figure 9: Intense grazing and lopping is responsible for a shrubby appearance of the forest. Logging activity completely removed larger trees. Shinaband, 06.05.2008.

Montane meadow belt Montane meadows are distributed above the forest belt around the smaller towns Lerik, Shingedulan and Iardimli. Due to deforestation, it has an open landscape character with gentle slopes ranging from 700 to 1200 m a.s.l. (Figure 4). Typical are meadows, which are used for hay making (Figure 10). Furthermore, agriculture and pastures are common. - 11 -


Study site

Figure 10: Hay making in the montane meadow belt. Muria, 05.06.2008.

Forest-like structures are rare, but along steep slopes the shrubby woodland stage is partly developed. Otherwise, shrubs are scattered along field paths or roads. Radde (1886a) described the area around Lerik 130 years ago also as an open landscape with agriculture.

Zuvand South of Lerik, the villages Mistan, Kialvaz, Allar, Gosmalian and Shonadzohla comprise the Zuvand upland (Figure 1). Here, three landscape types can be found:

Montane semi-desert Gentle slopes barely covered with thorny cushion-forming tragacanthic vegetation (e.g. Tragacantum, Acantholimon) feature the montane semi-desert (Atamov et al. 2006, Hajiyev 2006, Michael Succow Stiftung 2009) (Figure 11, Figure 12). Tree growth is prevented due to low annual precipitations and long summer droughts (Figure 3). - 12 -

Study site


However, a stand of Juniper excelsa in a runnel near Diwagatsh may indicate the distribution of a former Irano-Turanian steppe-forest which can be found in the adjacent Iran (Zohary 1973). The altitudinal distribution of the montane semi-desert ranges from 1400 to 2500 m a.s.l. (Figure 4, Figure 1). Humans use the semi-desert mainly as pastures. At a few locations, I observed transformations into cultivated land.

Figure 11: Montane semi-desert of the Zuvand. Divagach, 08.05.2008.

Figure 12: Thorny cushion-forming tragacanthic vegetation. Mistan, 19.05.2008.

Riparian forest Embedded in the barren montane semi-desert along small rivulets, open park-like riparian forests of willow species (Salix alba, S. purpurea, S. caprea) and Populus nigra can be found (Michael Succow Stiftung 2009). Riparian forests range from 1200 to 1800 m a.s.l. and are located for example around the villages Gosmalian, Mistan, Kialvaz and Govari. They are rather narrow stripes and restricted to water containing valley bottoms (Figure 13). The inhabitants of the villages along the forest created a complex irrigation system to water their lush and flower-rich meadows, fields and gardens (Figure 14). Large terraces have been built and are widely used as orchards or almond plantations. The vegetation structure of this type is similar to the open appearance of the park-like stage within the forest belt.

- 13 -


Figure 13: Like oases meander riparian forests through the montane semi-desert. Kialakhan, 01.06.2008.

Study site

Figure 14: Lush meadows and planted willow trees are a common aspect of the riparian forest. Shonadzhola, 04.05.2008.

Rocky habitats I found rocky outcrops at altitudes ranging from 1200-2500 m a.s.l., especially around the villages Mistan, Kialakhan, Digia, Pirasora, Bizeir and Khozavi. Steep, inaccessible cliffs and stony terrain is typical (Figure 15, Figure 16). At lower, more humid altitudes ranging from 1200 to 1600 m a.s.l., this type it is often mixed with dense shrubs or pastures. In drier, higher regions of the Zuvand, rocky habitats are often encircled by montane semidesert and include cushion-forming plant species.

Figure 15: Rocky habitats with cushion-forming plant Figure 16: Rocky habitats above 2000 m a.s.l. Mistan, 17.05.2008. species. Piresora, 30.05.2008.

- 14 -

Methods


3 Methods 3.1 Bird sampling I surveyed birds from 03 April to 08 June in 2008. This short period was necessary to guarantee the highest territorial activity of birds during the breeding season in spring. After finishing the survey on 08 June 2008, further excursions took place until 15 July 2008 to investigate more places in the Talish region. I visited the different landscape types equally during spring to avoid date-biased results, but the higher altitudes (montane semi-desert, rocky habitats) were studied later (Figure 17).

Figure 17: Data distribution per date. Each grey point represents a day at which transects were surveyed according to the landscape types.

I chose the line transect method, which was the most suitable method for this bird survey. This method is a compromise between point count method and mapping-census. The main advantage of the line transect method is the good ratio between less time effort and gained data (Flade 1994, Bibby et al. 1995, Südbeck et al. 2005). To accomplish this large area (3960 km²) in a short time span (03 April - 08 June), line transect method

- 15 -


Methods

would give the best results. Südbeck et al. (2005) recommend to survey each transect four times during the breeding season. This recommendation was refused and each transect was surveyed only once. This modification was necessary to achieve a widespread overview of the large Talish region. Furthermore, recovering an already visited transect would be difficult in remote regions without paths or the reliable accuracy of the GPS device (Garmin GPS 60), especially in dense forests. Methodological problems of transects without repeated counts are discussed in Pëterhofs & Priednieks (1989) and Hilden & Järvinen (1989). Main concern according to Hilden & Järvinen (1989) was that only 50% of the breeding pairs are detectable during one count. This problem was reduce as I walked transects slowly (ca. 1-1.5 km/hour) recommended in Südbeck et al. (2005), but faster in open landscapes, for example in montane semi-deserts where bird are not abundant and easier to recognise. A too slow pace could increase double countings of birds and was therefore avoided. Nevertheless, like most bird survey methods, this method is biased against quiet, secretive and nocturnal species (Bibby et al. 1995). It was not necessary to consider different detectability among species because I do not use statistical comparisons among species. A quiet and well-camouflaged treecreeper (Certhia) is always more difficult to recognise than a loud singing thrush (Turdus). An interspecific comparison would be inappropriate, but intraspecific comparisons of a single thrush species and their abundance of different habitats are legitimately. Surveying was done in the morning and evening when the diurnal bird activity peaked. No counts took place under poor weather conditions like rain or strong winds (Flade 1994, Bibby et al. 1995, Südbeck et al. 2005). Each study site and transect route was selected following the determination of landscape types (cf. chapter 2.3.). I avoided human settlings. The transect line was as straight as possible, which was at a few locations difficult to manage because of difficult topography, rivers or dense, thorny shrubs.

- 16 -

Methods


3.2 Habitat sampling I took every 500 m a waypoint to characterise each transect line. This point describes a 500 m segment of the transect line in detail. Each waypoint consist of date and time, landscape type, geographic coordinates (WGS 84), height above sea level, slope exposition, slope steepness, relief (plain, valley bottom, lower slope, middle slope, upper slope, ridge top) and a characterisation of vegetation structure including average height and cover of each vegetation layer (herb, shrub, lower tree and upper tree layer). For describing a complete transect line, I used the data of the waypoints to determine transect length, mean altitude, mean exposition, landscape type composition and vegetation structure, including mean height and cover of the vegetation layers. Along each transect line all birds and their activity (e.g. singing, calling, feeding, flying etc.) were continuously noted. I took the coordinates of every bird within the transects, which consisted of transect metre and distance of the bird to the transect line.

3.3 Data analysis I compiled the data with Microsoft Access 2003 and Microsoft Excel 2003. Processing the data and statistical calculations were performed using the statistical software package R version 2.9.2. (R Foundation for Statistical Computing 2009). For compiling maps the geographic information system QGIS was used (Quantum GIS Development Team 2009). Bird taxonomy follows the Clements Checklist to the Birds of the world 6.3.2 (Clements 2008)

3.3.1

Data treatment

The determination of “territories” was less constrained, because every transect was visited only once. Väisänen (1989) categorised the breeding evidences for bird species, which I modified in this study. I treated the following observations as a territory: a single bird (male, female or juvenile), a pair, an occupied nest or a family. Birds breeding in colonies (e.g. Delichon urbica, Apus apus, Merops apiaster) show no territorial behaviour and the

- 17 -


Methods

number of territories was estimated by counting the total number of adults dividing by two or counting the nests if possible (Bibby et al. 1995). Some bird species (e.g. Hirundo rustica, Merops apiaster, Sturnus vulagris) searched for food in the surrounding areas of their breeding colonies, which were not directly touched by the transect line or maybe overlooked. I treated these birds also as territorial because I expect their breeding colonies to be within surveyed landscape type. I treated flying individuals as territories as long as they show territorial behaviour or at least an association to the habitat. Migrating or high flying birds, non-breeding flocks or birds apart from their typical nesting habitats (e.g. Acrocephalus scirpaceus in park-like forest stage or Merops persicus in woody habitats) or apart from their breeding ranges (Lanius isabellinus, Locustella fluviatilis) were listed as no territories. Birds further than 200 metres from the zero transect line are not included in the calculations to guarantee their relation to the landscape types.

3.3.2

Species richness

To determine the relative species richness of each landscape type, I applied a comparison using log10-transformed values of species number and log10-transformed transect length for each landscape type originating from Table 1. The log10-values on both axes are necessary to fit a simple regression line. This linear regression analysis shows not a speciesarea relationship sensu the theory of island biogeography. It reveals a species-sampling effort relationship from which the relative species richness can be derived. However, this analysis lacks generally of quality, as an area-proportionate sampling, which also lead to time-based problems, could not be achieved (Mac Nally & Horrocks 2002) and results should interpreted with caution.

3.3.3

Calculation of relative abundances

For comparison of each bird species and their distribution according to each landscape type (not to each community), I calculated the relative abundances of every bird species using this formula:

A=T/L

A= relative abundance, T= number of territories, L= total transect length per landscape type (km) - 18 -

Methods


To assure comparability of the relative abundances of species to each landscape type the calculation was modified. According to the variable migratory behaviour of each bird species, are long-distance migrants not included in the early transects of this study, as they arrive later in spring. Using the total transect length per landscape type would underestimate their relative abundance. Hence, for all bird species the date of its first observation was used to sum the length of all kilometres per landscape (L) starting with this day. Therefore, the length used for calculation for long-distance migrants is shorter than for residents.

3.3.4

Nesting guilds

At the guild level, more coherent in the responses to forest degradation can be expected, because species belong to the same guild have similar life history traits, dispersal ability, and spatial distribution and may show similar responses to disturbance (Pearman 2002). To analyse, which guilds are affected by forest degradation, I partitioned them into four nesting guilds (ground nesters, shrub nesters, canopy nesters, cavity nesters) basing on literature review (Glutz von Blotzheim & Bauer 1991, Glutz von Blotzheim & Bauer 1994, Flade 1994, Glutz von Blotzheim & Bauer 1998, Urquhart, & Bowley 2002, Alström & Mild 2003, Patrikeev 2004, Andretzke et al. 2005, Kirwan et al. 2008) (Annex 2). If a bird species belongs to more than one guild, I used the most characteristical. Cuculus canorus had to be excluded from this analysis, because it is unclear which nesting guilds he parasitises most in the Talish mountains.

3.3.5

Statistical analysis

To classify breeding bird communities, I applied a hierarchical, agglomerative cluster analysis using a dissimilarity matrix with Bray–Curtis measure (Bray & Curtis 1957, Leyer & Wesche 2007) based on species abundance for each transect (territories per km) and Ward’s minimum variance method as linking procedure (Ward 1963). As the Ward’s minimum variance method is not compatible to dissimilarity matrix using semi-metric Bray-Curtis measure (Legendre & Legendre 1998, McCune & Grace 2002, Leyer & - 19 -


Methods

Wesche 2007), I took the square root of the dissimilarity matrix. The abundance data have not been transformed and contains rare species. Basing on the output of the cluster analysis, I then created a bird community table, where the communities are arranged following an altitudinal gradient. Within the forest belt, the order followed the intensity of degradation from natural forest stage to the shrubby woodland stage. The socio-ecological species groups were clustered by transposing the dissimilarity matrix to obtain a rough overview, and I finally arranged them manually to their best fit with the help of my expert knowledge and further observations besides the surveying time. I also used NMDS (non-metric multidimensional scaling) to plot ordinations using the function “metaMDS” of the package “vegan” of the statistical software package R. In these NMDS are the transects arranged in scatter-plots regarding to their species composition and abundance (territories/km per transect) to illustrate similarities and dissimilarities between the transect and the gradients which influences their composition (Leyer & Wesche 2007). NMDS base on Bray–Curtis measure (Bray & Curtis 1957) without the square root of the dissimilarity matrix. I then used the stress-value to deter-mine the number of dimensions for the two NMDS (McCune and Grace 2002). For identifying associations of species to communities or combinations of communities, the R package “indicspecies” (version 1.0) the function “multipatt” was applied using 1000 permutations. I rejected the indicator species analysis by Dufrene & Legendre (1997) as it identifies species indicating only one community but not a combination of communities.

In the following text, the code of the significance level is ‘*’ = p-value < 0.05, ‘**’ = pvalue < 0.01 and ‘***’ = p-value < 0.001.

- 20 -

Results


4 Results I surveyed 178 km (kilometre) consisting of 94 transects (Figure 18). The length of transects differed between 200 and 4500 metres. The average length was 1894 metres (n=94). I excluded three transects from the cluster analysis and NMDS analysis, as their length was below 500 m, which do not represent breeding bird communities adequately.

4.1 Species number and species richness I found 10104 individuals of 197 bird species (Annex 1). I expect about 147 species of them to breed in the Talish region (Table 1). (Further breeding birds of wetlands and costal habitats are not included.) From the faunistic aspect, the rediscovery of breeding Shikras (Accipiter badius) was a surprising highlight, as no confirmed breeding records existed since 1933. This is the only breeding site within the Western Palaearctic (Heiss & Gauger 2009). Further breeding species are new to that region or rediscovered after more than a century like Prunella ocularis, Rhodopechys sanguinea, Irania gutturalis or Bucanetes githagineus (Annex 1). Table 1: Totals of surveyed transect length and totals of found breeding bird species per landscape types. Note that bird species can occur in more than one landscape type. Landscape type

Transect length [km]

Species number

Caspian lowland

23.3

54

Natural forest stage

8.6

32

Slightly disturbed forest stage

12.1

40

Intermediate disturbed forest stage

26.1

46

Park-like stage

11.9

41

Shrubby woodland stage

20.4

59

Montane meadow belt

20.6

48

Riparian forest

15.7

44

Montane semi-desert

18.1

44

Rocky habitats

21.2

68

Total

178

147

- 21 -

Figure 18: Overview of the study site. Coloured lines illustrate the surveyed transects. The numbers are the transect numbers.


- 22 -

Results

Results


However, several species listed in Agaeva (1979) or Patrikeev (2004) could not be confirmed in this study like Poecile hyrcana, Hippolais languida, Emberiza buchanani, Gypaetus barbatus, Apus affinus, Halcyon smyrnensis, Picus canus(!), Phoenicurus erythrogaster, Carpospiza brachydactyla or Tichodroma muraria. The most species-rich landscape type are rocky habitats (Figure 19). Rather poor in species are the less disturbed forest degradation stages (natural, slightly and intermediate disturbed forest stages). According to this analysis, montane semi-desert and montane meadow belt are also regarded as relatively poor in species. Interestingly, the heavily degraded forest stages (park-like forest stage and shrubby woodland) show the highest relative species richness.

Figure 19: Species-sampling effort relationship illustrated by the landscape types (Casp = Caspian lowland, Nat = Natural forest stage, Sli = Slightly disturbed forest stage, Int = Intermediate disturbed forest stage, Park = Park-like forest stage, Shr = Shrubby woodland stage, Mead = Montane meadow belt, Rip = Riparian forest, Mon = Montane semi-desert, Rock = Rocky habitats) and a fitted regression line (red line) indicating the same amount of relative speciesrichness. Landscape types above the regression line are rich in species and below are poor in species.

- 23 -


Results

4.2 Breeding bird communities

Figure 20: The nine breeding bird communities of the Talish mountains represented by clusters. The cluster analysis bases on species abundance of each transect using Bray-Curtis measure (Bray & Curtis 1957) and Ward’s method (Ward 1963) as linking procedure. Cluster 6 consists of two merged clusters.

The cluster analysis identified ten breeding bird communities (Figure 20). I merged two clusters (cluster 6) by a local reduction of the cut level, because in field both had a shrubby appearance. Thus, I found nine breeding bird communities in the Talish mountains. The frequency table (Table 2) and the bird community table (Annex 3) show these 9 breeding bird communities and 27 socio-ecological species groups. Species of each socioecological group reflect the same ecological demands and are similar distributed to the communities. Due to the determination of clusters (communities) by breeding birds, some transects are not within their assumed landscape type, for example transect 6 in community 5 (Annex 3). This inaccuracy was neglected, because identifying habitats by plants, as Schwarnweber et al. (2007) did and which I used for landscape type determination, and birds are different approaches.

- 24 -

Results


Table 2: Frequency table of the breeding bird communities. This table is a simplification of the bird community table (Annex 3). Theoretically, a comparison of frequencies of transects with different lengths is not allowed due to species-area relationships. Therefore, the frequency values may give a slightly different impression than the bird community table does (for details see Annex 3). Community Number of transects Transect length [m] Mean altitude [m a.s.l.] Landscape composition: Caspian lowland [%] Natural forest stage [%] Sligthly disturbed forest [%] Intermediate disturbed forest [%] Park-like stage [%] Shrubby woodland stage [%] Montane meadow belt [%] Riparian forest [%] Montane semi-desert [%] Rocky habitats [%] Vegetation structure: Upper tree layer height [m] Upper tree layer cover [%] Lower tree layer height [m] Lower tree layer cover [%] Shrub layer height [m] Shrub layer cover [%] Herb layer height [m] Herb layer cover [%] Carduelis carduelis Cuculus canorus Parus major Hippolais pallida Acrocephalus schoenobaenus Circus aeruginosus Calandrella rufescens Motacilla flava Cercotrichas galactotes Streptopelia decaocto Merops persicus Passer montanus Remiz pendulinus Acrocephalus arundinaceus Francolinus francolinus Acrocephalus scirpaceus Ixobrychus minutus Sylvia mystacea Alcedo atthis Glareola pratincola Hirundo rustica Coracias garrulus Coturnix coturnix Melanocorypha calandra Emberiza melanocephala Motacilla alba Passer domesticus Merops apiaster Luscinia megarhynchos Lanius collurio Certhia familiaris Dryocopus martius

1 9 2589±1033 -4±24 . 100 . .

2 10 2160±1309 506±246 . . 41 46

3 13 1662±833 443±243 . . . 10

4 10 1980±1209 204±123 . . . .

5 11 1636±609 1060±290 . . . .

6 12 2042±1336 1045±485 . . . .

7 5 2740±864 1482±144 . . . .

8 10 1980±756 1685±209 . . . .

9 11 1355±610 1846±233 . . . .

. . . . . . . . 4.9±3.0 4.3±4.3 1.7±1.7 1.7±1.7 1.5±0.7 6.7±5.8 0.3±0.1 79.5±8.2 67 67 56 67 67 56 33 44 22 22 22 33 22 67 11 11 11 11 67 11 100 56 11 11 22 78 100 56 100 44 . .

13 . . . . . . . 25.1±2.1 42.3±18.4 13.5±4.9 22.2±12.6 1.8±1.2 8.2±3.8 0.2±0.1 39.2±19.6 10 70 80 . . . . . . . . . . . . . . . 10 . . . . . . . . . 20 10 30 10

56 23 11 . . . . . 20.1±2.5 20.2±10.9 9.6±2.5 21.3±14.0 1.3±0.6 16.1±11.3 0.1±0.0 30.2±12.4 50 43 86 . . . . . . . . . . . . . . . 7 . . . . . . 7 . . . . . .

38 35 27 . . . . . 19.2±4.2 14.1±9.3 9.5±4.2 24.8±21.8 1.6±0.4 14.2±9.5 0.2±0.0 20.7±11.6 90 70 100 . . . . . . . . . . . . . . . 10 . . . . . . 10 10 . 100 80 . .

18 . 54 . 11 . 17 . 10.7±7.2 8.8±8.6 6.3±2.8 13.2±10.2 2.0±0.8 38.6±22.3 0.1±0.0 37.4±22.2 36 36 100 . . . . . . . . . . . . . . . . . . . . . 18 27 9 . 45 27 . .

. . 12 82 . 6 . . 1.9±0.7 1.1±0.8 1.0±0.7 3.2±4.1 0.2±0.2 59.5±17.8 38 15 31 . . . . . . . . . . 8 . . . . . . 38 8 23 23 38 54 31 . 38 46 . .

. . . . 100 . . . 15.3±2.0 6.1±2.0 6.6±2.4 6.8±2.2 1.9±0.3 9.6±5.9 0.2±0.0 61.3±7.1 100 20 100 . . . . . . . . . . . . . . . . . . . . . . 100 60 60 80 80 . .

. . . . . 82 18 . 0.5±0.4 0.9±0.9 0.1±0.0 13.8±10.0 40 40 . . . . . . . . . . . . . . . . . . 10 . . . . 10 . 40 10 30 . .

. . . . . 13 87 . 3.0±0.0 0.2±0.0 0.9±0.4 5.3±4.6 0.1±0.0 24.0±12.8 45 73 18 . . . . . . . . . . . . . . . . . . . . . 9 9 . . 18 73 . .

- 25 -

Results


Columba oenas Pyrrhula pyrrhula Ficedula semitorquata Carduelis spinus Coccothraustes coccothraustes Sitta europaea Ficedula parva Hippolais icterina Dendrocopos minor Garrulus glandarius Periparus ater Erithacus rubecula Aegithalos caudatus Turdus philomelos Dendrocopos major Phoenicurus phoenicurus Picus viridis Prunella modularis Motacilla cinerea Cinclus cinclus Phylloscopus nitidus Troglodytes troglodytes Sylvia atricapilla Fringilla coelebs Turdus merula Turdus viscivorus Jynx torquilla Muscicapa striata Carduelis chloris Cyanistes caeruleus Corvus cornix Streptopelia turtur Oriolus oriolus Pica pica Lanius minor Sylvia communis Falco tinnunculus Upupa epops Emberiza calandra Dendrocopos syriacus Cettia cetti Falco subbuteo Acrocephalus palustris Sturnus vulgaris Carduelis cannabina Emberiza cia Lullula arborea Alectoris chukar Petronia petronia Oenanthe oenanthe Alauda arvensis Anthus campestris Melanocorypha bimaculata Eremophila alpestris Emberiza hortulana Sitta neumayer Aquila chrysaetos Buteo rufinus Oenanthe finschii Oenanthe isabellina Sylvia curruca Carpodacus erythrinus

. . . . . . . . . . . . . . . . . . . . 11 . . 11 . . . 22 22 33 100 33 33 89 33 33 11 67 89 22 11 11 11 78 . . . . . . . . . . . . . . . . . 11

40 50 50 60 70 100 100 50 . 20 100 100 70 100 100 10 70 20 20 . 60 100 100 100 100 10 10 60 70 100 . . . . . . . . . . . . 10 . 10 . . . . . . . . . . . . . . . . 20

29 14 21 29 93 79 50 14 29 21 93 79 50 93 86 21 50 7 36 7 7 79 64 93 93 7 21 21 79 93 14 7 . . . . . 7 . . . . . . . . . . . 7 . . . . . . . . . . . 7

. . . 20 60 80 60 90 10 . 60 50 60 40 90 10 40 . 30 10 40 40 80 100 100 10 30 70 70 60 70 10 80 . . . . 10 . . . . . . . . 30 . . . . . . . . . . . . . . .

- 26 -

18 . . . 9 18 9 . . 36 45 45 45 36 27 36 36 9 9 . 27 27 36 73 100 27 9 . 36 64 36 27 9 18 9 36 9 45 64 9 . . . . 45 27 73 . 9 9 9 9 . . . 18 . . . . 18 36

. . . . . . . . . 8 . . . . . 8 . . . . 8 . . 15 38 8 15 . . . 62 23 31 69 23 31 . 62 85 . . . 8 54 31 15 62 15 8 8 38 15 8 15 8 . . . . . . .

. . . . . . 20 . . . . . . . . 80 80 20 20 20 40 20 60 100 100 . . 60 . 100 80 . 40 100 80 80 60 100 100 80 100 40 40 100 . 40 60 40 60 . . . . . . 20 . . . . 60 60

. . . . . . . . . . . . . . . . . . . . . . . 10 . . . . . . 10 . 10 20 10 . 20 40 30 . . . . 50 80 70 100 100 70 30 30 90 30 90 20 60 10 10 60 60 . .

. . . . . . . . . . . . . . . . . . 9 . . 9 18 9 45 . . . . . . . 9 18 9 27 55 18 27 . . . . . 100 91 64 64 91 45 45 36 9 55 27 82 9 9 36 18 55 18

Results


Columba palumbus Apus apus Delichon urbicum Ptyonoprogne rupestris Tachymarptis melba Monticola saxatilis Monticola solitarius Oenanthe hisp. melanoleuca Serinus pusillus Turdus torquatus Bucanetes githagineus Saxicola torquatus Phoenicurus ochruros Rhodopechys sanguineus Irania gutturalis

. 67 33 . . . . . . . . . . . .

. . . . . . . . . . . . . . .

7 7 . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

18 . . . . . . . . . . 9 . . .

. 31 8 . . . . . . . . 8 . . .

60 . . . . . . . . . . . . . .

. . 20 . . 30 . 10 20 10 10 . 10 . .

9 27 27 18 27 100 18 18 64 45 18 18 91 9 9

Excluded species: No grouping: Saxicola rubetra Lanius senator Anthus spinoletta Ardea cinerea Prunella collaris Anthus trivialis Corvus corax Sylvia nisoria Excluded owls and raptors: Buteo buteo Aquila pomarina Accipiter badius Circaetus gallicus Neophron percnopterus Accipiter nisus Strix aluco Otus scops Breeding status unclear: Phylloscopus sindianus Crex crex

. . . 11 . 11 . . . . . . . . . . . . . . . .

. . 10 . . . . . . . . 20 . . . . . 10 . . . .

. . . . . . . . 7 . . 29 7 . . . . . . . . .

. . . . . . . . . . . 10 . 10 . . . . 10 . . .

. . . . . . . 36 9 9 . 9 . . 9 . . . 9 . 27 .

. . 8 8 . . . 23 23 . . 23 8 . . . 8 . . . . 8

. . . . . . . 20 . . . . . . . . . . 20 . . .

. . 10 . 10 . 10 . 20 . . . . . . . . . . . . .

. . . 9 . . . 9 18 9 . 18 . . . 9 . . . . . .

Annex 4 provides the detailed results of the indicator species analysis. Significant species associated to only one group based on the indicator species analysis, are marked with the significance code within the following description of the nine breeding bird communities. The landscape type composition is given in the header data of Table 2:

Community 1 (Caspian lowland) I found this community exclusively in the Caspian lowland. Several species are restricted to this community like typical steppe species (Hippolais pallida***, Cercotrichas galactotes**, Merops persicus*, Francolinus francolinus, Sylvia mystacea) or species of marshes and wetlands (Acrocephalus schoenobaenus***, Acrocephalus scirpaceus, Acrocephalus arundinaceus***, Circus aeruginosus***, Remiz pendulinus*, Glareola - 27 -


Results

pratincola, Motacilla flava***, Ixobrychus minutus). Most of them occur along draining channels. Species that generally prefer open landscapes can also be found here and in higher altitudes above the forest belt, mainly in the montane meadow belt (Emberiza melanocephala, Melanocorypha calandra, Emberiza calandra, Coturnix coturnix) or riparian forests (Dendrocopos syriacus, Cettia cetti, Falco subbuteo, Merops apiaster), where they belong to community 7 and 6 respectively. As the lowland is widely covered with human settlings, village related birds occur in high abundances (Passer domesticus***, Hirundo rustica***, Sturnus vulgaris, Apus apus, Pica pica, Motacilla alba).

Community 2 (Natural forest) Community 2 occurs along transects conducted in natural forest or slightly disturbed forests (both together: 87%). Certhia familiaris** and Dryocopus martius appear exclusively in this community. The social-ecological group containing Ficedula semitorquata**, Pyrrhula pyrrhula*** and Columba oenas are restricted to community 2 and 3, but only the first two species indicate community 2, whereas Columba oenas is an indicator species for the combination of community 2 and 3 (Annex 4). Typical forest species like Periparus ater, Sitta europaea, Dendrocopos major, Erithacus rubecula or Troglodytes troglodytes reach their highest abundances in this community (Annex 3), but occur also elsewhere within the forest belt in other breeding bird communities. Species of open woodland are scarce (Luscinia megarhynchos, Lanius collurio, Lullula arborea).

Community 3 (Intermediate disturbed forest) Due to the cluster analysis community 3 consists mainly of transects conducted in intermediate disturbed forests (56%) according to the classification of Scharnweber et al. (2007). It also presents bird species that are typical for forest habitats. The species composition of community 3 is similar to community 2 but open woodland species occur rarely (Streptopelia turtur, Corvis cornix). Lullula arborea is missing.

- 28 -

Results


Community 4 (Strongly disturbed forest) The proportion of transects made in park-like forests is relatively high (35%). Nevertheless, also transects of intermediate disturbed forests (38%) and shrubby woodland (27%) had been classified due to the cluster analysis to community 4. Community 4 contains typical forest species (Turdus merula, Fringilla coelebs, Sitta europaea, Periparus ater), but with lower relative abundances. In comparison with community 2 and 3, species of open habitats are more abundant (Lullula arborea, Corvus cornix, Streptopelia turtur, Oriolus oriolus, Luscinia megarhynchos, Lanius collurio). No indicator species are within community 3 and 4.

Community 5 (Shrubby woodland) The species composition and the species abundances of this community indicate this community as shrubby woodland. 54% of shrubby woodlands are included in this community. In comparison with community 2, 3 and 4, several forest species are missing in community 5 (Pyrrhula pyrrhula, Ficedula semitorquata, Carduelis spinus) or are less abundant (e.g. Sitta europaea, Ficedula parva, Dendrocopos major, Dendrocopos minor, Coccothraustes coccothraustes, Turdus philomelos, Fringilla coelebs, Periparus ater, Erithacus rubecula). Open woodland species (Corvus cornix, Streptopelia turtur) or generally open landscapes preferring species (Emberiza melanocephala, Emberiza calandra, Emberiza cia, Carduelis cannabina, Sylvia communis) are present and can be common. This degradation stage was mostly close to human settlements (Noack 2007, Scharnweber et al. 2007). Therefore, bird species that are also related to human settlements were present in small numbers (Upupa epops, Motacilla alba, Pica pica). The only indicator species of this community is Phylloscopus sindianus**, but it is unclear whether this species is a breeding bird in that region (Patrikeev 2004).

Community 6 (Montane meadows) I found this community mostly within the montane meadow belt (82%). It is closely related to community 1 (Figure 20). A few species are only distributed in these two com- 29 -


Results

munities such as Coracias garrulus, Coturnix coturnix* and Melanocorypha calandra. Community 6 lacks of typical forest species (Periparus ater, Ficedula parva, Sitta europaea) but includes species of open woodland (Corvus cornix, Streptopelia turtur, Lullula arborea). Emberiza melanocephala**, Lanius minor, Emberiza calandra, Carduelis cannabina prefer shrubs in an open landscape within their breeding territories and therefore also occur in community 6. As the montane meadow belt is dominated by agriculture, which includes many villages, also species assigned to human settlements can be found here (Hirundo rustica, Passer domesticus, Apus apus, Delichon urbicum).

Community 7 (Riparian forest) The breeding bird community 7 reflects the birds of the riparian forest in 1200 to 1800 m a.s.l. (Table 2). Typical species of the forest-interior do not occur in the riparian forests (Sitta europaea, Ficedula parva, Dendrocopos major, Periparus ater, Erithacus rubecula), except some widespread forest species (Fringilla coelebs, Turdus merula, Cyanistes caeruleus). As the riparian forest is situated within the montane semi-desert, some species, like Alectoris chukar and Petronia petronia, extend their breeding range into this community, but playing a minor role. A socio-ecological group that contains Dendrocopos syriacus***, Cettia cetti***, Falco subbuteo** and Acrocephalus palustris**, is restricted to community 7 and 1, but their species indicate the riparian forest community. Furthermore, species of open landscapes (Lanius minor**, Sylvia communis***, Emberiza calandra, Falco tinnunculus) or human settlements (Passer domesticus, Upupa epops) breed here. Sturnus vulgaris reaches here highest breeding densities. Further indicator species are Phoenicurus phoenicurus*** and Columba palumbus**.

Community 8 (Montane semi-desert) The breeding bird communities of community 8 and 9 are different to community 1-7. The species composition changed almost completely and several species occur only in these two communities. Community 8 includes typical montane semi-desert species (Melanocorypha bimaculata, Oenanthe isabellina***, Anthus campestris***), but also - 30 -

Results


species that breed in lower altitudes of community 6 (montane meadow belt), like Oenanthe oenanthe or Alauda arvensis, or in rocky habitats of community 9 (Oenanthe finschii, Eremophila alpestris, Emberiza hortulana). Along rocky outcrops within the montane semi-desert also typical rock associated species occur, but with generally low abundances like Sitta neumayer, Monticola saxatilis and Oenanthe finschii.

Community 9 (Rocky habitats) Rocky outcrops which I found at 1500 – 2300 m a.s.l feature community 9. It contains rock-dependent species, which are exclusively related to community 9 (Ptyonoprogne rupestris, Tachymarptis melba*, Monticola saxatilis***, Monticola solitarius, Oenanthe hispanica melanoleuca, Bucanetes githagineus, Phoenicurus ochruros***, Rhodopechys sanguineus) or exclusively related but depending on shrubs (Serinus pusillus***, Turdus torquatus**, Saxicola torquatus, Irania gutturalis). I observed species of this community also elsewhere, but predominantly in community 8 (Alectoris chukar, Petronia petronia, Emberiza cia, Carduelis cannabina, Lullula arborea, Sitta neumayer, Oenanthe finschii, Oenanthe oenanthe**). The only forest species occurring here are Turdus merula or rarely Troglodytes troglodytes, Sylvia atricapilla and Parus major. This is the only community where I found Apus apus and Delichon urbicum besides human settlings.

Indicator species for a combination of forest communities Additionally, the analysis of indicator species identified species occurring significantly in two or more communities. Especially, indicator species of the forest belt rarely appear significantly in only one community (Annex 4). For community 2 and 3, indicating a natural, slightly and intermediate disturbed forest, significant indicator species are Turdus philomelos***, Troglodytes troglodytes***, Erithacus rubecula*** and Columba oenas**. Species with a broader amplitude are Dendrocopos major***, Sitta europaea***, Periparus ater***, Coccothraustes coccothraustes***, Ficedula parva***, Carduelis chloris*** and Carduelis spinus**. They are significantly related to all forest degradation stages except the shrubby woodland stage (community 2+3+4). In contrast, Dendrocopos minor* is an indicator for breeding - 31 -


Results

bird communities preferring degraded forests (community 3+4). Aegithalos caudatus*** is the only species which is significantly related to all types of the forest belt (community 2+3+4+5), whereas Turdus merula***, Fringilla coelebs***, Sylvia atricapilla*** and Picus viridis*** are additionally indicator species for the riparian forest (community 2+3+4+5+7). This deeper insight into the species-community relationships reveals that several species indicate a large variety of landscape and habitat types. Even a generic species can be an indicator species like Turdus merula and Fringilla coelebs, as the most widespread forest species (Table 2, Annex 3).

4.3 Parameters influencing breeding bird communities The analysis of parameters is important to understand ecological communities. For this purpose, I carried out a NMDS analysis. This ordination is used to describe relationships between species composition patterns and the influencing parameters onto a two dimensional space. Figure 21 illustrates the nine breeding bird communities along two ordination axis. The altitudinal gradient is the strongest site parameter, which is underlined by an approach on species-level (Figure 24). The altitudinal gradient ranges from bottom left to top right with community 8 and 9 at the end of this vector representing the highest altitudes. NMDS1 axis follows the vegetation structure of tree and shrub layer. From left to right a gradient concerning naturalness of forest appearance can be derived. Open landscape communities are on the right and forested landscapes are on the left. A gradient following NMDS2 axis reflects the herb layer. Community 1 (Caspian lowland) contains the best-developed herb layer, followed by community 6 (montane meadows).

- 32 -

Results


Figure 21: Ordination of breeding bird communities using a NMDS generated from abundances of all species of all transects after 20 random starts of a two-dimensional solution (final stress = 16.06, rmse = 0.002, max residual = 0.0072). Highly significant site parameters with a p-value < 0.001 are plotted as red vectors and other are black (for details, see Annex 5). Lengths of vectors indicate the significance of each parameter. Transect numbers are given in symbols. Parameter abbreviations: ALTI = altitude, DATE = date, EXPO = slope exposition (after Parker 1982), SLO = slope steepness (after Parker 1982), HH = Herb layer height, HC = Herb layer cover, SH = Shrub layer height, SC = Shrub layer cover, LTH = Lower tree layer height, LTC = Lower tree layer cover, UTH = Upper tree layer height, UTC = Upper tree layer cover. Landscape abbreviations: Casp = Caspian lowland, Nat = Natural forest, Int = Intermediate disturbed forest, Stro = Strongly disturbed forest, Shr = Shrubby woodland, Mead = Montane meadows, Rip = Riparian forest, Semi = Montane semi-desert, Rock = Rocky habitats.

- 33 -


Results

Figure 22: Ordination of breeding bird communities within the forest degradation stages using a NMDS generated from abundances of all species of transects of community 2, 3, 4 and 5 after 100 random starts of a two-dimensional solution (final stress = 15.83, rmse = 0.00015, max residuals = 0.00054 after 81 tries). Highly significant site parameters with a p-value < 0.001 are plotted as red vectors and other are black (for details, see Annex 6). Lengths of vectors indicate the significance of each parameter. Transect numbers are given in symbols. Parameter abbreviations: ALTI = altitude, DATE = date, EXPO = slope exposition (after Parker 1982), SLO = slope steepness (after Parker 1982), HH = Herb layer height, HC = Herb layer cover, SH = Shrub layer height, SC = Shrub layer cover, LTH = Lower tree layer height, LTC = Lower tree layer cover, UTH = Upper tree layer height, UTC = Upper tree layer cover. Landscape abbreviations: Nat = Natural forest, Int = Intermediate disturbed forest, Stro = Strongly disturbed forest, Shr = Shrubby woodland.

- 34 -

Results


To get a deeper insight into the forest degradation stages and their explanatory variables, only transects of community 2, 3, 4 and 5 have been calculated again in a further NMDS (Figure 22). These transects were conducted within the forest belt (Figure 4). I excluded transect 5 and 6 as they belong to other landscape types, whereas transect 73 was treated as belonging to community 5, as it was securely made in the shrubby woodland stage and only due to the cluster analysis considered as belonging to community 6. NMDS1 axis reveals a degradation gradient (Figure 22). Vectors of the tree layer are pointing towards transects performed in natural and slightly disturbed forests and vectors of the shrub layer into the opposite direction towards transects of the shrubby woodland communities. The altitudinal gradient also plays an important role, but is less significant compared to the landscape-scaled approach in Figure 21. Each bird species and therefore each bird community has a different ecological demand. The differences of each community according to their preferred vegetation structures are given in Figure 23 and the header data of Table 2. Communities 2-5 and 7 represent forested landscapes. The height and cover of the upper tree layer decreases continuously from community 2 to 5, whereas the cover of the shrub layer increases. This gradient also appears in Figure 22. The breeding bird communities clearly show a response to the vegetation parameters resulting in different values. Interestingly, these results are similar to Scharnweber et al. (2007), which base on vegetation analysis. For comparison, communities besides the forest degradation stages are also included. Community 1, 6, 8 and 9 are different. They present open landscapes with a missing or reduced tree layer.

4.4 Relative abundances of bird species Due to the heterogeneity of several transects according to their landscape composition, especially within the forest belt, where transects consisted often of several degradation stages, the calculation of relative abundances could not be performed basing on bird

- 35 -


Results

Figure 23: Boxplots of height and cover of each vegetation layer for every breeding bird community.

- 36 -

Results


Figure 24: Boxplots of the altitudinal distribution of breeding bird species basing on the number of territories (n≥5). - 37 -


Results

Figure 25: Relative abundances values of the 15 most common bird species per landscape type (forest degradation stages).

- 38 -

Results


Figure 26: Relative abundances values of the 15 most common bird species per landscape type (outside the forest belt).

- 39 -


Results

communities. Thus the relative abundance values base on the landscape type determination (chapter 2.3). This was in field more precisely and therefore used for further calculations, especially to compare the species response in different forest degradation stages. Figure 25 shows the relative abundances of the most common bird species within the forest degradation stages (forest belt). The most abundant species in the lesser disturbed forests are Periparus ater and Fringilla coelebs. However, a comparison between bird species is theoretically not allowed, because of the different detectability among species. This comparison is also biased towards eye-catching and loud-voiced birds. A shift regarding species composition and relative abundance takes place in the parklike forest stage (Figure 25). The abundance of typical forest species decline with ongoing forest degradation (e.g. Periparus ater, Fringilla coelebs, Troglodytes troglodytes, Erithacus rubecula), whereas the relative abundance of open woodland species increases. Most increasing species are related to shrubs of open landscapes (Luscinia megarhynchos, Emberiza melanocephala, Sylvia communis, Lanius collurio).

4.5 Response to forest degradation on species-level I found 32 breeding bird species within the natural forest stage. The obtained species number is the result of a total transect of 8.6 km (Table 1). 28 species are regarded as ‘forest species’. I then converted the relative abundance values of each species per forest degradation stage into percent values to compare the impact of human activities within the forest belt. The natural forest stage serves as a gauge that gives information about a pristine forest without any human activities. I used the converted abundance values of the natural forest stage as reference values constituting 100%.

Negatively affected or strongly negatively affected Figure 27 shows a clear negative response of forest species to forest degradation. Most of these species prefer the forest interior. A decline of about 50% according to an unaffected population is visible in the intermediate disturbed forest stage. Some species, like

- 40 -

Results


Fringilla coelebs, are also common in the park-like forest stage, but show a rapid decline in the shrubby woodland stage. Only 6% or less of the bird population of strongly affected species remain in the shrubby woodland stage. 65% of the 28 forest species show a negative or strongly negative response to forest degradation.

Figure 27: Negatively and strongly negatively response of selected forest bird species to forest degradation. Values are scaled relative to the natural forest stage (reference stage) constituting 100%.

- 41 -


Results

Positively affected Generally, open landscapes and shrub formations preferring species are positively affected by forest degradation, e.g. Luscinia megarhynchos. Only a few species, treated as forest species or at least ‘edge species’, show a positive trend (Figure 28). These species have low abundances in natural forests and are common in degraded forests resulting in a vast increase of a few species, e.g. Carduelis chloris. The treatment of some of those species as forest species might be discussible. Only 21% of the 28 forest species show a positive response to forest degradation. Most of them show a peak in the park-like forest stage, but decrease in the shrubby woodland stage (Figure 28).

Figure 28: Positive response of selected forest bird species to forest degradation. Values are scaled relative to the natural forest stage (reference stage) constituting 100%.

4.6 Response to forest degradation on guild-level The guild concept is needed because it can help to identify the habitat characteristics that determine the structure of bird communities and because it would be possible to understand processes organising communities (Casenave et al. 2008). In total, the natural and slightly disturbed forest stages reach their highest relative abundance values within the forest belt. The shrubby woodland stage holds the lowest values (Figure 29). - 42 -

Results


The dominating guild within all degradation stages are cavity breeders. This guild includes typical cavity breeders, like Dendrocopos major or Cyanistes caeruleus, and semicavity breeders, like Certhia familiaris. Cavity breeders reach their highest relative abundances in natural and slightly disturbed forests with 25.2 and 24.8 territories per km, respectively. In contrast, only 6.5 territories per km of cavity breeders occur in the shrubby woodland stage. Their proportion of about 40% is in all degradation stages more or less equal, besides the shrubby woodland stage. Here, only 17% belong to the cavitybreeding guild owing to the lack of cavities in this stage. Canopy breeders are common in all degradation stages with 12.0 to 16.2 territories per km, except for the shrubby woodland stage with only 7.1 territories per km. The highest percentages can be found in intermediate disturbed forests and park with 32% each. Shrub breeders dominate the shrubby woodland stage. Here, the highest proportion (37%) and relative abundance (14.2 territories per km) can be found. Birds that breed on the ground show a decline in relative abundance and proportion concerning a degradation gradient from high values in natural forest stage and low values in park-like stage. On the contrary, the values of relative abundance and proportion rise again in the shrubby woodland stage.

Figure 29: Response of nesting guilds to forest degradation. (Nat = Natural forest stage, Sli = Slightly disturbed forest stage, Int = Intermediate disturbed forest stage, Park = Park-like forest stage, Shr = Shrubby woodland stage). - 43 -


Discussion

5 Discussion 5.1 Breeding bird communities The cluster analysis identified 9 breeding bird communities (Figure 20). Every community is representing more or less one of the landscape types. Especially transects conducted outside the forest belt show an apparent overlapping (header data of Table 2). Within the forest belt, a differentiation of forest degradation stages by breeding bird communities is imprecise owing to the problems of the survey method. For the cluster and NMDS analysis each transect was treated as a single record, despite the differing transect length and differing landscape type compositions. In a few cases, transects consisted of a mixture of several degradation stages. For example, a 2000 m long transect contains 1500 m parklike forest and 500 m shrubby woodland stage. The cluster algorithm may classifies this transect to a community representing the park-like stage including 500 m of shrubby woodland. A wrong determination of the landscape types in field could also cause blurred results concerning the landscape composition of each community. Within the forest degradation stages, it was sometimes difficult to separate the stages from each other. For example, the natural forest stage was difficult to distinguish from the slightly disturbed forest stage, as both are very similar. A further important point, regarding the landscape composition of each breeding bird community, is the early date of surveying of a few transects. The species composition of early transects (beginning of April) lacks most longdistance migrants. These species contribute also to the classification of transects to a community by the cluster analysis and if they are missing, the cluster analysis classifies the transects only by the resident species. This problem could not be adequately solved by beginning the survey when long-distance migrants completely arrived the Talish mountains (end of May/beginning of June). Several residents already fed chicks and show a reduced territorial behaviour, which leads to the opposite effect of ‘missing’ residents, when the territorial activity of migrants peaks. The only way to solve this problem would be a repeated surveying of each transect, but this was refused (chapter 2.6.). However, on this landscape-scaled approach, open habitats fit to the open land breeding bird communities. Within the forest belt with its degradation stages, it can only roughly be as- 44 -

Discussion


signed (header data of Table 2), but a seperation following a man-made degradation gradient is clearly visible (Figure 21, 22). Each community differs according to their species composition and relative abundance values (Table 2, Annex 3). Interestingly, on this landscape-scaled approach, which separated for example Caspian lowland from rocky habitats by breeding birds only, a differentiation between forest communities (community 2, 3, 4 and 5) is evident. This is remarkable as the clusters of the cluster analysis were cut at the same level, representing an equal degree of similarity (Figure 20). However, the differences in bird species composition between communities of the forest belt are low (Table 2, Annex 3). Within the forest communities, indicator species occurred only in community 2 (natural forest): Pyrrhula pyrrhula***, Certhia familiaris**, Ficedula semitorquata** (Annex 4). These few indicator species are in other forest communities missing, which underlines the strong similarity between these communities. As the cluster analysis separates clusters by species and their abundances, hence the forest communities are mainly differentiated by the relative abundance values. The differentiation by species inventory plays a minor role in the forest communities. Several forest bird species, like Periparus ater, Troglodytes troglodytes, Fringilla coelebs or Sitta europaea, reach their highest relative abundances in community 2 representing natural forests (Table 2, Annex 3). In contrast, Tomiałojć & Wesołowski (2004) found remarkably low abundance values in the primeval forest of Białowieża (Poland) compared to data from man-transformed places (e.g. fragmented forests of Western Europe). Species of community 2 are also present in communities representing strong forest degradation, for example community 5 containing 54% of the shrubby woodland stage. These differences of relative abundance values are responsible for the separation of bird communities by the cluster analysis within the forest belt (degradation stages), despite the low differences in species composition. However, to give community 2, 3, 4 and 5 the status of a breeding bird community might be discussable due to the similar species inventory. The separation of forest breeding bird communities by species only is difficult or almost impossible. Here, the main problem occurs for an practical use of this from the vegetation ecology derived concept. Other problems are the migratory life history traits of birds. To evaluate the

- 45 -


Discussion

quality of a landscape or a habitat by the development of a breeding bird community is under the circumstance of a not sessile living bird inaccurate. The reason for the presence or absence of a species might not be a loss of habitat quality on the breeding site, it can also be hunting pressure on migration routes or quality loss on the wintering sites. Plants are therefore more practical for the evaluation of habitat quality as they spend their whole life on one location. Furthermore, the determination by their abundance values is not practical in field, as it depends on a large amount of surveying time, which avoids a quick identification of each forest community. Nevertheless, human activity created different breeding bird communities within the forest belt, which are statistically verified in a landscape-scaled approach.

5.2 Parameters influencing bird communities The vegetation structure is the key parameter influencing the composition of breeding bird communities in the Talish mountains (Figure 21-23). Heterogeneous vertical and horizontal vegetation structures are an important factor influencing breeding bird communities (abundance and diversity), because it offers a large number of ecological niches for a broad variety of bird species. (Mac Arthur & Mac Arthur 1961, Mac Arthur 1968, Erdelen 1984, Tews et al. 2004, Kati & Sekercioglu 2006). Plant species diversity of (temperate) forests is less important and has nothing to do with bird species diversity (Mac Arthur 1961). Thus, I expect the high diversity of woody plant species in the Talish mountains has no effect on the species diversity of bird communities, although it was not tested in this study. For the formation of each breeding bird community is the development of the tree layers, as the main characteristic of the vertical vegetation structure with different tree heights and cover values, most important (header data of Table 2, Figure 23). In Figure 21 and Annex 5 are information given about the importance of each vegetation parameter. Every vegetation parameter is highly significant in this large-scaled approach. At a smaller-scaled approach, regarding forest communities (2, 3, 4 and 5) only, height and cover of the upper tree layer, height of the lower tree layer and cover of the shrub

layer

are

among

all

vegetation

parameters

- 46 -

these

with

the

highest

Discussion


Figure 30: Breeding bird response to forest degradation. Summarised schematic of the forest degradation stages.

significance (p-value < 0.0001) (Figure 22, Annex 6). High values of tree height and tree cover indicate the occurrence of forests, which is obviously crucial to the bird species composition of each community. If trees, as main contributor to the vertical vegetation structure, are missing, completely different bird communities are developed. The resulting open land communities consist mainly of open land species like Alauda arvensis, Hirundo rustica or Emberiza melanocephala. This separation from open land and forest communities is in agreement with other studies (e.g. Kati & Sekercioglu 2006). A removal of forests would cause a species turnover by extirpating of most forest bird species and colonisation of open land species (Figure 21, 30). This is also observable in tropical rain forests after intensive logging or clear-cutting (e.g. Thiollay 1999, Patten et al. 2010). A possible recolonisation of forest species could never bring back the whole assemblage due to the extinction of some bird species. The British avifauna for example, lacks of several forest bird species, for which no geographical or biological reasons exist, probably caused by a widespread removal of forests centauries ago (Fuller et al. 2007a, - 47 -


Discussion

Wesołowski 2007). Nevertheless, a complete removal of the Caspian forest in Azerbaijan is unlikely, but the current forest degradation changes the vertical vegetation structure, which affects most forest birds negatively (Figure 27). The differences in the vertical vegetation structure lead to the differing appearance of breeding bird communities, which roughly represent each degradation stage. As already mentioned, each forest community is more differentiated by the species abundance then the species composition. High trees and high foliage cover are typical for undisturbed primeval forests (Wesolowski 2007). Trees provide cavities for hole nesters and their crowns offer canopy breeders nesting opportunities. The older and larger the trees are the more cavities are available and the higher is the abundance of cavity breeders in the natural forest stage compared to other degradation stages (Enoksson et al. 1995, Poulsen 2002) (Figure 29). The high proportion of cavity breeders is typical for European oldgrowth forests (e.g. Wesołowski & Tomiałojć 1997, Saniga & Saniga 2004, Korṅan 2009). Old and large trees have also voluminous crowns and can therefore harbour more canopy breeders (Figure 29). In addition, also ground breeders achieve high values in natural forests. A good developed, diverse and undisturbed herb and shrub layer enables several species to breed on the ground. In contrast, within the park-like forest stage is the herb layer short due to grazing of cattle and sheep. Here, breeding possibilities for ground breeders are low and disturbances are frequent. Therefore, abundances of ground breeders are lower in the park-like forest stage (Figure 29). Compared to the natural forest stage with high and different aged trees and a multistorey structure, the vertical vegetation structure of each forest degradation stage is less diverse and complex. Furthermore, natural forests are structurally enriched by the ubiquitous presents of freshly fallen trees with their rootwads, fallen decaying logs, snags, various tree holes, temporary pools etc. (Tomiałojć & Wesołowski 2004). The diverse and complex structures in natural forests form a unique system of ecological niches that can support high species diversity, as it reduces the intraspecific and interspecific competition for resources, like nesting sites or food (Tomiałojć & Wesołowski 2004). There is no evidence for an interspecific competition caused by a shortage of nesting holes and food limitations in primeval forests (Wesołowski & Tomiałojć 1997, Wesołowski 2003,

- 48 -

Discussion


Wesołowski 2007). I regarded this assumption as the main reason why abundances of forest birds reach their highest values in natural or slightly disturbed forest (community 2). The only nesting guild, which is positively affected by forest degradation, are shrub breeders (Figure 29). I expect a shortage of food and nesting sites in the degraded forests due to the observed lower relative abundances of forest species. Furthermore, edge effects of fragmented forests increase the risk of predation resulting in lower abundances within degraded forests. Forest fragmentation is the replacement of large areas of native forest by other ecosystems leaving isolated forest patches, with deleterious consequences for most of the native forest biota (Saunders et al. 1991 cited in Murcia 1995). The maintenance of large, structurally complex patches of native vegetation is particularly important in landscapes where many species are area-sensitive and confined to native vegetation, and where locations outside these patches are entirely uninhabitable by many native species (Fischer et al. 2006). Human activities are the main driver for the development of the breeding bird communities within the forest belt of the Talish mountains, because they change the vertical vegetation structure. A removal of large trees due to logging, a destruction of the understorey and herb layer caused by fuel wood gathering or silvopastures reduce the breeding possibilities and success of each forest bird, which generally lead to lower abundances and a clear decline (Figure 25, 27, 29 and Annex 3). Additionally, the vertical vegetation structure is also triggered by climatic conditions. Tree growth is prevented or reduced in regions with a too dry and too cold climate. In the Talish mountains no trees occurred in the dry montane-semi desert. In higher altitudes above 1800 m a.s.l. (natural tree line), their growth is reduced due to the rougher climate. Here, the forest becomes the appearance of natural shrubby woodlands. Furthermore, fires, windthrows or landslides can locally completely deplete forests resulting in a changed vertical vegetation structure. Figure 21 reveals that the altitude is an important parameter influencing the breeding bird communities. This parameter was neglected, as the vegetation structure depends on the altitude, but birds depend on vegetation structure and not on the altitude.

- 49 -


Discussion

5.3 Species and species richness A high richness and diversity of species is for nature and bird conservation a main concern (e.g. Ramírez-Albores 2007). As the calculation of diversity indices was improper due to methodological reasons, only the species number of each landscape type was taken into account. The most species-rich landscapes are rocky habitats of the Zuvand upland (Figure 19). This landscapes type consists of a variety of different habitats. Rocks are a common aspect and enable several rock-dependent bird species to breed here. In addition, this type is often mixed with shrubs, pastures and montane semi-desert (tragacanthic vegetation), which offer further species breeding opportunities. Horizontal heterogeneity and a diverse landscape are important factors for high species richness (Pino et al. 2000, Sekercioglu 2002, Kati & Sekercioglu 2006). However, this fact should be interpreted with caution under nature conservation issues, because within a large extended primeval forest, horizontal heterogeneity can only be achieved due to fragmentation by grazing and logging. Unsurprisingly, the most species-rich forest degradation stages are the shrubby woodland stage and the park-like forest stage (Figure 19, Figure 30). Both are under intense human utilisation pressure resulting in a diverse horizontal heterogeneity with shrub formations, pastures, open and dense forests. This horizontal diversity enables breeding opportunities for a broad variety of open land and forest-dependent bird species in a relative small area (Table 1, Table 2). This consideration is in agreement with other studies (e.g. Herold 2005). Ugalde-Lezama (2010) made the ‘theory of intermediate disturbance’ (Connell 1978) responsible for his observations. In contrast, less disturbed forests like natural or slightly disturbed forests are relatively poor in species (Figure 19, Figure 30). Their horizontal structure is more uniform with only a few natural forest clearings due to windthrows, fires or landslides. Open land species are rare or missing, which reduces the total amount of species in these types. Tomiałojć & Wesołowski (2004) stated that primeval forests are generally rich in species, which might be the case for the Białowieża primeval forest, but the situation of the Caspian forest is different. This forest is isolated and surrounded by deserts and treeless steppes. No connection to the boreal forest zone exists and therefore many boreal species, which contribute to the high species number in Białowieża forest, are missing.

- 50 -

Discussion


Especially some non-migratory species of the genera owls, tits, warbler, woodpeckers and grouses do not occur. A general high richness in species is therefore not a feature of primeval forests of the temperate zone, it is a feature of the transition zone of deciduous broadleaf forest to boreal forest. Nevertheless, despite the lower total species richness of natural forests, compared to the degradation stages, the diversity of forest-interior species is higher (Figure 30, Annex 7). Inspired by the island biogeography theory, Askins et al. (1987) stated that the number of species and the density of forest birds is similar in forests of different sizes, but the composition of the bird community shows consistent differences. Interior-edge birds are equally common in large and small forests, while both the density and species richness of forest-interior birds are higher in larger forests. In short, not only the worsening of the habitat quality is responsible for the absence of some forest species, also the fragmentation and isolation of forest remnants lead to local extinctions, depending on varying spatial scales, of forest-interior bird species. A change in species composition per se can also trigger extinction cascades. The loss of individual species is particularly likely to trigger extinction cascades when ‘keystone’ or ‘strongly interacting’ species are involved because they exert a disproportionate effect on ecosystem function relative to their abundance (Chapin et al. 2000, Fischer & Lindenmayer 2007). This means that each species in a (pristine) forest ecosystem fulfills its ecological functions. For example, the loss of large predators due to hunting or habitat loss may lead to increasing numbers of herbivores, which leads to a decreasing tree recruitment. The absence of single bird species can also cause a reduced seed dispersal of different tree species (Lindenmayer et al. 2000). Insectivorous bird species control pest outbreaks, which reduces plant damages (Sekercioglu et al. 20004). Species diversity influences the resilience and resistance of ecosystems to environmental change (Chapin et al. 2000). However, species richness alone should not be used to guide conservation decisions. Other factors such as endemism, rarity, habitat specialisation, complementary of sites and biological organisation also need to be taken into account (Walther & Martin 2001). Bird endemism of the Caspian forest exists mainly on the subspecies-level. Almost all of them

- 51 -


Discussion

are more or less forest-dependent birds, like Phasianus colchicus talischensis, Dendrocopos minor quadrifasciatus, Dendrocopos major poelzami, Garrulus glandarius hyrcanus, Poecile hyrcana hyrcana, Periparus ater gaddi, Sitta europaea rubiginosa, Regulus regulus hyrcanus, Erithacus rubecula hyrcanus (Agaeva 1979, Patrikeev 2004, Clements 2008). Poecile hyrcana hyrcana was recently splitted from Poecile lugubris and is probably the only endemic species, but prefers degraded forests (Loskot 1978). Forestdependent endemic species or subspecies are directly affected by deforestation. To protect them is important to preserve the global biodiversity. Diversity at all organisational levels, ranging from genetic diversity within populations to the diversity of ecosystems in landscapes, contributes to global biodiversity (Chapin et al. 2000). The protection of the Caspian forest is therefore strongly linked with the protection of birds.

5.4 Conclusion and conservation implications Bird species richness and species abundance patterns of the Caspian forest are triggered by vertical and horizontal vegetation structures (Figure 30). A high heterogeneity of the vertical vegetation structure causes a high abundance of forest species. A high horizontal vegetation structure, caused by forest degradation, reduces the abundance of forest species and enables the breeding of open land species, which increases the species richness. Several forest-dependent species can still exist in strongly degraded park-like forests (Figure 25). The impact of forest degradation to species composition is negligible in slightly or intermediate disturbed forests. Both are able to accommodate a viable population of forest-dependent bird species. Therefore, most important for the conservation of forest species is the prevention of large-scaled natural forests with a diverse vertical heterogeneity. A sustainable and natural forestry with unused areas, also outside the Hirkan National Park, would protect the Caspian forest avifauna. Preservation of a system of small reserves rather than a large reserve of the same total area would probably result in the loss of area-sensitive species and increased rarity for other species of forest-interior birds that are present, but relatively uncommon, in small forests (Askins et al. 1987). Old-growth trees of these natural areas should also be prevented from log-

- 52 -

Discussion


ging. Mature deciduous trees are of great importance to many resident birds (Enoksson et al. 1995). Moreover, primeval forests are ‘windows into the past’ through which one would gain insights into the ecology of pristine plant and animal communities (Wesołowski 2003). The data collected here may, besides enhancing our understanding of the fundamental problems of community organisation, also serve as a gauge for the bird community studies made in transformed woodland habitats (Wesołowski & Tomiałojć 1997). Nowadays, when the natural habitats are exposed to a permanent negative impact resulting from human activities all over the world, the knowledge, gained here, is especially needed for forest biodiversity conservation, sustainable use and ecological restoration. Clear-cutting and further fragmentation should be avoided also outside the Hirkan National Park. In particular, smaller remnant fragments were highly vulnerable to ongoing disturbances as they were accessible for logging and clearance (Echeverria 2005 cited in Echeverria et al. 2007). Additionally, extensive infrastructure developments of remote areas, especially by the construction of roads, could cause further degradation and a further loss of native biodiversity. Roads offer easy accessibility and forest degradation is generally high along roadsides (Edenius 1996, Thiollay 1999). In addition, a touristic use of the protected areas of the Caspian forest should be carefully managed, as it may accompany with ecological degradation (Liu et al. 2001, Krüger 2005). Inaccessibility is the best protection for primeval forests (Joppa et al. 2008). To preserve such large-scaled forested areas is crucial to the whole flora and fauna. Birds, as a generally well studied taxonomic group (Flade 1994), may stand for the broad variety of further taxonomic groups like mammals, amphibians, reptiles or insects. Their reaction is completely unknown in that region, but their response can be derived from the observed response of birds. Paralleling the bird results, I expect further local extinctions of single species among these taxonomic groups accompanied by a loss of biodiversity with ongoing degradation and fragmentation. As long as the use of the forest is sustainable and in balance with ecological processes, an extinction of species and a loss of biodiversity is unlikely.

- 53 -


Discussion

5.5 Scope and limitations Some limitation of the present work should be considered. The sampling design was made for a landscape-scaled approach. Other methods, like point count method, might be more suitable for a smaller-scaled approach dealing with forest degradation stages only. Here, a detailed record of the vegetation with more parameters, especially concerning vegetation structure, could deliver better results. Most of the surveyed bird species were small passerines, which generally have smaller territories. They are probably not much affected by forest degradation, as their demand for large-scaled habitat structures is less complex. This study could not implement any suggestions about rare or nocturnal species. Most of them a large sized (e.g. Ciconia nigra, Aquila pomarina, Circaetus gallicus) with a need for large territories within the forest belt. These species are not adequately censused in this study and a further scientific research should also focus on them, for example to proof whether existing protected areas are sufficiently large to maintain viable populations. Monitoring species, which indicate natural or slightly disturbed forests such as Pyrrhula pyrrhula, Ficedula semitorquata and Certhia familiaris, or groups of birds e.g. woodpeckers (Drever & Martin 2010) is an efficient method to monitor the ecological state of forest communities and it gives a more direct insight into bird habitat quality. The paucity of consistent and complete knowledge about species biology or ecological processes, however, leads to believe that adequately identified indicator species can be useful for management, conservation, and restoration of natural and seminatural ecosystems, especially for large-scale projects (Bani et al. 2006).

- 54 -

Summary


6 Summary This study analyses the breeding bird communities and their response to forest degradation in the Talish mountains of Azerbaijan. Primeval deciduous broadleaf forests belong globally to the most threatened ecosystems. Only small, isolated remnants survived in Europe. The Caspian forest is the largest primeval deciduous broadleaf forest and situated in northern Iran and the adjacent Azerbaijan. In Azerbaijan, the Hirkan National Park protects 40358 ha. Outside, the forest is exposed to livestock grazing and logging resulting in forest degradation and fragmentation. I conducted this study in the breeding season from April to June 2008 using the line transect method. I then used the obtained data to link breeding bird communities with environmental parameters. For this purpose I calculated abundance values per transect for cluster analysis and ordination analysis (NMDS). Furthermore, I compared five forest degradation stages (natural forest, slightly disturbed forest, intermediate disturbed forest, park-like forest and shrubby woodland) according to relative abundances of bird species. From a faunistic aspect, the breeding records of Radde’s Accentor (Prunella ocularis), Crimson-winged Finch (Rhodopechys sanguinea), White-throated Robin (Irania gutturalis), Trumpeter Finch (Bucanetes githagineus) and Shikra (Accipiter badius) are of national importance, as they are newly discovered breeding birds or rediscovered after many decades. The cluster analysis revealed nine breeding bird communities. They are arranged mainly along an altitudinal gradient ranging from the Caspian lowland to montane semideserts. Four of the communities are within the forest belt and a result of forest degradation. 65% of the forest-dependent bird species are negatively affected by forest degradation. Several species, like Eurasian Bullfinch (Pyrrhula pyrrhula), Semi-collared Flycatcher (Ficedula semitorquata) and Eurasian Treecreeper (Certhia familiaris), are expected to become extinct with ongoing degradation. Nine endemic subspecies are threatened.

- 55 -


Summary

The differences between natural forests and only slightly disturbed forests are low regarding species composition and relative abundance. With further degradation, the relative abundance and the number of forest-dependent species decreases, but the total species number increases. The strong vertical vegetation heterogeneity of natural forests with high and old-growth trees, multistorey profiles and a large amount of dead wood cause a high abundance of forest-dependent species. Natural forests are horizontally homogenous landscapes resulting in a lower total species number. In contrast, degraded forests have a reduced vertical heterogeneity, which reduces the abundance of forest species. The strong horizontal heterogeneity of degraded forests including shrub formations, different-sized forest fragments and open lands (pastures) enables a broad variety of bird species breeding opportunities resulting in high species richness. A slightly use of the Caspian forest does not have a serious impact on the forest avifauna. Hence, a natural and sustainable forestry, conserving a rich vertical vegetation structure, would protect the forest avifauna and prevent a loss of global biodiversity.

7 Zusammenfassung Primäre Laubwälder gehören global zu den am stärksten bedrohten Ökosystemen. In Europa finden sich nur noch kleine, isolierte Reste primärer Laubwälder. Das weltweit größte zusammenhänge Primärwaldgebiet, der Kaspische Wald, befindet sich im Norden des Irans und im angrenzenden Aserbaidschan. In Aserbaidschan werden 40358 ha im Hirkan National Park geschützt. Außerhalb des Nationalparks führen Viehwirtschaft und intensive Holznutzung zu einer Degradierung und Fragmentierung des Waldes. Die vorliegende Arbeit untersucht die Brutvogelgemeinschaften des Talisch Gebirges in Aserbaidschan. Zudem wurde deren Verhalten unter dem Aspekt der fortschreitenden Walddegradierung betrachtet. Im Frühjahr 2008 führte ich die Geländeuntersuchung mittels der Transektmethode durch. Anhand der gewonnenen Daten wurden die errechneten Brutvogelgemeinschaften mit Umweltparametern korreliert. Dazu wurde die berechnete Abundanz pro Transekt für

- 56 -

Summary


die Clusteranalyse und Ordination (NMDS) benutzt. Des Weiteren verglich ich fünf Degradierungsstadien des Waldes (Naturwald, leicht gestörter Wald, mittel gestörter Wald, parkartiger Wald and Gebüschwald). Unter faunistischen Gesichtspunkten sind Brutnachweise bzw. Brutzeitfeststellungen von Steinbraunelle (Prunella ocularis), Rotflügelgimpel (Rhodopechys sanguinea), Weißkehlsänger (Irania gutturalis), Wüstengimpel (Bucanetes githagineus) und Schikra (Accipiter badius) von nationaler Bedeutung, da diese Arten neu entdeckt bzw. nach vielen Jahrzehnten wiederentdeckt wurden. Die Clusteranalyse ergab neun Brutvogelgemeinschaften. Diese sind hauptsächlich entlang eines Höhengradienten angeordnet, welcher vom Kaspischen Tiefland bis in die Gebirgshalbwüste verläuft. Vier der Gemeinschaften befinden sich im Waldgürtel und sind das Ergebnis der Walddegradierung. 65% der Waldvogelarten sind von der Walddegradierung negativ betroffen. Mehrere Arten, wie z.B. Gimpel (Pyrrhula pyrrhula), Halbringschnäpper (Ficedula semitorquata) und Waldbaumläufer (Certhia familiaris), werden mit fortschreitender Degradierung im Talisch Gebirge aussterben. Neun endemische Unterarten sind bedroht. Die Unterschiede zwischen Natur(Primär-)wäldern und nur leicht gestörten Wäldern sind gering bezogen auf das Arteninventar und deren relativen Abundanz. Mit zunehmender Degradierung verringert sich die Abundanz und Artenzahl von Waldvogelarten bei einer Zunahme der Gesamtartenzahl. Die starke Heterogenität der vertikalen Bestandsstruktur von Naturwäldern, mit hohen und alten Bäumen, vielen Vegetationsschichten und einem hohen Anteil an Totholz, verursachen eine hohe Abundanz von Waldvögeln. Aber ihre räumliche (horizontale) Homogenität verursacht eine vergleichsweise geringe Gesamtartenzahl. Im Gegensatz zu Naturwäldern haben degradierte Wälder eine verringerte Heterogenität der vertikalen Bestandsstruktur, die mit einer Reduzierung der Abundanz von Waldvögeln einhergeht. Die starke räumliche Heterogenität von degradierten Wäldern (z.B. durch Gebüsche, lichte und geschlossene Waldfragmente, Viehweiden), ermöglicht einer großen Anzahl von Vogelarten Brutmöglichkeiten, die eine hohe Gesamtartenzahl bewirken.

- 57 -


Summary

Eine geringfügige Nutzung der Kaspischen Wälder hat keinen schwerwiegenden Einfluss auf die Waldavifauna. Daher würde eine naturnahe und nachhaltige Forstwirtschaft, welche eine abwechslungsreiche vertikale Bestandsstruktur über weite Flächen bewahrt, dem Schutz der Waldavifauna dienen und damit eine globale Verarmung an Biodiversität verhindern.

8 Acknowledgements I am grateful to Prof. Dr. Michael Succow and Dr. Martin Flade for the supervision of my diploma thesis. The Ministry of Ecology and Natural Resources of Azerbaijan Republic kindly granted the necessary access to the Hirkan National Park. I thank Kai Gauger for the indispensable support during the fieldwork. I also wish to acknowledge Jan Peper for useful critics and the brilliant introduction to multivariate statistics. Furthermore, I thank Nigar Agaeva, Jonathan Etzold, Benjamin Herold, André Jankowski, Dr. Vladimir M. Loskot, Prof. Dr. Michael Manthey, Dr. Hartmut Müller, Jan Peters, Tobias Scharnweber and Sebastian Schmidt for providing literature and useful comments. The people of the Talish mountains, especially Novrus Hüseynov and his family, Akif Aliyev and Babakhan, also contribute with their support and overwhelming hospitality to this work. This study could not be accomplished without the financial support of the DAAD (German Academic Exchange Service) and the help of the Michael Succow Foundation.

- 58 -

References


9 References Abdollahpour, M., & Assadi Atui, A. R. (2005). Entwicklung der geregelten Forstwirtschaft im Iran. In: Nosrati, K., Marvie Mohadjer, R., Bode, W. and Knapp, H.D., Schutz der Biologischen Vielfalt und integriertes Management der Kaspischen Wälder (Nordiran). Bundesamt für Naturschutz. Bonn, 99-120. Agaeva, T. A. (1979). [Avifauna of the Talish mountains]. Dissertation. Azerb. Univ. Baku. Agaeva, T. A. (1972). [Short notes on some birds of Talish Mountains] (report 1). Proc. Azerb. Univ., boil. ser., no. 2, 54-57. Agaeva, T. A. & Mustafaev, G. T.(1973). [Short notes on some birds of Talish Mountains] (report 2). Proc. Azerb. Univ., boil. ser., no. 3, 51-54. Alström, P. & Mild, K. (2003). Pipits & Wagtails of Europe, Asia and North America. Christopher Helm. Andretzke, H., Schikore, T. & Schröder, K. (2005). Artsteckbriefe. In: Südbeck, P., Andretzke, H., Fischer, S., Gedeon, K., Schikore, T. Schröder, K. & Sudfeldt, C. (eds.): Methodenstandards zur Erfassung der Brutvögel Deutschlands. Radolfzell. 135-695. Askins, R. A., Philbrick, M. J. & Sugeno, D. S. (1987). Relationship Between the Regional Abundance of forest and the Composition of Forest Bird Communities. Biological Conservation 39: 129-152. Atamov, V. V., Cabbarov, M. & Gurbanov, E. (2006). The Pytosociological Characteristics of Ecosystems of Mountain of Talysh Region of Azerbaijan. Asian Journal of Plant Sciences 5 (5): 889-904. Bani, L., Massimino, D., Bottoni, L. & Massa, R. (2006). A Multiscale Method for Selecting Indicator Species and Priority Conservation Areas: a Case Study for Broadleaved Forests in Lombardy, Italy. Conservation Biology, vol. 20, no. 2: 512526. Bibby, C. J., N. D. Burgess & D. A. Hill (1995): Methoden der Feldornithologie. Radebeul, 270 S. Bray, J. R. & Curtis, J. T. (1957). An Ordination of the Upland Forest Communities of Southern Wisconsin. Ecological Monographs, Vol. 27, No. 4, 326-349.

- 59 -


References

Casenave, J. L., Cueto, V. R. & Marone, L. (2008). Seasonal dynamics of guild structure in a bird assemblage of the central Monte desert. Basic and Applied Ecology 9: 7890. Chapin, F. S., Zavaleta, E. S., Eviner, V. T., Naylor, R. L., Vitousek, P. M., Reynolds, H. L., Hooper, D. U., Lavorel, S., Sala, O. E., Hobbie, S. E., Mack, M. C. & Diaz. S. (2000). Consequences of changing biodiversity. Nature, vol. 405: 234-242. Clements, J. F. (2008). Birds of the World: A Checklist – 6 th Edition. Ibis Publishing, Vista, CA. Connell, J. H. (1978). Diversity in Tropical Rain Forests and Coral Reefs. Science, New Series, vol. 199, no. 4335: 1302-1310. Drever, M. C. & Martin, K. (2010). Response of woodpeckers to changes in forest health and harvest: Implications for conservation of avian biodiversity. Forest Ecology and Management 259: 958-966. Dufrene, M. and Legendre, P. (1997). Species assemblages and indicator species: the need for a flexible asymmetrical approach. Ecol. Monogr. 67: 345-366. Echeverria, C., Newton, A. C., Lara, A., Benayas, J. M. R. & Coomes, D. A. (2007). Impacts of forest fragemtation on species composition and forest structure in the temperate landscape of southern Chile. Global Ecol. and Biogeogr. 16: 426-439. Edenius, L. & Elmberg, J. (1996). Landscape level effects of modern forestry on bird communities in North Swedish boreal forests. Landscape Ecology. vol. 11 no. 6: 325-338. Enoksson, B., Angelstam, P. & Larsson, K. (1995). Deciduous forest and resident birds: the problem of fragmentation within a coniferous forest landscape. Landscape Ecology vol. 10 no. 5: 267-275. Erdelen, M. (1984). Bird communities and vegetation structure: 1. Correlations and comparisons of simple and diversity indices. Oecologia 61: 277-284. Fischer, J. Lindenmayer, D. B. & Manning, A. D. (2006). Biodiversity, ecosystem function, and resilience: ten guiding principles for commodity production landscapes. Front Ecol. Environ. 4 (2): 80-86. Fischer, J. & Lindenmayer, D. B. (2007). Landscape modification and habitat fragmentation: a synthesis. Global Ecol. Biogeogr. 16: 265-280. Flade, M. (1994). Die Brutvogelgemeinschaften Mittel- und Norddeutschlands. IHWVerlag, Eching. - 60 -

References


Fuller, R. J., Gaston, K. J. & Quine, C. P. (2007a). Living on the edge: Brithish and Irish woodland birds in a European context. Ibis 149 (Suppl.2): 53-63. Fuller, R. J., Smith, K. W., Grice, P. V., Currie, F. A. & Quine, C. P. (2007b). Habitat change and woodland birds in Britain: implications for management and future research. Ibis 149 (Suppl. 2): 261-268. Glutz von Blotzheim, U. & Bauer, K. (eds.) (1991): Handbuch der Vögel Mitteleuropas, vol. 12, Aula Verlag, Wiesbaden. Glutz von Blotzheim, U. & Bauer, K. (eds.) (1994): Handbuch der Vögel Mitteleuropas, vol. 9, Aula Verlag, Wiesbaden. Glutz von Blotzheim, U. & Bauer, K. (eds.) (1998): Handbuch der Vögel Mitteleuropas, vol. 11, Aula Verlag, Wiesbaden. Google Earth (2010). Satellite Images. http://earth.google.de . Accessed: April 2010. Gregory, G. (2007). Shikra in the Western Palearctic. Birding World 20: 159–161. Grossheim, A. A. (1926). [Flora of the Talish mountains]. Isdanie Narkomzema Az. SSR. Hajiyev, V. (2006). Biodiversity of the flora of the Hirkan Biosphere Park. In: Aliyev, I., Engels, B., Hedden-Dunkhorst, B., Heinrichs, A. K. & Okada, C. (Eds.), Protection and Utilisation of Biological Resources for a Sustainable Development, with Special Reference to Azerbaijan - Proceedings of a NATO CCMS supported Seminar. Bundesamt für Naturschutz. Bonn, 16-19. Heiss, M. & Gauger, K. (2009). The rediscovery of breeding Shikras Accipiter badius in the Western Palaearctic. Sandgrouse 32: 134–137. Herold, B. (2005). The Avian Communities and their Habitat Associations of the Walnut Fruit Forests of Southern Kyrgyzstan. Diploma thesis. Ernst-Moritz-ArndtUniversity. Greifswald. Hilden, O. & Järvinen, A (1989). Efficiency of the line-transect method in mountain birch forest. Ann. Zool. Fennici 26: 185-190. Islamic Republic of Iran Meteorological Organization (IRIMO). Climatic Statistics. http://www.irimo.ir/. Accessed: 15.04.2010. IUCN Red List (2010). IUCN Red List of Threatened Species. http://www.iucnredlist.org/ Accessed: 15.05.2010.

- 61 -


References

Joppa, L. N., Loarie, S. R. & Pimm, S. L. (2008). On the protection of “protected areas”. PNAS, vol. 105, no. 18: 6673-6678. Kati, V. I. & Sekercioglu, C. H. (2006). Diversity, ecological structure, and conservation of the landbird community of Dadia reserve, Greece. Diversity Distrib. 12: 620-629. Kirwan, G. M., Boyla, K. A., Castell, P., Demirci, B., Özen, M., Welch, H. & Marlow, T. (2008). The Birds of Turkey - The Distribution, Taxonomy and Breeding of Turkish Birds. Christopher Helm. Knapp, H. D. (2005). Die globale Bedeutung der kaspischen Wälder. In: Nosrati, K., Marvie Mohadjer, R., Bode, W. and Knapp, H.D., Schutz der Biologischen Vielfalt und integriertes Management der Kaspischen Wälder (Nordiran). Bundesamt für Naturschutz. Bonn: 45-70. Korňan, M. (2009). Structure of the breeding bird assemblage of a primeval alder swamp in the Šúr National Nature Reserve. Biologia 64/1: 165—179, 2009. Krüger, O. (2005). The role of ecotourism in conservation: panacea or Pandora’s box? Biodiversity and Conservation 14: 579-600. Legendre, P. & Legendre, L. (1998). Numerical Ecology – Second English Edition. Elsevier B. V. Leyer, I. & Wesche, K. (2007). Multivariate Statistik in der Ökologie - Eine Einführung. Springer-Verlag Berlin Heidelberg. Lindenmayer, D. B., Margules, C. R. & Botkin, D. B. (2000). Indicators of Biodiversity for Ecologically Sustainable Forest Management. Conservation Biology, vol. 14, no. 4: 941-950. Liu, J., Linderman, M., Ouyang, Z., An, L., Yang, J. & Zhang, H. (2001). Ecological Degradation in Protected Areas: The Case of Wolong Nature Reserve for Giant Pandas. Science 292. 98-101. Loskot, V. M. (1978). [The Hyrkanian Tit, Parus hyrcanus Sarudny et Loundon]. Proc. of Zool. Inst. of Acad. Sci. of the USSR, vol. 76: 46-60. Mac Arthur, R. H (1968). The theory of the niche. In R. C. Lewontin, ed., Population Biology and Evolution: 159-176. Mac Arthur, R. H., & Mac Arthur, J. W. (1961). On Bird Species Diversity. Ecology 42: 594-598.

- 62 -

References


Mammadov, G.S., M.P. Babaev, Z.R. Movsumov, C.M. Ceferova & V.H. Hesenov (2007). Torpaq Atlasi (Soil Atlas, in Azeri and Russian). Baki Kartografiya Fabriki. McCune, B. & Grace, J. B. (2002). Analysis of Ecological Communities. MjM Software Design. Michael Succow Stiftung (2009). Potential Analysis for Further Nature Conservation in Azerbaijan – A Spatial and Political Investment Strategy. Geozon Science Media. Ministry of Ecology and Natural Resources of Azerbaijan Republic (MENR) (2004). Nomination of the "Hirkan forests" of Azerbaijan as UNESCO World Heritage Site. Baku: unpublished manuscript. Ministry of Ecology and Natural Resources of Azerbaijan Republic (MENR) (2010). Biodiversity. http://www.eco.gov.az. Accessed: 30.04.2010. Mohadjer, R. M. (2005). Prinzipien der naturnahen Waldwirtschaft. In: Nosrati, K., Marvie Mohadjer, R., Bode, W. and Knapp, H.D., Schutz der Biologischen Vielfalt und integriertes Management der Kaspischen Wälder (Nordiran). Bundesamt für Naturschutz. Bonn, 121-127. Mühr, B. (2007). Das Klima in Lenkoran. http://www.klimadiagramme.de/Asien/lenkoran.html. Accessed: 15.04.2010. Murakami, M., Hirao, T., Iwamoto, J. & Oguma, H. (2008). Effects of windthrow disturbance on a forest bird community depend on spatial scale. Basic and Applied Ecology 9: 762-770. Murcia, C. (1995), Edge effects in fragmented forests: implication of conservation. Tree vol. 10 no. 2: 58-62. Mustafaev, G. T. & Agaeva, T. A. (1981). [The birds of the Talish Mounains at Azerbaijanian border]. Azerb. gos.107-115. Baku. Mac Nally, R. & Horrocks, G. (2002). Proportionate spatial sampling and equal-time sampling of mobile animals: A dilemma for inferring areal dependence. Austral Ecology 27: 405-415. NASA (2004). The Caucasus-Caspian region (30.09.2004). http://visibleearth.nasa.gov. Accessed: 12.02.2010. Noack, F. & Hidayatov, A. (2007). The Economics of Forest Depletion in Southern Azerbaijan. Archiv für Naturschutz und Landschaftsforschung , 2 (46), 109-131.

- 63 -


References

Parker, A. J. (1982). The topographic relative moisture index: an approach to soilmoisture assessment in mountain terrain. Phys. Geogr. 3(2):160-168. Patrikeev, M. (2004). The Birds of Azerbaijan. Pensoft, Sofia/Moscow. Patten, M. A., de Silva, H. G. & Smith-Patten, B. D. (2010). Long-term changes in the bird community of Palenque, Chiapas, in response to rainforest loss. Biodivers Conserv 19: 21-36. Pearman, P. B. (2002). The scale of community structure: Habitat variation and avian guilds in tropical forest understory. Ecological Monographs, 72, 19–39. Pëterhofs, E. & Priednieks, J. (1989). Problems of applying the line-transect method without repeated counts when the breeding season is long. Ann. Zool. Fennici 26: 181-184. Pino, J., Rodá, F., Ribas, J., Pons, X. (2000). Landscape structure and bird species richness: implications for conservation in rural areas between natural parks. Landscape and Urban Planning 49: 35-48. Poulsen, O. B. (2002). Avian richness and abundance in temperate Danish forests: tree variables important to birds and their conservation. Biodiversity and Conservation 11: 1551-1566. Prilipko, L., (1954). Lesnaja Rastitelnost Aserbaidschana. Baku: Isdatelstwo Akademi Nauk Aserbaidschanskoi SSR. Quantum GIS Development Team (2009). Quantum GIS Geographic Information System. Open Source Geospatial Foundation Project. http://qgis.osgeo.org R Foundation for Statistical Computing (2009). R: A Language and Environment for Statistical Computing. Vienna, Austria. http://www.r-project.org Radde, G. (1884). Ornis Caucasica. Kassel: Theodor Fischer. Radde, G. (1886a). Reisen an der persisch-russischen Grenze -Talysch und seine Bewohner. Leipzig: F.A.Brockhaus. Radde, G. (1886b). Die Flora und Fauna des südwestlichen Caspi-Gebietes – Wissenschaftliche Beträge zu den Reisen an der Persich-Russischen Grenze. Leipzig: F.A.Brockhaus. Ramírez-Albores,J.E.(2007). Bird diversity and conservation of Alto (Southwestern Puebla), Mexico. Rev. Biol. Trop. Vol. 55 (1): 287-300.

- 64 -

Balsas

References


Rietschel, M. (2010). Multitemporale Untersuchungen Hyrkanischer Wälder mittels Change Detection auf der Basis von Landsat-Daten. Diploma thesis. Ernst-MoritzArndt-University. Greifswald. Saffari, N. G. (2005). Landnutzung, ländliche Entwicklung und Gefährdung der Kaspischen Wälder. In: Nosrati, K., Marvie Mohadjer, R., Bode, W. and Knapp, H.D., Schutz der Biologischen Vielfalt und integriertes Management der Kaspischen Wälder (Nordiran). Bundesamt für Naturschutz. Bonn, 89-97. Saniga, M. & Saniga, M. (2004). Influence of forest stand structure on the occurence of bird community in Skalná Alpa National Nature Reserve in the Vel’ká Fatra Mts. (West Carpathians). J. For. Sci. 50: 219-234. Scharnweber, T., Rietschel, M., & Manthey, M. (2007). Degradation stages of the Hyrcanian forests in southern Azerbaijan. Archiv für Naturschutz und Landschaftsforschung , Vol.46 ( Heft 2), 133-156. Seifollahian, M., Ebrahimi Rastaghi, M. & Hedayati, M.A. (2005). Die Bedeutung der Pflanzenarten der Kaspischen Wälder. In: Nosrati, K., Marvie Mohadjer, R., Bode, W. and Knapp, H.D., Schutz der Biologischen Vielfalt und integriertes Management der Kaspischen Wälder (Nordiran). Bundesamt für Naturschutz. Bonn, 71–80. Sekercioglu, C. H. (2002). Effects of forestry practices on vegetation structure and bird community of Kibale National Park, Uganda. Biological Conservation 107: 229240. Sekercioglu, C. H., Daily, G. C., Ehrlich, P. R. (2004). Ecosystem consequences of bird declines. PNAS, vol.101, no.52: 18042-18047. Skvorzov, G. A. (1976a). Topographical Map 1:100 000 J-39-37 Lerik. General Staff of Soviet Military. Skvorzov, G. A. (1976b). Topographical Map 1:100 000 J-39-38 Lenkoran. General Staff of Soviet Military. Südbeck, P., Andretzke, H., Fischer, S., Gedeon, K., Schikore, T. Schröder, K. & Sudfeldt, C. (2005): Methodenstandards zur Erfassung der Brutvögel Deutschlands. Radolfzell. Sultanov, E. (2008). Recent Waterbird Counts in Gyzylagach, the Azerbaijan Republic’s Most Important Ramsar Site, with Comments on Conservation. Podoces, 3(1/2): 3138. Tews, J., Brose,U. ,Grimm, V., Tielbörger, K., Wichmann, M. C., Schwager, M. & Jeltsch, F. (2004). Animal species diversity driven by habitat - 65 -


References

heterogeneity/diversity: the importance of keystone structures. J. Biogeogr. 31: 7992. The State Statistical Committee of the Republic of Azerbaijan (2009). Demographic indicators. http://www.azstat.org. Accessed: 03.03.2010. Thiollay, M-J. (1999). Responses of an avian community to rain forest degradation. Biodiversity and Conservation 8: 513-534. Tomiałojć, L. & Wesołowski, T. (2004). Diversity of the Białowieża Forest avifauna in space and time. J. Ornithol. 145: 81–92. Ugalde-Lezama, S., Alcántara-Carbajal, J. L., Valdez-Hernández, J. I., Ramírez-Valverde, G., Velázquez-Mendoza, J. & Tarángo-Arámbula, L. A. (2010). Richness, abundance and diversity of birds in the temperate forest with different conditions of perturbance. Agrociencia 44: 159-169. Urquhart, E. & Bowley, A. (2002). Stonechats - A Guide to the Genus Saxicola. Christopher Helm. Väisänen, R. A. (1989). Renewal of methodology in the second bird atla of Finland, 19861989. Ann. Zool. Fennici 26: 167-172. Walther, B. A. & Martin, J-L. (2001). Species richness estimation of bird communities: how to control for sampling effort? Ibis 143: 413-419. Ward, J. H. (1963). Hierarchical Grouping to Optimize an Objective Function. American Statistical Ass.58 (301), 236-244. Wesołowski, T. & Tomiałojć, L. (1997). Breeding bird dynamics in a primeval temperate forest: long-term trends in Białowieża National Park (Poland). Ecography 20: 432– 453. Wesołowski, T. (2003). Bird community dynamics in a primaeval forest – is interspecific competition important? Ornis Hungarica 12-13: 51-62. Wesołowski, T. (2007). Primeval conditions – what can we learn from them? Ibis 149 (Suppl. 2), 64–77. Zohary, M. (1973). Geobotanical Foundations of the Middle East. Stuttgart: G.Fischer.

- 66 -

Appendices


10 Appendices Annex 1: Commented list of the observed bird species of the Talish mountains region including breeding status (b = breeding, n = non-breeding), based on own observations and interpretations, and IUCN Red List (2010) category (EN = Endangered, VU = Vulnerable, NT = Near Threatened, Least Concern not mentioned).

No. 1 2 3

Species Cygnus cygnus Cygnus olor Anas querquedula

4

Alectoris chukar

5

Tetraogallus caspius Francolinus francolinus

6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

Status Comment n 1 immature bird on 20 May near Narimanabad. n Up to 85 individuals in wetlands near Liman. n 1 bird near Narimanabad on 20 May Common at rock outcrops in Zuvand upland. 1 nest with 20 eggs b found near Kialakhan on 2 June. b On 9 July a calling bird was heard in rocks near Khozavi.

Only 2 displaying males observed on 5 May near Tazakend (Caspian lowland). Rare. A displaying male heard on 3 April near Aliabad (montane Perdix perdix meadow belt). 2 birds in montane semi-desert near Nalabin on 24 June. Singing males were common in the montane meadow belt and Coturnix coturnix the Caspian lowland. Phasianus No observation. Only a plucking of a female or juvenile found in colchicus shrubs at steep slope near Lerik on 14 April. A colony of 49 breeding pairs in wetlands near Liman on 20 Podiceps cristatus May. 2 chicks also present. Phalacrocorax 13 birds in wetlands near Liman on 20 May. carbo Phalacrocorax Some observations around Liman, Narimanabad and Lenkoran pygmaeus river. Ardea cinerea Single birds throughout the Caspian coast and lowland. Ardea purpurea Found in lowland, especially in wetlands near Liman. Ardea alba 1 adult in costal lagoons near Narimanabad. Egretta garzetta Common in wetlands of the Caspian lowland. Ardeola ralloides Up to 30 birds on 20 May in wetlands near Liman. 2 observations in the Caspian lowland in May. On 6 May 5 birds Bubulcus ibis foraging in montane meadow near Aliabad. Nycticorax Regularly seen in the lowland, especially in wetland near Liman. nycticorax Ixobrychus Occurred in the lowland. 4 migrants on 15 May in riparian forest minutus near Gosmalijan. Plegadis Up to 200 birds in wetlands near Liman on 2 July. falcinellus Ciconia nigra Occasionally seen in the Caspian lowland and the forest belt. Ciconia ciconia Breeds in villages in the lowland, but is rather rare. Pernis apivorus A few migrants seen from 7 to 14 May. Milvus migrans 1 bird on 10 May near Ashagy Bilnia and 1 bird on 27 May in

- 67 -

IUCN

b b b b b n (b) (b) (b) (b) (b) (b) (b) (b) b (b) b b (b) (b)

NT


Appendices

Moscow forest. 26

Neophron percnopterus

27

Gyps fulvus

28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54

Only a few times seen near Lerik and in Zuvand.

Only observed in Zuvand. 1 near Divagach on 8 May and 16 near Mistan on 15 May. Circaetus gallicus Regularly seen at altitudes below 1200 m a.s.l. Circus Breeds in the Caspian lowland. aeruginosus Circus macrourus Migrant. Aegypius 1 individual on 3 April near Aliabad. monachus 9 territories were found in the Caspian lowland around Masally Accipiter badius and Lenkoran (Heiss & Gauger 2008). Accipiter nisus Nine observations from April to July . Accipiter gentilis Twice seen in April in the lower foothills. Buteo buteo Common in the forest and montane meadow belt. Regularly seen in Zuvand upland with territorial behavior of a Buteo rufinus pair west of Pirasora Displaying individuals in forest near Günesli and montane Aquila pomarina meadows near Ashagy Bilnia. Regularly seen within the forest belt. Aquila nipalensis Several migrating birds seen from 3 April to 20 Mai. Regularly seen in Zuvand. 1 abandoned nest found in rocks near Aquila chrysaetos Shonadzhola. Aquila pennata Few observations within the forest belt. A small colony with 6 breeding pairs was found on 19 April in a Falco naumanni building in Khialakhan. Breeds in the riparian forest and in rocky habitats. Only 1 Falco tinnunculus observation in the Caspian lowland on 20 May near Boladi. Nests in the riparian forest, but was also observed in the Caspian Falco subbuteo lowland (near Tazakend, Moscow forest). Falco biarmicus An adult seen on 6 July on rocks near Pirasora. Falco peregrinus 1 individual on 9 July west of Khozavi. One singing male on 10 May in monane meadows near Ashagy Crex crex Bilnia. Gallinula An adult with a juvenile observed at a back water of the chloropus Lenkoran river in the lowland. Charadrius Was seen at costal lagoons near Narimanabad and Lenkoran dubius river. Himantopus 10 birds in wetlands near Liman on 10 July. himantopus Migrant. 70 birds on 2 July and 30 birds on 10 July in wetlands Limosa limosa near Liman. Tringa totanus 10 birds on 2 July in wetlands near Liman. Migrant. Observed along rivulets of the riparian forest and Tringa ochropus Lenkoran river near Vel. Two displaying individuals on 22 April at Lake Xanbulan are maybe migrants. Resting birds were found in the beginning of Actitis hypoleucos July at several locations in the Caspian lowland (costal lagoons near Narimanabad, wetlands near Liman, Lenkoran river). Arenaria 1 at a costal lagoon near Narimanabad on 20 and 24 May.

- 68 -

(b)

EN

n b b n n

NT NT

b b (n) b b b n b b b

VU

b b n (n) (b)

NT

b (b) b n n n (b)

n

NT

Appendices

55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81


interpres Philomachus pugnax Glareola pratincola Larus cachinnans Chroicocephalus ridibundus Chroicocephalus genei Thalasseus sandvicensis

1 bird in wetlands near Liman on 10 July. Occurred only in the lowland. 2 birds near Sirabil on 21 May and 4 birds at costal lagoons near Narimanabad on 20 May. Regularly at the Caspian coast. Regularly at costal lagoons and Lenkoran river (lowland). 30 individuals on 20 May near Liman. 10 birds on 20 May at the Caspian coast near Narimanabad.

A few observations in wetlands near Liman and the Caspian coast. A few observations in wetlands near Liman and the Caspian Sternula albifrons coast. Chlidonias Common, with up to 2500 individuals in wetlands near Liman. hybrida Chlidonias Common in wetlands near Liman. 1 migrating flock with 150 leucopterus individuals was seen on 2 May at 2300 m a.s.l. near Mistan. Stercorarius An adult seen at the Caspian coast near Narimanabad. parasiticus Columba livia A wild bird was seen on 1 June near Lialiakeran (Zuvand). Columba oenas Common in the forest belt. Common in the riparian forest. Only 3 observations within the Columba forest belt with 1 singing male near Günesli and 1 singing male palumbus near Tankivan. Streptopelia A few singing males observerd in the montane meadow belt and turtur Caspian lowland. Streptopelia Occasionally seen in the Caspian lowland, e.g. Masally. decaocto Streptopelia 1 bird at the Lenkoran river near Vel on 10 July. senegalensis Cuculus canorus Common in the Talish mountains at all altitudes. Several singing males with the montane meadow belt and the Otus scops riparian forest. Only 1 singing male in the lowland near Hirkan village on 3 July. 3 fledged juveniles found on 4 and 5 July in a rocky cliff east of Bubo bubo Mistan. Strix aluco Almost everywhere heard at night within the forest belt. Seen in montane meadow belt near Bilaband and the riparian Athene noctua forest. First seen on 22 April near lake Xanbulan. Singing males were Caprimulgus found from the Caspian lowland to rocky habitats throughout all europaeus habitat types. 2 chicks were found on 11 July near Vel (lowland). Tachymarptis Regularly seen along rocky cliffs e.g. Pirasora, Nalabin, Mistan melba and Khozavi. Apus apus Breeds in rocky habitats and towns throughout all altitudes. Occurred common along channels of the Caspian lowland. A few Alcedo atthis observations were made along rivers within the forest belt. 1 colony with about 100 pairs found north of Kumbashi (Caspian Merops persicus lowland). Sterna hirundo

- 69 -

n (b) b (n) (n) n n n b (b) n b b b b b b b b b b b b b b b b


82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106

During migration common throughout the Talish mountains. Breeds only at sandy cliffs of the montane semi-desert and along Merops apiaster rivers of the lowand. Only small colonies were found with 2-3 pairs each. Displaying individuals were found in the Caspian lowland and Coracias garrulus rarely in the montane meadow belt. Regularly seen in villages throughout the Talish mountains, Upupa epops especially in riparian forest, montane meadow belt and Caspian lowland. Several singing males within the forest belt, especially the parkJynx torquilla like degradation stage. Dendrocopos Rare breeder of the forest belt. Occurred in the lowland only in minor the Moscow forest. Dendrocopos Common in the forest belt. major Dendrocopos Common in the riparian forest, but was also observed in the syriacus Caspian lowland near Kumbashi and Tazakend. Dryocopus 1 singing male on 27 April in primeval forest near Siov at 800 m martius a.s.l. Picus viridis Common in the forest belt and riparian forest. Common wherever shrubs are present, especially in the montane Lanius collurio meadow belt, Caspian lowland, riparian forest and shrubby woodland stage. Migrant. 1 individual on 9 May near Ambu and 1 individual on Lanius isabellinus 18 May near Shinaband. Common in the riparian forest, but occurred also in the montane Lanius minor meadow belt and Caspian lowland. Only a few single birds in rocky habitats, riparian forest, Lanius senator montane meadow belt and Caspian lowland. Garrulus Regularly seen in the forest belt. glandarius Common breeder of Caspian lowland, montane meadow belt and Pica pica riparian forest. Pyrrhocorax A pair was several times observed in rocks west of Pirasora. pyrrhocorax Common around Kialvas, where it breeds in a colony in riparian Corvus frugilegus forest. Common breeder in open habitats, e.g. Caspian lowland, Corvus cornix montane meadow belt, riparian forest. Regularly seen in the forest belt and rocky habitats, where it Corvus corax breeds. Melanocorypha Occurred in the montane meadow belt near Aliaband and Muria calandra and in the Caspian lowland west of Masally. Melanocorypha Regularly in montane semi-desert above 1500 m a.s.l. bimaculata Calandrella Locally common in the Caspian lowland around Sirabil, rufescens Sarchuvar and west of Masally. Galerida cristata Only 1 bird seen west of Masally. Lullula arborea Common in the montane meadow belt and Zuvand. Common in the montane meadow belt around Lerik and montane Alauda arvensis semi-deserts and subalpine meadows e.g. around Mistan, Orand and Pirasora. Did not occur below 1000 m.a.s.l.

- 70 -

Appendices b

b b b b b b b b b n b b b b b b b b b b b b b b

NT

Appendices 107 108 109 110 111 112 113 114 115 116 117 118


Eremophila alpestris Ripparia ripparia Ptyonoprogne rupestris Hirundo rustica Delichon urbicum Periparus ater Parus major Cyanistes caeruleus Remiz pendulinus Aegithalos caudatus Sitta europaea Sitta neumayer

Breeds in the montane semi-desert. The first fledged juvenile was observed on 15 May near Mistan. Common migrant. No breeding colonies found. Regularly seen at steep, rocky cliffs. 2 nests found near Nalabin. Breeds in villages and is common in the Caspian lowland. Breeds in rocky habitats and towns throughout all altitudes. Common breeder of the forest belt. Common throughout the Talish mountains. Common within the forest belt and riparian forest. Occurred in the Caspian lowland. Regularly seen in the forest belt.

Common in the forest belt. Common in rocky habitats. Only 4 singing males observed. 2 near Siov in natural forest and 119 Certhia familiaris 2 near Sifiakeran in slightly disturbed forest. Troglodytes Common in the forest belt. A rare breeder in riparian forest with 120 troglodytes a nest found near Mistan. Breeds along rivers throughout the Talish mountains except the 121 Cinclus cinclus Caspian lowland. This species was distributed in two regions. It was common in riparian forest above 1200 m a.s.l., e.g. near Gosmalian, 122 Cettia cetti Shonadzhola and Mistan and occurred also in the Caspian lowland near Boladi and Liman along channels. Rare. 1 singing male on 28 April near Sim at a forest clearing 123 Locustella naevia was maybe a migrant. Furthermore, 1 singing male on 5 June near Bilaband in the montane meadow belt. Migrant. On 16 to 17 May a singing male in riparian forest near Locustella 124 Gosmalian. On 19 May a singing male near Kumbashi. 2 singing fluviatilis males near Boladi on 20 May. Was treated as a breeding bird, but is probably only a migrant. Acrocephalus 125 Singing males were regularly seen along channels of the schoenobaenus lowlands, but no observation was done after 1 June. Due to the hidden behavior and the similar song with Hippolais pallida maybe overlooked. Only 3 observations. A singing male Acrocephalus on 27 May in a park-like forest stage near Piran and a singing 126 scirpaceus male in the riparian forest near Mistan on 1 June are migrants. A further singing male on 20 May near Boladi (Caspian lowland) occupied probably a territory. 1 singing male on 19 May in bushes near Kumbashi. This species Acrocephalus is not listed in Patrikeev (2004), but obviously a regular migrant 127 dumetorum in autumn (Gauger 2005). This is the first spring record to Azerbaijan. Was treated as a breeding bird, but is probably only a migrant. Acrocephalus Was seen mostly in the riparian forest, but occurred rarely also in 128 the montane meadow belt and the lowland from 11 May to 1 palustris June. Common along channels and reedbeds in the lowland. 1 territory Acrocephalus 129 was found near Aliabad with a singing male from 6 May to 5 arundinaceus June.

- 71 -

b b b b b b b b b b b b b b b b

n n b b

n

(b)

b


130 Hippolais pallida 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156

Common in the Caspian lowland. First observed on 5 May. Common in lower altitudes of the forest belt (0-500 m a.s.l.). Hippolais icterina Singing males observed from 23 April to 21 June. Phylloscopus Several singing migrants from mid-April to mid-May. trochilus Maybe only a migrant. Several singing males heard at the Phylloscopus begining of April. However, 4 singing males heard near collybita Shinaband on 5 June. Phylloscopus Some singing males found in shrubs at higher altitudes above sindianus 1000 m a.s.l., but no sure signs of breeding. Phylloscopus Common in the forest belt. nitidus Sylvia atricapilla Common in the forest belt. Migrant. 1 bird on 16 May near Gosmalian, 1 bird on 17 May Sylvia borin near Mistan and 1 bird on 18 May near Pirasora. Sylvia communis Common in riparian forest. Several singing males in rocky habitats with bushes and in dense Sylvia curruca shrubs above 1000 m a.s.l. Probably an uncommun breeder in dense shrubs of higher Sylvia nisoria altitudes. Observed near Shinaband, Mistan and Khosavi. Breeds rarely at a few locations in the Caspian lowland. Sylvia mystacea Occurred as migrant also elsewhere, e.g. Zuvand. Muscicapa striata Common in the forest belt and riparian forest. Ficedula A male bird was observed on 4 April in riparian forest near hypoleuca Divagach. Ficedula Regularly seen in less degraded forests, especially around Siov. semitorquata Common breeder of the forest belt. Was at 3 April already Ficedula parva present. Erithacus Common in the forest belt. rubecula Luscinia Common in montane meadow belt, riparian forest, shrubby megarhynchos woodland stage and Caspian lowland. Breeds in rocky habitats with bushes above 2000 m a.s.l. in Zuvand upland. 2 singing males and a female found on 30 May Irania gutturalis west of Orand. A total of 4 pairs found near Mistan. 2 families observed on 4 July with 3 and 2 juveniles each. Cercotrichas On 21 May 3 birds near Garibljar and on 24 May 1 singing male galactotes near Boladi. Phoenicurus Common in rocky habitats. ochruros Phoenicurus Regularly in park-like forest stage and riparian forest. phoenicurus Rare. Observed e.g. at forest clearings near Siov or near Mistan Saxicola rubetra and Pirasora in Zuvand. Saxicola Breeds in rocky habitats of Zuvand upland. torquatus Oenanthe Locally common in rocky habitats of Zuvand, e.g. near Pirasora. oenanthe Also a rare breeder of the montane meadow belt. Common at rocky outcrops of Zuvand, especially near Pirasora, Oenanthe finschii Kialvas and Kialakhan. Oenanthe Rare breeder in rocky habitats of Zuvand.

- 72 -

Appendices b b n n (b) b b n b b b b b n b b b b b

b b b b b b b b

NT

Appendices


hispanica melanoleuca Oenanthe 157 isabellina Monticola 158 saxatilis 159

Monticola solitarius

160 Turdus torquatus 161 Turdus merula Turdus 162 philomelos

Common breeder of montane semi-desert. Common in rocky habitats. This species was only regularly seen in rocks between Shinaband and Nalabin. A pair was observed on 22 June near Kialakhan (K. Gauger pers. comm.) Locally common in rocky habitats above 1500 m a.s.l., e.g. near Shinaband, Mistan, Nalabin, Kialakhan, Orand. Common. Common in the forest belt.

Regularly seen in the forest belt, especially at the lower and upper treeline and the park-like forest stage. A family with 3 juveniles were found near Mamedoba at the lower tree line. 164 Oriolus oriolus Occurred in forest habitats. Common breeder of the riparian forest. Occurred also in montane 165 Sturnus vulgaris meadow belt and Caspian lowland. Several migrating flocks seen in the beginning of May. Occurred 166 Pastor roseus also elsewhere, but disappeared at the end of May. Rare breeder in rocky habitats of altitudes above 2000 m a.s.l., 167 Prunella collaris e.g. Mistan, Pirasora. The specimen of the species was shot in June 1880 at the Kyziurdi mountain (Radde 1886a) and since then never again seen in the region (Patrikeev 2004). Dr. H. Müller rediscovered 168 Prunella ocularis this species on 17 May in a rocky cliff with Juniperus spec. northeast of Mistan, where it was also seen on 22 June and 5 July by K. Gauger, J.Etzold and the author. Prunella Rarely seen in the forest belt. Occurred also in riparian forest 169 modularis near Mistan. 170 Motacilla alba Regularly seen in villages, Caspian lowland and riparian forest. 171 Motacilla citreola A bird on migration on 9 May near Dshangemiran. Motacilla flava A locally common breeder of the lowland, especially around 172 feldegg Boladi, Sirabil and Sarchuvar. Motacilla f. flava A pair was observed on 17 May near Pirasora. 173 Motacilla cinerea Regularly seen along rivers of the forest belt. Several territories found in Zuvand in 1400-2300 m.a.s.l in 174 Anthus campestris montane semi-desert. Occurred in April in migrating flocks of about 5 birds. Singing 175 Anthus trivialis males were observed since May e.g. around Dshangemiran, Shonadzhola or Pirasora. 176 Anthus pratensis 1 migrating bird near Aliabad on 4 April. 177 Anthus cervinus 1 migrating bird near Ashagy Bilnia on 10 May. 8 singing males and a food carrying bird was seen at 2000-2300 178 Anthus spinoletta m.a.s.l. on a sub-alpine meadow between Razgov and Mistan. 179 Emberiza cia Common in rocky habitats. Emberiza Common in rocky habitats around 2000 m a.s.l. Occurred during 180 hortulana migration also elsewhere. Emberiza Abundant in the montane meadow belt and also present in drier 181 melanocephala areas of the Caspian lowland. Arrived on 24 April from its

b b b b b b b

163 Turdus viscivorus

- 73 -

b b n b b

b b n b n b b b n n b b b b


Emberiza 182 calandra 183 Fringilla coelebs 184 185 186 187 188 189 190 191 192

193 194 195 196 197

Carpodacus erythrinus Carduelis chloris Carduelis spinus Carduelis carduelis Carduelis cannabina

wintering grounds. Common in the montane meadow belt and Caspian lowland. Occurred also along the riparian forest. Common and widespread species of the forest habitats. Occurred also in the lowlands along afforstations. Regularly seen in the riparian forest above 1400 m a.s.l., but occurred also elsewhere. Common below 1000 m a.s.l., above a rare sight. Regularly seen within the forest belt. Common in the whole Talish mountains.

Occurred above 1200 m a.s.l. near Lerik and was common in Zuvand upland. Locally common in rocky habitats with a few shrubs, e.g. near Serinus pusillus Nalabin, Khozavi, Orand. Pyrrhula pyrrhula Only a few observations within the forest belt. Coccothraustes Common in the forest belt. coccothraustes 1 pair at rocky cliffs northeast of Mistan and another pair, Rhodopechys including a singing male, on a rocky outcrop north of Pirasora on sanguineus 15 and 29 May, respectively. Fist seen on 8 May in Zuvand upland. Occurred regularly with Bucanetes up to 4 individuals near Divagach, Pirasora and Kialakhan. A githagineus fledged juvenile was seen in July near Divagach (C. Völlm pers. comm.). Common in villages, especially in the Caspian lowland with a Passer domesticus large flock of 500 individuals near Hirkan village on 2 July. Passer 3 birds near Divagach on 22 June (K. Gauger pers. comm.). hispaniolensis Passer montanus Only a total of 18 birds seen in the Caspian lowland. Petronia petronia Locally common breeder in rocks, e.g. west of Orand, Khozavi.

Appendices

b b b b b b b b b b b b

b (b) b b

Annex 2: Nesting guilds of selected bird species based on literature review (Flade 1994, Andretzke et al. 2005, Glutz von Blotzheim & Bauer 1991, Glutz von Blotzheim & Bauer 1994, Glutz von Blotzheim & Bauer 1998, Urquhart, & Bowley 2002, Kirwan et al. 2008, Alström & Mild 2003, Patrikeev 2004). Bird species

Nesting guild

Bird species

Nesting guild

Aegithalos caudatus

canopy

Lanius minor

shrub

Alauda arvensis

ground

Lanius senator

shrub

Alcedo atthis

cavity

Lullula arborea

ground

Alectoris chukar

ground

Luscinia megarhynchos

ground

Anthus campestris

ground

Melanocorypha bimaculata

ground

Anthus trivialis

ground

Melanocorypha calandra

ground

Aquila pennata

canopy

Motacilla alba

cavity

Aquila pomarina

canopy

Motacilla cinerea

cavity

Athene noctua

cavity

Motacilla flava

ground

Buteo buteo

canopy

Muscicapa striata

cavity

Caprimulgus europaeus

ground

Oenanthe finschii

ground

Carduelis cannabina

shrub

Oenanthe hi.melanoleuca

ground

- 74 -

Appendices


Carduelis carduelis

canopy

Oenanthe isabellina

ground

Carduelis chloris Carduelis spinus

canopy

Oenanthe oenanthe

ground

canopy

Oriolus oriolus

canopy

Carpodacus erythrinus

shrub

Otus scops

cavity

Certhia familiaris

cavity

Parus major

cavity

Ciconia nigra

canopy

Passer domesticus

cavity

Cinclus cinclus

cavity

Passer montanus

cavity

Circaetus gallicus

canopy

Periparus ater

cavity

Coccothraustes coccothraustes

canopy

Petronia petronia

cavity

Columba oenas

cavity

Phoenicurus ochruros

cavity

Columba palumbus

canopy

Phylloscopus collybita

ground

Coracias garrulous

cavity

Phylloscopus nitidus

ground

Corvus corax

canopy

Phylloscopus sindianus

ground

Corvus cornix

canopy

Phylloscopus trochilus

ground

Cuculus canorus

x

Pica pica

canopy

Cyanistes caeruleus

cavity

Picus viridis

cavity

Dendrocopos major

cavity

Prunella modularis

shrub

Dendrocopos minor

cavity

Pyrrhula pyrrhula

canopy

Dendrocopos syriacus

cavity

Remiz pendulinus

canopy

Dryocopus martius

cavity

Saxicola rubetra

ground

Emberiza calandra

ground

Saxicola torquatus

ground

Emberiza cia

ground

Sitta europaea

cavity

Emberiza melanocephala

shrub

Streptopelia turtur

shrub

Erithacus rubecula

ground

Strix aluco

cavity

Falco subbuteo

canopy

Sturnus vulgaris

cavity

Falco tinnunculus

cavity

Sylvia atricapilla

shrub

Ficedula hypoleuca

cavity

Sylvia communis

shrub

Ficedula parva

cavity

Sylvia curruca

shrub

Ficedula semitorquata

cavity

Sylvia nisoria

shrub

Fringilla coelebs

canopy

Troglodytes troglodytes

ground

Garrulus glandarius

canopy

Turdus merula

shrub

Hippolais icterina

shrub

Turdus philomelos

canopy

Hippolais pallida

shrub

Turdus torquatus

shrub

Hirundo rustica

cavity

Turdus viscivorus

canopy

Jynx torquilla

cavity

Upupa epops

cavity

Lanius collurio

shrub

Annex 3: Bird community table of the nine breeding bird communities including parameter values and species abundance per transect (territory/km).

Please see attatched extra sheet!

- 75 -


Annex 4: All indicator species of each community and every combination of communities. Community

Species

Stat

p-value

1

Hirundo rustica

0.963

0.000999***

1

Passer domesticus

0.937

0.000999***

1

Acrocephalus schoenobaenus

0.816

0.000999***

1

Hippolais pallida

0.816

0.000999***

1

Acrocephalus arundinaceus

0.802

0.000999***

1

Alcedo atthis

0.749

0.000999***

1

Circus aeruginosus

0.745

0.000999***

1

Coracias garrulus

0.723

0.000999***

1

Motacilla flava

0.667

0.000999***

1

Calandrella rufescens

0.577

0.003996**

1

Passer montanus

0.577

0.002997**

1

Cercotrichas galactotes

0.471

0.00999**

1

Merops persicus

0.471

0.012987*

1

Remiz pendulinus

0.471

0.012987*

1

Streptopelia decaocto

0.471

0.00999**

2

Pyrrhula pyrrhula

0.639

0.000999***

2

Ficedula semitorquata

0.622

0.003996**

2

Certhia familiaris

0.548

0.007992**

5

Phylloscopus sindianus

0.522

0.00999**

6

Emberiza melanocephala

0.576

0.006**

6

Coturnix coturnix

0.478

0.038*

7

Cettia cetti

0.983

0.000999***

7

Dendrocopos syriacus

0.802

0.000999***

7

Phoenicurus phoenicurus

0.707

0.000999***

7

Sylvia communis

0.672

0.000999***

7

Lanius minor

0.615

0.003996**

7

Columba palumbus

0.591

0.002997**

7

Acrocephalus palustris

0.585

0.003996**

7

Falco subbuteo

0.567

0.005994**

8

Anthus campestris

0.727

0.000999***

8

Oenanthe isabellina

0.7

0.000999***

9

Monticola saxatilis

0.93

0.000999***

9

Phoenicurus ochruros

0.907

0.000999***

9

Serinus pusillus

0.768

0.000999***

9

Turdus torquatus

0.637

0.001998**

9

Oenanthe oenanthe

0.608

0.006993**

9

Tachymarptis melba

0.522

0.015984*

1+7

Sturnus vulgaris

0.853

0.000999***

1+9

Apus apus

0.634

0.007**

2+3

Turdus philomelos

0.903

0.000999***

2+3

Troglodytes troglodytes

0.884

0.000999***

- 76 -

Appendices

Appendices


2+3

Erithacus rubecula

0.865

0.000999***

2+3

Columba oenas

0.562

0.006993**

2+4

Hippolais icterina

0.816

0.000999***

3+4

Dendrocopos minor

0.466

0.039*

5+7

Carpodacus erythrinus

0.539

0.028*

7+9

Sylvia curruca

0.686

0.000999***

7+9

Falco tinnunculus

0.633

0.002997**

8+9

Carduelis cannabina

0.831

0.000999***

8+9

Alectoris chukar

0.831

0.000999***

8+9

Emberiza cia

0.83

0.000999***

8+9

Eremophila alpestris

0.792

0.000999***

8+9

Sitta neumayer

0.782

0.000999***

8+9

Oenanthe finschii

0.69

0.000999***

1+4+7

Oriolus oriolus

0.672

0.000999***

1+5+6

Streptopelia turtur

0.505

0.025*

1+6+7

Emberiza calandra

0.917

0.000999***

1+6+7

Pica pica

0.856

0.000999***

1+6+7

Motacilla alba

0.762

0.000999***

1+7+8

Merops apiaster

0.707

0.000999***

2+3+4

Dendrocopos major

0.947

0.000999***

2+3+4

Sitta europaea

0.931

0.000999***

2+3+4

Periparus ater

0.914

0.000999***

2+3+4

Coccothraustes coccothraustes

0.871

0.000999***

2+3+4

Ficedula parva

0.824

0.000999***

2+3+4

Carduelis chloris

0.792

0.000999***

2+3+4

Carduelis spinus

0.603

0.004995**

2+3+5

Garrulus glandarius

0.507

0.02*

2+3+7

Cyanistes caeruleus

0.914

0.000999***

2+4+7

Muscicapa striata

0.74

0.000999***

5+6+7

Anthus trivialis

0.499

0.023*

6+8+9

Alauda arvensis

0.62

0.011*

7+8+9

Petronia petronia

0.868

0.000999***

1+3+4+7

Carduelis carduelis

0.769

0.003**

1+4+5+7

Luscinia megarhynchos

0.846

0.000999***

1+4+6+7

Corvus cornix

0.812

0.000999***

2+3+4+5

Aegithalos caudatus

0.754

0.000999***

2+4+5+7

Phylloscopus nitidus

0.639

0.000999***

4+6+7+9

Lanius collurio

0.749

0.000999***

1+5+6+7+8

Upupa epops

0.724

0.000999***

2+3+4+5+7

Turdus merula

0.974

0.000999***

2+3+4+5+7

Fringilla coelebs

0.96

0.000999***

2+3+4+5+7

Sylvia atricapilla

0.825

0.000999***

2+3+4+5+7

Picus viridis

0.728

0.000999***

- 77 -


5+6+7+8+9

Lullula arborea

0.84

0.000999***

1+2+3+4+5+7

Parus major

0.922

0.000999***

Appendices

Annex 5: Statistical analysis of the site parameters of the NMDS including all transects (Figure 21). ALTI = altitude, DATE = date, EXPO = slope exposition (after Parker 1982), SLO= slope steepness (after Parker 1982), HH = Herb layer height, HC = Herb layer cover, SH = Shrub layer height, SC = Shrub layer cover, LTH = Lower tree layer height, LTC = Lower tree layer cover, UTH = Upper tree layer height, UTC = Upper tree layer cover. Site parameters

r²

Pr(>r)

ALTI

0.7225

0.0001***

SLOP

0.4568

0.0001***

UTH

0.7977

0.0001***

UTC

0.5126

0.0001***

LTH

0.6683

0.0001***

LTC

0.4452

0.0001***

SH

0.2757

0.0001***

HC

0.4185

0.0001***

SC

0.1629

0.0004***

HH

0.1579

0.0006***

DATE

0.0979

0.0109*

EXPO

0.0543

0.09119

Annex 6: Statistical analysis of the site parameters of the NMDS including all forest transects (Figure 22). ALTI = altitude, DATE = date, EXPO = slope exposition (after Parker 1982), SLO= slope steepness (after Parker 1982), HH = Herb layer height, HC = Herb layer cover, SH = Shrub layer height, SC = Shrub layer cover, LTH = Lower tree layer height, LTC = Lower tree layer cover, UTH = Upper tree layer height, UTC = Upper tree layer cover. Site parameters

r²

Pr(>r)

UTH

0.7847

0.0001***

UTC

0.4377

0.0001***

LTH

0.5234

0.0001***

SC

0.3934

0.0001***

ALTI

0.3213

0.0007***

LTC

0.2197

0.005199**

DATE

0.1995

0.010799*

HH

0.1323

0.055594

SH

0.1348

0.058194

HC

0.1127

0.082692

EXPO

0.0654

0.258474

SLO

0.0414

0.431857

- 78 -

Appendices


Annex 7: Relative abundance values (territory/km) of each bird species per landscape type. Species Accipiter badius Accipiter nisus Acrocephalus arundinaceus Acrocephalus palustris Acrocephalus schoenobaenus Acrocephalus scirpaceus Aegithalos caudatus Alauda arvensis Alcedo atthis Alectoris chukar Anthus campestris Anthus spinoletta Anthus trivialis Apus apus Aquila chrysaetos Aquila pomarina Ardea cinerea Bucanetes githagineus Buteo buteo Buteo rufinus Calandrella rufescens Carduelis cannabina Carduelis carduelis Carduelis chloris Carduelis spinus Carpodacus erythrinus Cercotrichas galactotes Certhia familiaris Cettia cetti Cinclus cinclus Circaetus gallicus Circus aeruginosus Coccothraustes coccothraustes Columba oenas Columba palumbus Coracias garrulus Corvus corax Corvus cornix Coturnix coturnix Crex crex Cuculus canorus Cyanistes caeruleus Delichon urbicum Dendrocopos major Dendrocopos minor Dendrocopos syriacus Dryocopus martius Emberiza calandra Emberiza cia Emberiza hortulana Emberiza melanocephala Eremophila alpestris Erithacus rubecula Falco subbuteo Falco tinnunculus Ficedula parva Ficedula semitorquata Francolinus francolinus

Caspian lowland

Natural forest

Slightly disturbed forest

Intermediate disturbed forest

Park-like forest

Shrubby woodland

Montane meadow belt

Montane semidesert

Riparian forest

Rocky habitats

. . 0.99 0.05 1.73 0.06 . . 0.69 . . . . 2.36 . . 0.06 . . . 3.05 . 0.73 0.17 . 0.04 0.21 . 0.13 . . 0.30 . . . 0.43 . 1.42 0.04 . 0.69 0.17 0.47 . . 0.13 . 3.95 . . 0.34 . . 0.04 0.04 . . 0.04

. . . . . . 0.58 . . . . . . . . 0.12 . . 0.23 . . . . 0.12 0.58 0.12 . . . . . . 0.12 0.23 . . . . . . 0.58 0.70 . 3.84 . . 0.19 . . . . . 5.70 . . 3.60 0.47 .

. . . 0.44 . . 0.83 . 0.23 . . . . . . . . . 0.17 . . 0.08 0.08 0.66 0.25 0.08 . 0.25 . . . . 0.50 0.33 . . . . . . 0.58 1.40 . 3.47 . . . . . . . . 4.30 . . 3.14 0.41 .

0.13 . . . . . 0.38 . 0.11 . . . . 0.14 . 0.04 . . 0.08 . . . 0.50 0.96 0.23 0.04 . . . 0.08 . . 0.80 0.08 0.04 . 0.08 0.11 . . 0.66 1.61 . 2.11 0.15 . . . . . . . 1.95 . . 1.76 0.12 .

. . . . . . 0.08 . . . . . . . . . . . 0.09 . . . 2.02 2.35 0.42 0.09 . . . . . . 1.18 0.34 0.08 . . 0.84 . . 1.38 1.85 . 2.94 0.17 . . . . . . . 0.50 . . 1.26 0.09 .

. . . . . . 0.83 0.05 0.11 0.21 0.05 . 0.34 . . . . . 0.16 . . 0.53 0.59 0.54 0.05 0.56 . . . . 0.13 . 0.29 0.25 0.05 . . 0.44 . . 0.40 0.64 . 0.44 . . . 0.49 0.28 . 2.43 . 0.34 . . 0.05 . .

. 0.05 0.07 0.14 . . . 1.26 . . 0.05 . 0.24 0.43 . 0.15 . . 0.36 . . 0.05 0.39 . . . . . . . . . . . . 0.07 0.15 1.07 0.98 0.09 0.11 . 0.76 . . . . 7.23 . . 6.11 . . . . 0.05 . .

. . . . . . . 1.49 . 0.99 0.94 . . . . . . 0.18 . . . 1.44 0.06 . . . . . . . . . . . 0.06 . 0.06 0.11 . . 0.28 . 0.22 . . . . 0.66 1.22 0.28 . 2.49 . . 0.17 . . .

. . . 1.13 . . . . . 0.22 . . 0.06 . . . . . . . . . 1.97 . . 0.58 . . 2.55 0.29 . . . . 0.57 . . 0.89 . . 0.15 1.53 . . . 1.72 . 1.78 0.29 . . . . 0.16 0.32 0.19 . .

. . . . . . . 0.80 . 1.43 0.38 0.60 0.05 1.48 0.11 . . 0.26 0.09 0.11 . 1.60 0.57 0.05 . 0.28 . . . . . . . . 0.05 . 0.14 0.09 . . 0.66 . 1.04 . . . . 0.33 3.92 0.16 0.11 0.82 . . 0.38 . . .

- 79 -


Fringilla coelebs Garrulus glandarius Glareola pratincola Hippolais icterina Hippolais pallida Hirundo rustica Irania gutturalis Ixobrychus minutus Jynx torquilla Lanius collurio Lanius minor Lanius senator Lullula arborea Luscinia megarhynchos Melanocorypha bimaculata Melanocorypha calandra Merops apiaster Merops persicus Monticola saxatilis Monticola solitarius Motacilla alba Motacilla cinerea Motacilla flava Muscicapa striata Neophron percnopterus Oenanthe finschii Oenanthe hispanica melanoleuca Oenanthe isabellina Oenanthe oenanthe Oriolus oriolus Otus scops Parus major Passer domesticus Passer montanus Periparus ater Petronia petronia Phoenicurus ochruros Phoenicurus phoenicurus Phylloscopus nitidus Phylloscopus sindianus Pica pica Picus viridis Prunella collaris Prunella modularis Ptyonoprogne rupestris Pyrrhocorax pyrrhocorax Pyrrhula pyrrhula Remiz pendulinus Rhodopechys sanguineus Saxicola rubetra Saxicola torquatus Serinus pusillus Sitta europaea Sitta neumayer Streptopelia decaocto Streptopelia turtur Strix aluco Sturnus vulgaris Sylvia atricapilla Sylvia communis Sylvia curruca Sylvia mystacea Sylvia nisoria Tachymarptis melba Troglodytes troglodytes

0.04 . 0.11 . 2.48 4.81 . 0.05 . 0.34 0.21 0.04 . 3.86 . 0.15 0.52 1.42 . . 0.90 . 1.12 0.13 . . . . . 0.47 . 0.90 8.97 0.23 . . . . 0.04 . 2.06 . . . . . . 0.23 . . . . . . 0.13 0.21 . 2.40 . 0.13 . 0.04 . . .

7.21 0.35 . 0.35 . . . . . . . . . 0.12 . . . . . . . 0.23 . 0.35 . . . . . . . 0.81 . . 8.72 . . . 2.79 . . 0.58 . 0.47 . . 1.05 . . . . . 5.47 . . . . . 2.91 . . . . . 5.93

9.50 0.08 . 1.41 . . . . 0.11 0.12 . . . 0.11 . . . . . . . 0.41 . 1.18 . . . . . . . 1.49 . . 7.11 . . 0.08 2.39 . . 0.50 . . . . 0.08 . . 0.12 . . 4.63 . . . 0.08 . 1.65 . . . . . 5.04

7.39 0.08 . 1.21 . . . . . 0.13 . . . 1.69 . . . . . . 0.08 0.54 . 0.81 . . . . . 0.80 0.04 1.76 0.04 . 4.18 . . 0.04 0.44 0.06 . 0.34 . . . . . . . . . . 2.30 . . . . . 1.46 . . . . . 2.30

- 80 -

6.30 0.17 . 1.10 . . . . 1.10 2.09 . . 0.25 2.86 . . . . . . . . . 1.65 . . . . 0.08 1.45 . 3.61 . . 3.19 . . 0.83 0.28 . . 0.34 . . . . 0.08 . . . . . 2.52 . . 0.18 . . 1.18 . . . . . 0.76

2.70 0.34 . 0.50 . . . . 0.11 1.43 0.29 . 0.59 6.79 . . . . . . 0.10 0.20 . . . . . . . 0.21 0.05 3.33 0.11 . 0.39 . . 0.05 1.07 0.19 0.10 0.25 . 0.28 . . 0.05 . . . . . 0.10 0.05 . 0.45 . . 1.81 1.43 0.24 . 0.20 . 0.15

0.15 0.05 . . . 0.63 . . 0.11 1.54 0.13 0.07 0.73 1.35 . 0.73 . . . . 0.63 . . . . . . . 0.10 0.40 . 0.29 0.43 . . 0.11 . 0.10 0.05 . 0.87 . . . . . . . . 0.07 0.05 . . . . 0.27 . 1.02 . 0.94 . . . . .

Appendices 0.06 . . . . 0.06 . . . 0.27 0.21 . 2.10 0.06 0.41 . 0.50 . 0.44 . 0.06 . . . . 0.61 0.06 1.27 0.28 0.27 . . 0.06 . . 0.83 0.21 . . . 0.28 0.11 . . . . . . . . 0.11 0.06 . 0.33 . . . 0.61 . 0.07 . . . . .

2.87 . . . . . . . . 2.13 0.82 . 0.25 3.87 . . 0.22 . . . 1.15 0.13 . 1.23 . . . . . 0.49 0.13 3.44 0.88 . . 0.44 . 1.02 1.46 . 1.59 0.57 . 0.22 . . . . . . . . . 0.13 . . . 7.20 0.51 3.52 0.51 . . . 0.06

0.14 . . . . . 0.13 . . 0.66 0.05 0.07 0.66 0.60 0.07 . 0.11 . 1.89 0.44 0.24 0.09 . . 0.05 1.32 0.27 0.05 0.71 0.05 . 0.24 0.14 . . 2.26 2.80 0.05 . . 0.09 . 0.07 . 0.22 0.08 . . 0.07 0.05 0.19 0.90 . 3.07 . . . 0.19 0.09 0.40 0.49 . 0.07 0.14 0.05

Appendices


Turdus merula Turdus philomelos Turdus torquatus Turdus viscivorus Upupa epops

. . . . 0.39

3.95 1.28 . . .

4.79 2.56 . 0.11 .

3.22 1.76 . 0.05 .

5.29 0.76 . 0.37 0.25

4.02 0.39 . 0.23 0.25

0.19 . . 0.05 0.92

0.11 . 0.21 . 0.55

2.36 . . . 1.59

0.85 . 0.77 . 0.33

Annex 8: Relative abundance values (individuals/km) of each bird species per landscape type (including migrants and other non-breeders).

Species Accipiter badius Accipiter nisus Acrocephalus arundinaceus Acrocephalus palustris Acrocephalus schoenobaenus Acrocephalus scirpaceus Actitis hypoleucos Aegithalos caudatus Alauda arvensis Alcedo atthis Alectoris chukar Anthus campestris Anthus cervinus Anthus pratensis Anthus spinoletta Anthus trivialis Apus apus Aquila chrysaetos Aquila nipalensis Aquila pennata Aquila pomarina Ardea cinerea Ardea purpurea Bubulcus ibis Bucanetes githagineus Buteo buteo Buteo buteo vulpinus Buteo rufinus Calandrella rufescens Carduelis cannabina Carduelis carduelis Carduelis chloris Carduelis spinus Carpodacus erythrinus Cercotrichas galactotes Certhia familiaris Cettia cetti Chlidonias hybrida Ciconia ciconia Ciconia nigra Cinclus cinclus Circaetus gallicus Circus aeruginosus Circus macrourus Coccothraustes coccothraustes Columba oenas Columba palumbus Coracias garrulus Corvus corax Corvus cornix Corvus frugilegus

Caspian lowland

Natural forest

Slightly disturbed forest

Intermediate disturbed forest

Park-like forest

Shrubby woodland

Montane meadow belt

Montane semidesert

Riparian forest

Rocky habitats

. 0.04 1.03 0.05 1.77 0.06 . . . 0.69 . . . . . . 4.08 . . . . 0.06 0.23 0.04 . . . . 3.26 . 1.55 0.17 . 0.04 0.28 . 0.13 0.23 0.13 0.04 . . 0.30 . . 0.47 . 0.43 . 1.89 .

. . . . . . . 1.51 . . . . . . . . . . . . 0.12 . . . . 0.23 . . . . . 0.12 1.28 0.12 . . . . . . . . . . 0.12 0.23 . . . . .

. . . 0.44 . . . 1.16 . 0.23 . . . . . . . . . . . . . . . 0.17 . . . 0.08 0.08 0.83 2.81 0.08 . 0.25 . . . . . . . . 0.58 0.33 . . . . .

0.25 . . . . . . 0.73 . 0.11 . . . . . . 0.29 . . . 0.04 . . . . 0.08 . . . . 0.54 1.07 0.88 0.04 . . . . . . 0.16 . . . 1.00 0.08 0.04 . 0.08 0.15 .

. . . . . 0.14 . 0.17 . . . . . . . . 3.67 . . . . . . . . 0.09 . . . . 2.18 2.35 2.27 0.09 . . . . . . . . . . 2.02 0.42 0.08 . . 1.01 .

. . . . . . . 2.11 0.05 0.11 0.29 0.05 . . . 0.64 0.68 . . 0.11 . . . . . 0.21 . . . 0.85 0.88 0.64 0.15 0.56 . . . . . 0.05 . 0.25 . . 0.34 0.34 0.05 . . 0.59 .

. 0.05 0.07 0.14 . . . . 1.31 . . 0.05 0.09 0.05 . 0.24 0.76 . 0.05 . 0.19 . . . . 0.72 . . . 0.05 0.53 . . . . . . . . 0.05 . . . 0.07 . 0.73 . 0.13 0.24 1.65 0.13

. . . . . . . . 1.49 . 1.16 1.05 . . . . 1.82 . . . . . . . 0.35 . . . . 2.10 0.33 . . . . . . . . . . . . . . 0.22 0.06 . 0.06 0.11 .

. . . 1.13 0.13 0.37 0.08 . . . 0.22 . . . . 0.06 0.51 . 0.06 . . . . . . . . . . . 2.36 . . 0.66 . . 2.55 . . . 0.29 . . . . . 0.57 . . 1.08 .

. . . . . . . . 0.80 . 1.65 0.38 . . 0.60 0.05 3.24 0.11 . 0.05 . . . . 0.40 0.09 1.37 0.11 . 2.12 1.27 0.09 . 0.28 . . . . . . . . . . . 0.09 0.05 . 0.19 0.09 2.91

- 81 -


Coturnix coturnix Crex crex Cuculus canorus Cyanistes caeruleus Delichon urbicum Dendrocopos major Dendrocopos minor Dendrocopos syriacus Dryocopus martius Egretta garzetta Emberiza calandra Emberiza cia Emberiza hortulana Emberiza melanocephala Eremophila alpestris Erithacus rubecula Falco subbuteo Falco tinnunculus Ficedula hypoleuca Ficedula parva Ficedula semitorquata Francolinus francolinus Fringilla coelebs Garrulus glandarius Glareola pratincola Hippolais icterina Hippolais pallida Hirundo rustica Irania gutturalis Ixobrychus minutus Jynx torquilla Lanius collurio Lanius isabellinus Lanius minor Lanius senator Locustella fluviatilis Locustella naevia Lullula arborea Luscinia megarhynchos Melanocorypha bimaculata Melanocorypha calandra Merops apiaster Merops persicus Milvus migrans Monticola saxatilis Monticola solitarius Motacilla alba Motacilla cinerea Motacilla citreola Motacilla flava Muscicapa striata Neophron percnopterus Nycticorax nycticorax Oenanthe finschii Oenanthe hispanica melanoleuca Oenanthe isabellina Oenanthe oenanthe Oriolus oriolus Otus scops Parus major Passer domesticus Passer montanus Pastor roseus Periparus ater Pernis apivorus

0.04 . 0.69 0.52 0.69 . . 0.13 . 0.20 3.99 . . 0.34 . . 0.09 0.04 . . . 0.04 0.04 . 0.11 . 2.48 8.33 . 0.05 . 0.34 . 0.30 0.04 0.10 . . 3.91 . 0.15 2.36 3.22 . . . 1.16 . . 1.33 0.17 . 0.21 . . . . 0.60 . 1.03 17.21 0.34 3.22 . .

. . 0.70 0.81 . 3.95 . . 0.19 . . . . . . 5.70 . . . 3.72 0.47 . 7.44 0.35 . 0.35 . . . . . . . . . . 0.20 . 0.12 . . 0.58 . . . . . 0.23 . . 0.35 . . . . . . . . 0.81 . . . 9.77 .

. . 0.58 1.57 . 3.72 . . . . . . 0.12 . . 4.46 . . . 3.55 0.41 . 9.83 0.08 . 1.41 . . . . 0.11 0.12 . . . . . . 0.11 . . . . . . . . 0.41 . . 1.18 . . . . . . . . 1.65 . . . 7.85 .

. . 0.70 1.99 . 2.11 0.23 . . . . . . . . 1.99 . . . 1.80 0.12 . 7.55 0.11 . 1.28 . 0.11 . . . 0.20 . . . . . . 1.69 . . 0.13 . . . . 0.08 0.57 . . 0.87 . . . . . . 0.80 0.04 2.18 0.04 . . 4.44 .

- 82 -

. . 1.65 2.27 . 2.94 0.17 . . . . . . . . 0.50 . . . 1.26 0.09 . 6.39 0.17 . 1.10 . 0.76 . . 1.10 2.64 . . . . . 0.25 2.86 . . 1.10 . . . . . . . . 1.87 . . . . . 0.08 1.74 . 4.03 . . . 3.53 .

. . 0.45 0.88 . 0.44 . . . . 0.49 0.40 0.07 2.50 . 0.34 . . . 0.05 . . 2.94 0.44 . 0.50 . 0.25 . . 0.11 1.64 . 0.36 . . . 0.59 6.79 . . 1.38 . . . . 0.10 0.34 . . . 0.06 . . . . . 0.21 0.05 3.77 0.23 . . 0.54 .

0.98 0.09 0.11 . 1.47 . . . . . 7.48 . . 6.38 . . . . . 0.05 . . 0.15 0.05 . . . 1.17 . . 0.11 1.81 0.07 0.20 0.07 . . 0.73 1.35 . 0.83 0.88 2.55 0.09 . . 0.78 . 0.07 0.33 . . . . . . 0.10 0.40 . 0.39 1.63 . 0.40 . 0.14

Appendices . . 0.28 . 0.28 . . . . . 0.72 1.49 0.28 . 3.20 . . 0.17 . . . . 0.06 . . . . 0.61 . . 0.06 0.41 . 0.27 . . . 2.27 0.06 0.68 . 0.94 . . 0.44 . 0.06 . . 0.06 0.07 0.06 . 0.61 0.06 1.49 0.39 0.27 . . 0.06 . 3.42 . 0.07

. . 0.15 2.10 0.15 . . 1.78 . . 1.85 0.29 . . . . 0.16 0.45 0.06 0.19 . . 3.06 . . . . 2.61 . . . 2.62 . 0.98 . . . 0.25 4.16 . . 2.12 . . . . 1.46 0.13 . . 1.23 . . . . . . 0.49 0.13 4.52 1.09 . . . .

. . 0.71 . 1.98 . . . . . 0.33 4.29 0.22 0.11 0.88 . . 0.47 . . . . 0.19 . . . . . 0.20 . . 0.77 . 0.05 0.07 . . 0.71 0.60 0.07 . 0.11 . . 2.08 0.44 0.24 0.09 . . . 0.05 . 1.54 0.33 0.05 0.80 0.05 . 0.24 0.14 . 3.30 . .

Appendices Petronia petronia Phoenicurus ochruros Phoenicurus phoenicurus Phylloscopus collybita Phylloscopus nitidus Phylloscopus sindianus Phylloscopus trochilus Pica pica Picus viridis Prunella collaris Prunella modularis Ptyonoprogne rupestris Pyrrhocorax pyrrhocorax Pyrrhula pyrrhula Remiz pendulinus Rhodopechys sanguineus Ripparia ripparia Saxicola rubetra Saxicola torquatus Serinus pusillus Sitta europaea Sitta neumayer Streptopelia decaocto Streptopelia turtur Strix aluco Sturnus vulgaris Sylvia atricapilla Sylvia borin Sylvia communis Sylvia curruca Sylvia mystacea Sylvia nisoria Tachymarptis melba Tringa ochropus Troglodytes troglodytes Turdus merula Turdus philomelos Turdus torquatus Turdus viscivorus Upupa epops


. . . . 0.04 . 0.82 3.69 . . . . . . 0.34 . 2.10 . . . . . 0.17 0.30 . 12.10 . . 0.13 . 0.04 . . . . . . . . 0.43

. . . . 2.79 . 0.58 . 0.70 . 0.47 . . 1.05 . . . . . . 5.81 . . . . . 2.91 . . . . . . . 5.93 4.07 1.28 . . .

. . 0.08 0.25 2.39 . 0.45 . 0.50 . . . . 0.08 . . . 0.12 . . 5.62 . . . 0.08 . 1.65 . . . . . . . 5.12 4.96 2.56 . 0.11 .

. . 0.04 0.15 0.44 0.06 0.52 . 0.34 . . . . . . . . . . . 2.57 . . . . . 1.46 . . . . . . . 2.30 3.26 1.76 . 0.05 .

- 83 -

. . 0.83 0.17 0.28 . 0.33 . 0.34 . . . . 0.08 . . . . . . 2.86 . . 0.18 . . 1.18 . . . . . . . 0.76 5.38 0.76 . 0.55 0.25

. . 0.05 1.32 1.07 0.19 0.13 0.20 0.25 . 0.28 . . 0.05 . . . . . . 0.10 0.05 . 0.51 . . 1.86 . 1.43 0.24 . 0.20 1.59 . 0.15 4.12 0.39 . 0.23 0.25

0.11 . 0.10 . 0.05 . 0.59 1.36 . . . . . . . . 0.13 0.07 0.10 . . . . 0.60 . 18.79 . . 0.94 . . . . . . 0.19 . . 0.05 1.07

0.94 0.21 . . . . . 0.39 0.11 . . . . . . . . . 0.11 0.06 . 0.33 . . . 0.94 . . 0.07 . 0.06 . . . . 0.11 . 0.21 . 0.66

0.51 . 1.15 0.64 1.53 . 1.24 2.36 0.57 . 0.22 . . . . . 2.05 . . . . 0.13 . . . 13.38 0.51 0.13 3.69 0.51 . . . 0.06 0.06 2.42 0.06 . . 1.59

2.64 2.91 0.05 . . . 0.11 0.09 . 0.07 . 0.38 0.17 . . 0.13 . 0.05 0.24 1.13 . 3.44 . . . 1.46 0.09 . 0.40 0.49 . 0.07 0.19 . 0.05 0.94 . 0.82 . 0.33


Appendices

Erklärung zur Diplomarbeit Hiermit erkläre ich, die vorliegende Diplomarbeit mit dem Thema: „The breeding bird communities of the Talish mountains (Azerbaijan) and their response to forest degradation“ selbstständig verfasst und keine anderen Hilfsmittel als die angegebenen verwendet zu haben. Aus anderen Werken in Wortlaut oder Sinngehalt entnommene Inhalte sind durch Quellenverweis, auch für Sekundärliteratur, kenntlich gemacht.

Greifswald, den 30.06.2010

Michael Heiß

- 84 -

Annex3:Bi r dcommuni t yt abl eoft heni nebr eedi ngbi r dcommuni t i esi ncl udi ngpar amet erval uesandspeci esabundancepert r ansect( t er r i t or y/ km) .

ToDi pl omaThesi s‘ Thebr eedi ngbi r dcommuni t i esoft heTal i shmount ai ns( Azer bai j an)andt hei rr esponset of or estdegr adat i on’ byMi chaelHei ß