Next Article in Journal
Resilience of Terraced Landscapes to Human and Natural Impacts: A GIS-Based Reconstruction of Land Use Evolution in a Mediterranean Mountain Valley
Previous Article in Journal
Book Review: Bourgon, L. Tree Thieves: Crime and Survival in North America’s Woods; Little Brown Spark: New York, NY, USA, 2022; ISBN: 978-0316497442
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Land
Volume 13
Issue 5
10.3390/land13050591
land-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles thumb_up
...
Endorse textsms
...
Comment
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Relative Pollen Productivity Estimates for Mediterranean Plant Taxa: A New Study Region in Turkey

by
Esra Ergin
1,2,*,
Laurent Marquer
2,
Florence Mazier
3,
Ugo Bisson
2 and
Hasan Nüzhet Dalfes
1
1
Eurasia Institute of Earth Sciences, Istanbul Technical University, Maslak, Istanbul 34469, Turkey
2
Department of Botany, University of Innsbruck, Sternwartstrasse 15, 6020 Innsbruck, Austria
3
Laboratoire Géographie de l’Environnement, UMR 5602, CNRS, Université de Toulouse-Jean Jaurès, 31058 Toulouse, France
*
Author to whom correspondence should be addressed.
Land 2024, 13(5), 591; https://doi.org/10.3390/land13050591
Submission received: 18 February 2024 / Revised: 5 April 2024 / Accepted: 12 April 2024 / Published: 29 April 2024
Browse Figures
Review Reports Versions Notes

Abstract

:
This study estimates relative pollen productivity (RPP) for plant taxa from Southern Anatolia, an important region in the Mediterranean with a long history of human settlements. RPP estimates are required for quantitative pollen-based reconstruction of past land cover modelling. The application of the reconstruction by the REVEALS model in the Mediterranean basin is constrained due to the scarcity of the RPP values specific to the region. To better understand the relationship between vegetation cover and land use in the Mediterranean area, the present study aims to provide a set of RPPs for Turkey and the Mediterranean region. The study area centres around Gölhisar Lake in southwestern Turkey. Modern pollen data are collected from moss pollsters from 21 sites together with vegetation surveys. RPP estimates for the main taxa characteristic of the Mediterranean region are obtained (referenced to evergreen Quercus t.) using the extended R-value (ERV) model through the analysis of modern pollen assemblages. The most reliable results are acquired with the ERV sub-model 2 and Prentice’s taxon-specific method (using a Gaussian plume dispersal model) to distance-weighted vegetation data, corresponding to a Relative Source Area of Pollen (RSAP) value of 102 m. RPPs of dominant taxa in the study area are obtained for Quercus coccifera/Fagaceae (1 ± 0), Juniperus/Cupressaceae (0.279 ± 0.001), Fabaceae (0.008 ± 0.000), Pinus/Pinaceae (5.782 ± 0.011), and Poaceae (0.112 ± 0.001) and are comparable with other RPPs obtained in the Mediterranean region.

1. Introduction

Holocene vegetation changes in the southwestern Anatolian region hold important records for understanding the specific environmental and historical context of the area [1,2,3]. Palynological studies in this region have revealed an increase in human activities including agricultural practices, grazing, and fires, during the Holocene, as well as the relationship between human settlements and vegetation changes [1,4,5,6,7]. As a result of the human-related influences, the current vegetation comprises a mosaic of highly diverse landscapes characterized by both arboreal and non-arboreal species [5]. However, this high ecological diversity has presented a challenge for interpreting fossil pollen records from the late Holocene period where human impact has been recognized as a factor that has modulated vegetation dynamics for a long time [2]. To overcome these difficulties, modern pollen studies can assist in understanding paleoecological reconstructions based on fossil pollen analysis [3]. So far, modern pollen studies have been conducted in this region to examine the pollen production and dispersal mainly from moss pollsters, along with a limited number of surface soil, surface sediment, and pollen traps. However, these methods have been limited in the integration of modern pollen characteristics on paleoecological reconstructions based on fossil pollen analysis in southwestern Anatolia [2,3,5,8,9,10,11]. Therefore, our study introduces an additional approach that has not been applied in Turkey so far.
This study focuses on the area surrounding Gölhisar Lake in southwestern Turkey, which is renowned for its rich historical and cultural heritage. Particularly, the region is abundant in palynological and archaeological evidence, providing insights into human-based vegetation changes [1,4,5,12]. It is known for several ancient settlements such as Kibyra, Balboura, Oenoanda, Bubon, Kremna, and Sagalassos, each with their roots dating back to the Hellenistic period around the 3rd century BC [13]. This period can also be traced in some fossil pollen records, indicating a transition period within the Late Holocene, referred to as the Beyşehir Occupation Phase (3500 to 1300 calendar years ago), and revealing the clearance of Pinus and Quercus forests and simultaneous expansion due to agricultural practices. The reoccurrence of Pinus forests is thought to have occurred later, following the abandonment of the area [2,3,14]. Along with these long-term human-induced vegetation changes, climate changes such as wet and dry periods may have contributed to vegetation shifts, by expanding wetlands, and transforming woodlands into steppes [1,2,3].
Although fossil pollen evidence is widely recognized for its potential to provide a fundamental framework for understanding the relationships between ancient human settlements and changes in vegetation and climate, it does have certain limitations. One notable constraint is the non-linear relationship between pollen percentages and actual plant abundances [15]. Differences in pollen productivity, morphology, dispersal characteristics among taxa, and the size of sedimentary basins influence the relationship between pollen assemblage and the related plant abundance [16,17,18]. Therefore, pollen data transformed into quantitative reconstructions of past land cover play an important role in assessing the long-term perspectives of climate and anthropogenic land cover changes [19]. The Landscape Reconstruction Algorithm (LRA) is a method to quantitatively reconstruct vegetation abundance at regional and local spatial scales using pollen counts [16]. This approach reduces biases resulting from variations in pollen productivity and dispersal among different taxa, and differences in basin size. It enables the reconstruction of past abundances of individual plant taxa around pollen sites at a given spatial resolution. Regional Estimates of Vegetation Abundance from Large Sites (REVEALS) is one of the LRA models, a valuable tool for reconstructing past regional vegetation cover based on pollen data. The REVEALS model has been increasingly applied to investigate regional vegetation history across Europe including northern Europe [20], northwestern Europe, western Europe north of the Alps, and eastern Europe [21,22], as well as part of the eastern Mediterranean-Black Sea-Caspian corridor, respectively [23,24]. However, the application of the REVEALS model in the Mediterranean basin has been limited due to the scarcity of the relative pollen productivity estimates (RPPs) values specific to the Mediterranean region. Most of the RPP has been conducted in Europe and the northern hemisphere; however, only two studies have been carried out in southern France and southeastern Romania for the Mediterranean plant taxa so far [23,25]. To enhance the model’s applicability and reliability, it is important to gather more data on RPP values specific to the Mediterranean, thus enabling more accurate reconstructions of past vegetation in this area. The study area is situated within a mountainous landscape structure in the Oro-Mediterranean vegetation zone which displays distinctive biodiversity, differing from the European Mediterranean vegetation [3,26,27]. The pollen production can vary due to factors like species, geography, and climate [28,29,30]. Considering this, it is important to generate RPP values for vegetation reconstruction in southwestern Turkey based on data obtained from this region. To better understand the relationship between vegetation cover and land use in the Mediterranean area, the present study aims to provide a new set of RPPs specific to Turkey and the Mediterranean region.

2. Materials and Methods

2.1. The Study Area

This study encompasses a 50 km radius of Gölhisar Lake, which is geographically located between latitudes 37.58242° and 36.65060° N and longitudes 29.62034° and 29.57386° E (Figure 1a). The study area is situated within a mountainous landscape structure in the Oro-Mediterranean vegetation zone, with elevations ranging from 700 m to 1800 m (Figure 1a) [3,26,31,32]. A 50 km radius vegetation map was created using the data from the forest management plans of the General Directorate of Forestry of Turkey. According to this, prevalent vegetation communities within the 50 km radius are pine forests (Pinus brutia and Pinus nigra), juniper forests (Juniperus excelsa, Juniperus foetidissima), high mountain steppe, and Quercus coccifera shrublands (Figure 1b) [32,33].
In the study area, Pinus nigra forests extend between altitudes of 1200 to 2000 m, exhibiting a wide distribution [31,32,33]. They are generally accompanied by Cistus latifolius, especially in areas where black pine forests are typically damaged by fire [33,34]. Pinus nigra forests also form mixed forests with species such as Quercus cerris, Q. ithaburensis, Q. pubescens, Q. coccifera, as well as Platanus orientalis in humid places [33,34]. Juniper forests typically appear at high elevations on limestone cliffs and often mix with Pinus nigra in the Oro-Mediterranean vegetation zone [31]. Juniper forests have a sparse structure due to the elevation, rocky terrain, and high mountain steppe vegetation prevailing in these clearings [33,35]. Juniperus foetidissima is common at higher elevations and J. excelsa is dispersed in the lower and middle regions [30]. Another dominant vegetation formation, Quercus coccifera maquis vegetation, which is distributed between 800 to 1300 m a.s.l., is seen together with Berberis crataegina and Juniperus oxycedrus subsp. oxycedrus, often occurring with cushion-like plant communities such as Astragalus, Alyssum, and Acantholimon in the clearings [33]. The high mountain steppe vegetation is mostly located in the eastern part of Gölhisar and is distributed between 800 to 1800 m a.s.l. [35]. This vegetation is characterized by a form of thorny cushion formations comprising Astragalus, Acantholimon, and also other herbaceous species such as Verbascum, Silene, Salvia, Thymus, Thymbra, Origanum, Lamium, Alyssum, Teucrium, Fumana, Potentilla, Centaurea, and Poaceae family species [33,35].
The long-term climate data for Gölhisar extending from 2005 to 2022, sourced from the Turkish State Meteorological Service for the Gölhisar station (37.14275° N, 29.526° E), reveals that the mean temperature is 2 °C in January and 23 °C in July. The annual average precipitation amounts to 447.85 mm, with approximately 41% occurring during the winter and 13% during the summer (Figure 2).

2.2. Site Selection

Sampling sites were randomly selected within a 50 km radius around Gölhisar Lake using QGIS software version 3.22.9-Białowieża. The random selection criteria ensured a minimum distance of 1500 m between each sampling location to prevent autocorrelation [36]. Out of the randomly selected 50 points, 13 were situated within agricultural zones, 10 were positioned at elevations exceeding 1800 m with access difficulties, and 3 were located in sites near hydroelectric power stations (HES) and tree plantations. Consequently, these specific points were excluded from the fieldwork, which was carried out in September 2021.
Finding moss species in the southern part of Turkey poses certain challenges. Out of the initially selected 50 random sites, a total of 24 sites were deemed suitable for data collection; however, moss samples were collected from 21 sites as no moss samples were encountered at 3 of the sites which were found to be sparsely vegetated, mostly covered with rocks, eroded and located in steppe vegetation formation. The moss samples were typically found in forest areas and Quercus coccifera shrublands among sampling sites. The visited 21 sites were situated at elevations ranging from 668 to 1842 m above sea level, as detailed in Table 1.
It is important to note that the geography of the study area with elevations of 1000 m and above poses physical challenges in accessing the sampling points. The Taurus Mountains constitute an extended mountain range situated in the Mediterranean area of Turkey. Stretching for approximately 560 km in parallel to the Mediterranean shoreline, they are used as the southern boundary of the Anatolian plateau [27].

2.3. Pollen Data

Studies show that moss pollsters preserve several years of pollen loads and they represent modern vegetation according to seasonal variation in species abundance [36,37,38,39]. In our study, moss pollsters were collected from a 1-square-meter area at each site, with soil particles removed, before the vegetation survey. In the laboratory, any remaining soil particles in the moss were eliminated prior to pollen extraction. The pollen samples underwent standard procedures outlined in Faegri and Iversen’s 1976 protocols [40]. These procedures involved the treatment of samples with a 10% HCl solution to remove carbonates, a 10% KOH solution to eliminate humic acids, a 46% hot HF solution, and a hot acetolysis process to ensure the removal of any remaining cellular content, calcareous and mineral components from the sample. Additionally, after the HCl and KOH treatments, the moss underwent sieving through a 0.25 mm mesh to filter out any larger mineral or organic residues. A minimum of 1000 pollen grains per sample were counted and identified using identification pollen keys [41,42] and the reference pollen collection at the University of Innsbruck (Department of Botany).

2.4. Vegetation Data

A detailed vegetation survey was conducted within a radius of 0–100 m from each pollen sampling point, following the main guideline of the Crackles Bequest Project protocol [36]. To assess the percentage cover of every plant species, including trees, shrubs, and herbs, visual estimations were made within a 1-square-meter quadrat.
Vegetation was surveyed in two main zones. In the first zone, between 0 to 10 m, the survey involved the use of 1-square-meter quadrats (Figure 3). This included a central quadrat positioned around the pollen sample site, along with 20 additional quadrats placed at specific distances of 1, 2.25, 4.5, and 8 m in all four cardinal directions (N, E, S, W), as well as at 7.5 m in the diagonal directions (NE, SE, SW, NW), totalling 21 quadrats. The composition of plant taxa was assessed by estimating the total plant cover as a percentage of the quadrat area.
In the second zone, ranging from 10 to 100 m, the boundaries between different vegetation communities were drawn using compass directions and measured distances obtained from a handheld GPS device (Figure 3). Following, and during, the mapping process, the coverage of plant taxa within each community was estimated. For open communities, 1-square-meter quadrats were used, whereas semi-open and forest communities were determined using a 6 m rope in all cardinal directions with point surveys. The major vegetation types identified during the field survey are detailed in Table 1.
For the zones ranging from 100 to 5000 m, a vegetation map was created using the data from the forest management plans of the General Directorate of Forestry of Turkey along with local biodiversity and vegetation studies [33,34]. Subsequently, certain features were associated with specific vegetation and land cover type categories, totalling 16 in all (Table 2). These categories were defined based on the vegetation communities identified during the field surveys. To assign taxa compositions to each land cover type, we assumed that they closely resembled the vegetation communities surveyed within a 100 m radius during fieldwork at each study site. To address vegetation units not covered during the research, botanical expert opinions were formed utilizing information obtained from the flora and vegetation of Turkey. Subsequently, land cover maps for distances ranging from 100 m to 5000 m at the study sites were generated using QGIS software version 3.22.9-Białowieża. The distance increment from the centre of the moss sample area was set at 1 m increments from 100 m to 5000 m. This involves calculating the percentage coverage of each vegetation type within each concentric ring, with increments of 1 m (Figure 3).

2.5. ERV Model Runs

In this study, we applied the ERV model using ERV-Analysis software version 2.5.4, a program developed by Sugita (unpublished), to determine the RPPs for different plant taxa. Each sub-model calculates the RPP and background pollen content by comparing pollen and vegetation datasets gathered from various pollen sampling sites. Besides the pollen and vegetation data, the ERV model required additional input including the FSP rates and wind speed (taken as 3 m s−1) to run properly. There are three ERV sub-models: ERV models 1 and 2 estimate a species’ constant background in pollen proportion, whereas ERV model 3 estimates a constant background pollen loading instead of pollen proportion [18,43]. We applied the ERV model to three different distance-weighting functions: the inverse distance (1/d), Prentice’s model (uses a Gaussian plume diffusion model (GPM)) [44], and the Lagrangian stochastic dispersal model (LSM) [45].
To identify the most dependable estimates for the RPP, various combinations of ERV sub-models and distance-weighting methods were employed, and the outcomes were subsequently compared. In conducting the ERV model, we selected specific plant/pollen taxa for RPP estimation, focusing on species present both in the vegetation and the pollen samples. For initial analysis, we selected eleven pollen types comprising Pinus, Cedrus, Juniperus, evergreen Quercus t., deciduous Quercus t., Olea, Phillyrea, Asteraceae, Fabaceae, Chenopodiaceae, and Poaceae (Table 3). We selected FSP estimates from the Mediterranean studies except Olea [23,25]. The Cedrus estimate is from Borrel (2012) [46]. The Olea FSP was calculated for this study from the reference taxon collection presented by Istanbul University, Institute of Marine Sciences and Management (Table 3). A minimum of 30 grains of Olea pollen were measured for both their long and short axes on reference slides. These measurements were used to calculate the fall speed of pollen (FSP) based on Stokes’ Law of particle settling velocity [47] (Table 3). Relative Source Area of Pollen (RSAP) was estimated statistically using a moving window regression approach and the moving window was 200 m. A reference taxon with an RPP value of 1 was used for calculating RPPs for other taxa.
Typically, the most common reference taxon for ERV models is Poaceae, given its wide distribution which includes many plant species with differing ecologies [29,48,49,50]. This variety may lead to variations in pollen production among regions [29]. Our study region did not provide an adequate abundance of Poaceae for ERV modelling. Therefore, we chose a single-species reference taxon, Quercus coccifera, as the reference taxon, given its abundance in both pollen and vegetation data.
During this process, we identified taxa that exhibited the most linear relationships between pollen and vegetation data. The procedure was repeated, gradually reducing the number of taxa, until the log-likelihood and RSAP curve assumed the theoretically optimal shape which reached an asymptote at a certain distance.

3. Results

3.1. Pollen Assemblages

A total of 48 pollen types were identified in 21 moss samples. The pollen types such as Geraniaceae, Ericaceae, Cerealia t., Abies, Artemisia, Potentilla, Xanthium, Juglans, Urtica, Polygonum, Plantago, Corylus, Convolvulus, Alnus, Pistacia, and Cannabis were only observed in the pollen slides and were not encountered in the vegetation survey. Pinus (79%), evergreen Quercus t. (6%), and Cedrus (5%) were the main taxa encountered in pollen. Although Juniperus species were observed in high numbers within the vegetation, their pollen counts did not reflect a similar abundance. Pinus pollen was abundant across all sites, yet its abundance was significantly lower in Site 23 compared to the other sites (Figure 4). The vegetation at Site 23 was composed of a Cedrus-Juniperus forest, which explains the higher abundance of Cedrus pollen and the relatively lower presence of Pinus in this sample (Figure 4 and Figure 5). The Platanus pollen was found only in Site 45, as the site constituted a mixed forest of Pinus and Platanus. Although the deciduous Quercus type was represented in almost all samples, its actual representation in the local vegetation was relatively low. Olea and Phillyrea were recorded only at one location, a red pine (Pinus brutia) forest at a low altitude (Site 13), although they were present in many of the pollen samples. The most common taxon among herbaceous plants was Poaceae, especially in the steppe sites 22, 49, 34 (Figure 4 and Figure 5). Although Fabaceae, Brassicaceae, Caryophyllaceae, and Asteraceae (Cichorioideae, Leucanthemum (Anthemis) t., Artemisia, Centaurea) had higher vegetation abundances, their representation in the pollen samples was low.

3.2. Vegetation Data

At the local scale, our field survey revealed a great prevalence of Pinus spp. in the vegetation composition. Among the 21 sampled points, 7 were representative of coniferous forests, another 7 of mixed forests where Pinus spp. was the dominant species, 4 of Quercus coccifera maquis, and 3 of steppe vegetation. In the field survey, 34 plant families and 90 genera were identified. The primary arboreal species in coniferous forests typically included Pinus spp., Juniperus spp., and Cedrus libani. However, it is important to note that only a single site had a Cedrus libani community. Land cover types in a 5 km zone (Figure 5) show that arable land is a common land cover type among all sites. Pine (Pinus spp.) forest land cover type is abundant across all sites; however, it is observed in notably lower percentages in the steppe sites (22, 34, 49), the Cedrus libani-Juniperus spp. forest type (Site 23), and the Q. coccifera shrubland type (Site 33). Pinus brutia forest (Site 13) is the only forest located at a lower elevation exhibiting a nearly homogenous vegetation community within a 5 km radius. Similarly, Sites 22, 34, and 49 are situated in the steppe area and exhibit a homogenous vegetation community. Maquis vegetation is present only at Site 50 due to its proximity to the southern coastal area and the corresponding decrease in elevation. Deciduous Quercus forest-type abundance is lower and also seen mixed with pine forest in Sites 41 and 50.

3.3. Relevant Source Area of Pollen

For the first runs of ERV models, we assessed the log-likelihood curve and identified the RSAP, then plotted the pollen/vegetation relationships from the three sub-models at the RSAP distance and evaluated these relationships. Eleven pollen types comprising Pinus, Cedrus, Juniperus, evergreen Quercus t., deciduous Quercus t., Olea, Phillyrea, Chenopodiaceae, Asteraceae, Fabaceae, and Poaceae were selected for the first analysis. However, weak correlations were detected between six of these taxa due to the low vegetation abundances. Hence, five taxa, Q. coccifera, Juniperus, Fabaceae, Pinus, and Poaceae (listed in Table 4) were selected for the final ERV analysis.
The RSAP estimates were obtained from three ERV sub-models and three distance-weighting models, along with the corresponding log-likelihood function curves for nine combinations of ERV sub-models and distance-weighting methods, as shown in Figure 6 and Table 4. Within a given distance-weighting method, ERV sub-model 1 consistently exhibited the weakest performance, characterized by the lowest log-likelihood values among the three ERV sub-models and a log-likelihood curve that deviated from the expected pattern. On the other hand, ERV sub-models 2 and 3 employed similar equations to linearize the pollen–vegetation relationship but differed in RSAP values (Figure 6a, Table 4). The combinations of ERV sub-model 2 with Prentice’s and 1/d vegetation distance-weighting methods produced highly similar log-likelihood values and trends (Figure 6a).
Meanwhile, the combination of ERV sub-model 2 with LSM distance-weighting produced similar log-likelihood values yet exhibited significant variation and produced low RSAP values. The ERV sub-model 2 paired with Prentice’s and 1/d vegetation distance-weighting method produced log-likelihood curves that aligned most closely with expectations, characterized by a gradual increase towards an asymptote, and thus, was deemed the most favorable outcome. In the optimal model combination (ERV sub-model 2 with Prentice’s vegetation distance-weighting method), the log-likelihood curve stabilized and maintained consistent values after approximately 100 m (Figure 6b). Consequently, we defined the RSAP as 102 m in this study

3.4. Pollen–Vegetation Relationships

The scatter plots illustrating the original (percentage-based) and ERV-adjusted pollen and vegetation values, as calculated with sub-model 2 and Prentice’s model, are depicted in Figure 7. Among the taxa examined, Quercus coccifera demonstrates the relationship that most closely resembles an ideal linear correlation, with Juniperus coming in as the second closest (see Figure 7). In contrast, Fabaceae and Poaceae exhibit a broad range of values, where high plant cover tends to coincide with low pollen values. And finally, Pinus presents a notably elevated background pollen loading.
We obtained the RPP value of Fabaceae, despite the nonlinear relationship between its vegetation proportion and adjusted pollen proportion. The Fabaceae family is the second most common among the top three families in Turkey [34]. One of its most prevalent genera, Astragalus, is widely distributed in Turkey, with 440 species [34], as frequently recorded during our fieldwork. Therefore, we decided to include it for future comparison in similar studies conducted in Turkey.

3.5. Relative Pollen Productivity Estimates

RPP estimates, their corresponding alpha values, and standard deviation values (SD) for five taxa are illustrated in Figure 8 and Table 4. These estimates were computed by applying three different ERV sub-models and employing three alternative distance-weighting methods.
In this study, the values obtained using ERV sub-model 1 consistently exhibited lower RPP values for Poaceae, Pinus and Juniperus, ERV sub-model 3 exhibited lower RPPs for Poaceae, Juniperus and Fabaceae, while ERV sub-model 2 yielded higher values for Poaceae, Juniperus and Fabaceae, than the other two sub-models regardless of the distance-weighting method used. According to the results, Pinus (1.42–6.06) is the highest pollen producer while Poaceae, Juniperus, and Fabaceae exhibited lower values among the three ERV sub-models. Therefore, we focused on ERV sub-model 2 and distance-weighting methods (Prentice’s (GPM), 1/d, and LSM) which tended to produce similar RPP values except Pinus (5.91–5.78–6.06). Consequently, we obtained the most accurate results, with lower RPP values for Pinus (5.78) and higher for the other taxa, from the ERV sub-model 2 and Prentice’s taxon-specific method.

4. Discussion

This study provides the first RPP estimates for Turkey, in a region characterized by Oro-Mediterranean vegetation. This new set of RPP opens the door for quantitative Holocene vegetation reconstructions in the Mediterranean. The RPP results obtained in this study can help us to understand historical landscape changes in the Southwestern part of Turkey, through an analysis of fossil pollen assemblages obtained from lake or marsh sediments. This approach enables the assessment of various processes related to vegetation patterns, thereby enhancing our understanding of human-induced landscape changes and biodiversity.

4.1. Challenges of Fieldwork in the Mediterranean Region

The fieldwork was carried out in September when pandemic restrictions were lifted, and permissions for field access were regranted. Despite September being late in the season for assessing plant diversity in the Mediterranean region, there were no problems in identifying trees and perennial herbaceous plants. Additionally, various annual herbaceous plants were identified, owing to the beginning of autumn precipitation. It is important to note that conducting the fieldwork during the spring season has the potential to substantially enhance the quality of the study and the RPP estimates through the facilitation of a more comprehensive collection of vegetation data, particularly for taxa such as Poaceae and other herbaceous species [39,51]. Studies show that the timing is crucial for vegetation surveys, especially for grassland communities, that exhibit variation throughout the fieldwork season, while synchronization is less critical for woodlands [38,39]. Out of the initially selected 50 random sites, 24 sites were visited; however, the moss samples were collected from 21 sites. No moss samples were encountered at three of the sites, as these sites were sparsely vegetated, mostly covered with rocks, and located in steppe vegetation formation. The moss samples were typically found in forest areas and Quercus coccifera shrublands among sampling sites. Finding moss samples in the southern part of Turkey poses certain challenges. The study area’s geographical structure with elevations of 1000 m and above poses physical challenges in accessing the sampling points. Finally, to generate reliable RPP values that will encompass the main Mediterranean taxa (Cedrus libani, Abies cilicica, Olea europea, Phillyrea, deciduous Quercus t., etc.) within all this diversity, the boundaries of the study area should be expanded and the number of randomly selected points increased.

4.2. Reliable RPPs for Southwestern Turkey

In the first run of the ERV model, we were unable to produce RPP values for Cedrus, Olea, and Phillyrea species due to the low vegetation abundance of these plants, according to the distance-weighted model results, especially because only one cedar forest (Site 23) and one red pine forest (Site 13) could be visited. Similarly, there were only two deciduous Quercus forests (Site 41 and Site 50) visited and deciduous Quercus t. was excluded from the ERV model analysis due to its low abundance in the vegetation and the distance-weighted model results. As a result, we were unable to calculate RPP values for the deciduous Quercus type, Cedrus, Olea, and Phillyrea.
In the second run of the ERV model, we have selected six pollen types, including Pinus, Juniperus, evergreen Quercus t., Fabaceae, Asteraceae, and Poaceae which are well represented in both vegetation and pollen records. With the initial results, we observed that the SD for Asteraceae exceeded the RPP value, so we opted to exclude the Asteraceae from the final run. Consequently, we ran the ERV model with five taxa, Pinus, Juniperus, evergreen Quercus t., Fabaceae, and Poaceae.
Quercus coccifera was selected as the reference taxon as it is appeared well represented in both vegetation and pollen records. We observed that every sampling site with evergreen Quercus t. pollen was also abundant as a vegetation type (i.e., Quercus shrublands) within a 5 km radius, as depicted in Figure 5.
According to the results of our fieldwork, there is a clear dominance of Pinus pollen abundance (79%) in the data obtained from moss samples (Figure 4). In a study conducted in southwestern Turkey at the end of the 1990s, a comparison was made between moss and trap samples [52]. This Pinus dominance was also observed in the BNP (Blackdown Hills Natural Futures Project), where the comparison of moss samples and Tauber traps (averaging 2-year pollen data) indicated that Pinus proportions in moss pollsters were, on average, twice as high [53]. These variations are typically attributed to the unique structures of moss cushions, which result in differential efficiency in trapping and preserving pollen of varying sizes and shapes [54,55]. Additionally, pollen deposition in mosses may be influenced by occasional extreme meteorological events or pollen addition by insects [50,56].
The Fabaceae recorded in the vegetation survey were generally in the form of perennial low shrubs. Particularly, cushion-form species belonging to the Astragalus genus have been commonly observed. Similarly, although in smaller quantities, shrubby forms of Dorycnium, Spartium junceum, Calicatome, and Cytisus species were also recorded. While Fabaceae’s actual representation in the vegetation yielded a higher value, it presented low pollen abundance, as well as a very low RPP value. Fabaceae are primarily insect-pollinated, thereby influencing their pollen abundance and resulting in low pollen production.
Although Poaceae were recorded during the fieldwork, they exhibited low pollen abundance. This could be attributed to the fact that the fieldwork was conducted in the autumn or the possibility that they were not well preserved in the moss samples.
There is still a need to have RPP estimates for Cedrus, Olea, deciduous Quercus t.and Phillyrea, among othere for the Mediterranean area. Therefore, it may be beneficial to consider increasing the number of randomly selected points to overcome the accessibility challenges arising from elevation variations and to prevent points from overlapping with agricultural areas. Additionally, it would be beneficial to expand the study area to encompass lower elevations in order to generate RPP values for taxa such as Pistacia, Ericaceae, and other maquis plant taxa.

4.3. Relevant Source Area of Pollen (RSAP)

RSAP is influenced by several factors, including the size and type of sediment basin, as well as the spatial distribution of taxa and vegetation patches within the studied landscape [18,57,58]. Furthermore, the boundary separating vegetation surveying strategies within zones of 10–100 m and beyond 100 m can influence RSAP outcomes in pollen modelling research [36]. Our research area displays a diverse landscape comprising various elements, including forests, semi-open landscapes, high mountain steppes, and Quercus coccifera maquis. In our study, the RSAP value is 102, while in similar studies, 100 m in the forest steppe landscape of southeastern Romania [25], and 145 m in cultural landscapes in central eastern China were calculated [48]. Although the study area exhibits diverse habitat types, the sites where vegetation surveys were conducted generally contain homogeneous plant covers. This could explain the low RSAP values found in our study. The data pertaining to vegetation units beyond 100 m were obtained from the forest management plans of the General Directorate of Forestry of Turkey. Plant abundances based on these data are relatively coarse. This could be another possible reason explaining the low RSAP values. It is therefore, recommended that the vegetation data beyond 100 m should be further examined and assessed.

4.4. Relative Pollen Productivity (RPP) Estimates

In determining the RPP values, we assessed both the log-likelihood scores for each model and the RPP values associated with each taxon, ultimately concluding that sub-model 2 was the most appropriate choice. Estimates of RPP and SDs from the three ERV sub-models for five pollen taxa and comparisons are shown in Table 4.
The research carried out in southern France [23] has generated new FSP and RPP values specifically for seven taxa found in the sub-Mediterranean and Mediterranean regions. These taxa include Buxus sempervirens, Carpinus orientalis, Castanea sativa, Ericaceae (Mediterranean species), Phillyrea, Pistacia, and the evergreen Quercus type t. [23]. When we compare our RPP values with those reported in previous research conducted in the Mediterranean region, we observe both similarities and differences. While we used Quercus coccifera as the reference taxon in our study, Poaceae was often used as a reference taxon in other studies. In this study, we determined the average RPP values for the Mediterranean region using software developed by Sugita (unpublished), to calculate Standard Errors based on the Delta Method.
To facilitate comparison, we recalculated our RPP values to use Poaceae as the reference taxon, as shown in Table 5. Our RPP values for evergreen Quercus t. (8.97) and Juniperus (2.52) are relatively similar to the values found in France (11.04 and 1.61). However, our estimate for Fabaceae (0.07) is lower than that reported in Romania (0.4). One of explanations could be that in the Romanian study, the Fabaceae predominantly constituted herbaceous plants, whereas, in our study, they were predominantly composed of perennial Fabaceae species in the form of cushions or bushes. In addition, Fabaceae species are primarily insect-pollinated, thereby influencing their pollen abundance and resulting in low RPP. Wind-pollinated (anemophilous) plants tend to produce a high amount of pollen compared with insect-pollinated (entomophilous) plants [59]. Insect-pollinated plant’s pollen is not carried by the wind as well and tends to be under-represented in the pollen records [59].
The RPP estimates for Pinus in Europe generally tend to be lower than our estimates. Our estimation for Pinus (51.85) is notably higher than findings from forested regions in northern China [60] and northern Europe [61], although it is similar to the value reported in Poland (51.38) [53]. In our study area, Pinus brutia and Pinus nigra species are prevalent and pollen production can vary depending on factors such as plant taxa, pollen type, anthropogenic impact, geography, and climate [45,60,61] which can further affect the RPPs.
Fossil pollen studies conducted in southwestern Anatolia [2,14,62] reveal that the most common taxa observed include Cedrus, Olea, deciduous Quercus t., Phillyrea, Asteraceae, Centaurea, Artemisia, Cichoridae, Anthemideae, Brassicaceae, Chenopodiaceae, Cyperaceae, Apiaceae, Polygonum, Plantago, and Chenopodiaceae. Thus, there is still a need for new RPP values for those taxa.

5. Conclusions

We have estimated the first values of RSAP and RPP for the main plant taxa in the Mediterranean area of southwestern Turkey, based on ERV sub-model 2 and Prentice’s dispersal model. Pinus, evergreen Quercus t., Juniperus, Poaceae, and Fabaceae are the main pollen types in the study area. The results of ERV analysis suggest that RSAP in surface samples is ca. 102 m. The inconsistencies in RSAP results across different Mediterranean regions may be caused by the size of the vegetation patch and landscape openness. The RPP estimates obtained using ERV sub-model 2 analysis show that the highest producer among arboreal pollen taxa is Pinus (5.782) a wind-pollinated plant, and the lowest is Fabaceae (0.008), an insect-pollinated plants. The variations in RPP values can be explained by different sample materials, variations in vegetation composition, and vegetation landscape openness. The RPP values from the current study area are comparable to those with similar vegetation composition and structure, and climate settings from other Mediterranean regions. The differences observed might be attributed to various factors including landscape patchiness, species composition, climate., and land use. Methodological approaches such as vegetation surveys and sampling strategies may have also contributed to differences in RPPs between our study area and other regions in the Mediterranean.
The RPP results obtained in this study can help us to understand historical landscape changes in the southwestern part of Turkey, based on fossil pollen assemblages and pollen-based modelling.

Author Contributions

Conceptualization, E.E. and H.N.D.; methodology and analysis, E.E., L.M. and F.M.; writing—original draft preparation, E.E., L.M. and F.M.; writing—review and editing, E.E., L.M., F.M. and H.N.D.; visualization, E.E. and U.B. All authors have read and agreed to the published version of the manuscript.

Funding

This research received no external funding. This study is a part of the Ph.D. thesis of E.E. and was not funded as a project; however, during her studies, E.E. was awarded the 100/2000 doctoral scholarship (4 years) by the Council of Higher Education of Turkey (CoHE) and short-term granted (6 months) by the TUBITAK 2214-A (Scientific and Technological Research Council of Turkey).

Data Availability Statement

All data used in this study are open access except for the forest management plans. The original contributions presented in the study are included in the article, further inquiries can be directed to the corresponding author.

Acknowledgments

Special thanks to Demet Biltekin for her invaluable assistance and support with laboratory analyses during the pollen extraction process. I am also grateful to the members of ITU/EMCOL Lab for graciously allowing us to utilize their laboratory facilities. Special thanks to Stéphane Binet for providing R codes used in the calculation of land cover types. Furthermore, I wish to express my gratitude to the University of Innsbruck Department of Botany and the University of Toulouse-Jean Jaurès of Laboratoire Géographie de l’Environnement for their hospitality and for granting me access to their laboratory facilities during my research visit. I am also deeply thankful to Nazlı Tecimer for her invaluable editorial assistance.

Conflicts of Interest

The authors declare no conflicts of interest.

References

  1. Woodbridge, J.; Roberts, C.N.; Palmisano, A.; Bevan, A.; Shennan, S.; Fyfe, R.; Eastwood, W.J.; Izdebski, A.; Çakırlar, C.; Woldring, H.; et al. Pollen-Inferred Regional Vegetation Patterns and Demographic Change in Southern Anatolia through the Holocene. Holocene 2019, 29, 728–741. [Google Scholar] [CrossRef]
  2. Bottema, S.; Woldring, H. Late Quaternary Vegetation and Climate of Southwestern Turkey Part II. Palaeohistoria 1984, 26, 123–149. [Google Scholar]
  3. van Zeist, W.; Woldring, H.; Stapert, D. Late Quaternary Vegetation and Climate of Southwestern Turkey. Palaeohistoria 1975, 17, 55–143. [Google Scholar]
  4. Eastwood, W.J.; Roberts, N.; Lamb, H.F. Palaeoecological and Archaeological Evidence for Human Occupance in Southwest Turkey: The Beyşehir Occupation Phase. Anatollian Stududies 1998, 48, 69–86. [Google Scholar] [CrossRef]
  5. Kaniewski, D.; De Laet, V.; Paulissen, E.; Waelkens, M. Long-term Effects of Human Impact on Mountainous Ecosystems, Western Taurus Mountains, Turkey. J. Biogeogr. 2007, 34, 1975–1997. [Google Scholar] [CrossRef]
  6. Kaniewski, D.; Paulissen, E.; De Laet, V.; Dossche, K.; Waelkens, M. A High-Resolution Late Holocene Landscape Ecological History Inferred from an Intramontane Basin in the Western Taurus Mountains, Turkey. Quat. Sci. Rev. 2007, 26, 2201–2218. [Google Scholar] [CrossRef]
  7. Kaniewski, D.; Paulissen, E.; De Laet, V.; Waelkens, M. Late Holocene Fire Impact and Post-Fire Regeneration from the Bereket Basin, Taurus Mountains, Southwest Turkey. Quat. Res. 2008, 70, 228–239. [Google Scholar] [CrossRef]
  8. Karlıoğlu Kılıç, N.; Şenkul, Ç.; Memiş, T.; Doğan, M. Salurtepe Dağı (Elmalı-Antalya) Ardıç Ormanında Güncel Polen Dağılımının İncelenmesi. J. Geogr. 2019, 8, 11–22. [Google Scholar] [CrossRef]
  9. Şenkul, Ç. Abies cilicica Ormanının (Karlık Dağı/Burdur-Bucak) Güncel Polen Dağılımı: Polen Tuzakları, Kara Yosunu Örnekleri ve Yüzey Sediman Örneği Arasındaki İlişkiler. Süleyman Demirel Üniversitesi Fen-Edeb. Fakültesi Sos. Bilim. Dergisi. 2018, 1, 205–226. [Google Scholar]
  10. Şenkul, Ç.; Karlioğlu Kiliç, N.; Doğan, M.; Eastwood, W.J. Modern Pollen Distribution of the Teke Peninsula Forests: The Case of the Ördübek Highland. Eurasian J. For. Sci. 2018, 6, 58–75. [Google Scholar] [CrossRef]
  11. Vermoere, M.; Vanhecke, L.; Waelkens, M.; Smets, E. Modern Pollen Studies in the Territory of Sagalassos (Southwest Turkey) and Their Use in the Interpretation of a Late Holocene Pollen Diagram. Rev. Palaeobot. Palynol. 2001, 114, 29–56. [Google Scholar] [CrossRef]
  12. Woodbridge, J.; Roberts, N.; Fyfe, R. Pan-Mediterranean Holocene Vegetation and Land-cover Dynamics from Synthesized Pollen Data. J. Biogeogr. 2018, 45, 2159–2174. [Google Scholar] [CrossRef]
  13. Smith, T.J. Highland Gods: Rock-Cut Votive Reliefs from the Pisidian Survey. Anatol. Stud. 2011, 61, 133–150. [Google Scholar] [CrossRef]
  14. Eastwood, W.J.; Roberts, N.; Lamb, H.F.; Tibby, J.C. Holocene Environmental Change in Southwest Turkey: A Palaeoecological Record of Lake and Catchment-Related Changes. Quat. Sci. Rev. 1999, 18, 671–695. [Google Scholar] [CrossRef]
  15. Prentice, I.C.; Webb, T. Pollen Percentages, Tree Abundances and the Fagerlind Effect. J. Quat. Sci. 1986, 1, 35–43. [Google Scholar] [CrossRef]
  16. Sugita, S. Theory of Quantitative Reconstruction of Vegetation I: Pollen from Large Sites REVEALS Regional Vegetation Composition. Holocene 2007, 17, 229–241. [Google Scholar] [CrossRef]
  17. Sugita, S. Theory of Quantitative Reconstruction of Vegetation II: All You Need Is LOVE. Holocene 2007, 17, 243–257. [Google Scholar] [CrossRef]
  18. Sugita, S. Pollen Representation of Vegetation in Quaternary Sediments: Theory and Method in Patchy Vegetation. J. Ecol. 1994, 82, 881. [Google Scholar] [CrossRef]
  19. Gaillard, M.J.; Sugita, S.; Mazier, F.; Trondman, A.K.; Broström, A.; Hickler, T.; Kaplan, J.O.; Kjellström, E.; Kokfelt, U.; Kuneš, P.; et al. Holocene Land-Cover Reconstructions for Studies on Land Cover-Climate Feedbacks. Clim. Past 2010, 6, 483–499. [Google Scholar] [CrossRef]
  20. Marquer, L.; Gaillard, M.J.; Sugita, S.; Trondman, A.K.; Mazier, F.; Nielsen, A.B.; Fyfe, R.M.; Odgaard, B.V.; Alenius, T.; Birks, H.J.B.; et al. Holocene Changes in Vegetation Composition in Northern Europe: Why Quantitative Pollen-Based Vegetation Reconstructions Matter. Quat. Sci. Rev. 2014, 90, 199–216. [Google Scholar] [CrossRef]
  21. Trondman, A.K.; Gaillard, M.J.; Mazier, F.; Sugita, S.; Fyfe, R.; Nielsen, A.B.; Twiddle, C.; Barratt, P.; Birks, H.J.B.; Bjune, A.E.; et al. Pollen-Based Quantitative Reconstructions of Holocene Regional Vegetation Cover (Plant-Functional Types and Land-Cover Types) in Europe Suitable for Climate Modelling. Glob. Chang. Biol. 2015, 21, 676–697. [Google Scholar] [CrossRef]
  22. Antoine, E.; Marquer, L.; Muigg, B.; Tegel, W.; Bisson, U.; Bolliger, M.; Herzig, F.; Heussner, K.-U.; Hofmann, J.; Kontic, R.; et al. Legacy of Last Millennium Timber Use on Plant Cover in Central Europe: Insights from Tree Rings and Pollen. Sci. Total Environ. 2024, 922, 171157. [Google Scholar] [CrossRef] [PubMed]
  23. Githumbi, E.; Fyfe, R.; Gaillard, M.-J.; Trondman, A.-K.; Mazier, F.; Nielsen, A.-B.; Poska, A.; Sugita, S.; Woodbridge, J.; Azuara, J.; et al. European Pollen-Based REVEALS Land-Cover Reconstructions for the Holocene: Methodology, Mapping and Potentials. Earth Syst. Sci. Data 2022, 14, 1581–1619. [Google Scholar] [CrossRef]
  24. Serge, M.A.; Mazier, F.; Fyfe, R.; Gaillard, M.J.; Klein, T.; Lagnoux, A.; Galop, D.; Githumbi, E.; Mindrescu, M.; Nielsen, A.B.; et al. Testing the Effect of Relative Pollen Productivity on the REVEALS Model: A Validated Reconstruction of Europe-Wide Holocene Vegetation. Land 2023, 12, 986. [Google Scholar] [CrossRef]
  25. Grindean, R.; Nielsen, A.B.; Tanţău, I.; Feurdean, A. Relative Pollen Productivity Estimates in the Forest Steppe Landscape of Southeastern Romania. Rev. Palaeobot. Palynol. 2019, 264, 54–63. [Google Scholar] [CrossRef]
  26. Paul Ozenda Sur Les Étages de Végétation Dans Les Montagnes Du Bassin Méditerranéen. Available online: http://ecologie-alpine.ujf-grenoble.fr/item/95/ (accessed on 6 December 2023).
  27. Atalay, I.; Efe, R.; Öztürk, M. Effects of Topography and Climate on the Ecology of Taurus Mountains in the Mediterranean Region of Turkey. Procedia Soc. Behav. Sci. 2014, 120, 142–156. [Google Scholar] [CrossRef]
  28. Mazier, F.; Gaillard, M.-J.; Kuneš, P.; Sugita, S.; Trondman, A.-K.; Broström, A. Testing the Effect of Site Selection and Parameter Setting on REVEALS-Model Estimates of Plant Abundance Using the Czech Quaternary Palynological Database. Rev. Palaeobot. Palynol. 2012, 187, 38–49. [Google Scholar] [CrossRef]
  29. Broström, A.; Nielsen, A.B.; Gaillard, M.-J.; Hjelle, K.; Mazier, F.; Binney, H.; Bunting, J.; Fyfe, R.; Meltsov, V.; Poska, A.; et al. Pollen Productivity Estimates of Key European Plant Taxa for Quantitative Reconstruction of Past Vegetation: A Review. Veg. Hist. Archaeobotany 2008, 17, 461–478. [Google Scholar] [CrossRef]
  30. Bunting, M.J.; Hjelle, K.L. Effect of Vegetation Data Collection Strategies on Estimates of Relevant Source Area of Pollen (RSAP) and Relative Pollen Productivity Estimates (Relative PPE) for Non-Arboreal Taxa. Veg. Hist. Archaeobotany 2010, 19, 365–374. [Google Scholar] [CrossRef]
  31. Atalay, I. Vegetation BT—The Soils of Turkey; Kapur, S., Akça, E., Günal, H., Eds.; Springer International Publishing: Cham, Switzerland, 2018; pp. 15–24. ISBN 978-3-319-64392-2. [Google Scholar]
  32. Atalay, I. Vegetation Formations of Turkey. Trav. Inst. Geogr. Reims 1986, 65–66, 17–30. [Google Scholar] [CrossRef]
  33. Özçelik, H.; Çinbilgel, İ.; Belkıs, M.; Ahmet, K.; Tavuç, İ.; Bebekli, Ö. Burdur Ili Karasal ve Iç Su Ekosistem Çeşitliliği, Koruma ve Izleme Çalışmaları. Süleyman Demirel Üniversitesi Fen Edeb. Fakültesi Fen Derg. 2014, 9, 12–43. [Google Scholar]
  34. Ketenoğlu, A.; Vural, M.; Kurt, L.; Körüklü, S. Vejetasyon. In Resimli Türkiye Florası; Güner, A., ve Ekim, T., Vakfi, A.N.G., Dernegi, F.A., Eds.; Turkiye Is Bankasi Kultur Yayinlari: Istanbul, Turkey, 2014; Volume 1, pp. 172–182. ISBN 978-605-332-220-7. [Google Scholar]
  35. Atalay, I.; Altunbaş, S.; Khan, A.A.; Coşkun, M. The Mountain Ecology of the Taurus Mountains and Its Effects on Nomadism. In Proceedings of the International Geography Symposium on the 30th Anniversary of TUCAUM, Ankara, Turkey, 3–6 October 2018; pp. 623–640. [Google Scholar]
  36. Bunting, M.J.; Farrell, M.; Broström, A.; Hjelle, K.L.; Mazier, F.; Middleton, R.; Nielsen, A.B.; Rushton, E.; Shaw, H.; Twiddle, C.L. Palynological Perspectives on Vegetation Survey: A Critical Step for Model-Based Reconstruction of Quaternary Land Cover. Quat. Sci. Rev. 2013, 82, 41–55. [Google Scholar] [CrossRef]
  37. Bradshaw, R.H.W. Modern Pollen-Representation Factors for Woods in South-East England. J. Ecol. 1981, 69, 45. [Google Scholar] [CrossRef]
  38. Räsänen, S.; Hicks, S.; Odgaard, B.V. Pollen Deposition in Mosses and in a Modified ‘Tauber Trap’ from Hailuoto, Finland: What Exactly Do the Mosses Record? Rev. Palaeobot. Palynol. 2004, 129, 103–116. [Google Scholar] [CrossRef]
  39. Farrell, M.; Bunting, M.J.; Middleton, R. Replicability of Data Collected for Empirical Estimation of Relative Pollen Productivity. Rev. Palaeobot. Palynol. 2016, 232, 1–13. [Google Scholar] [CrossRef]
  40. Faegri, K.; Iversen, J. Textbook of Pollen Analysis. Kew Bull. 1976, 31, 192. [Google Scholar] [CrossRef]
  41. Moore, P.D.; Webb, J.A.C.M. Pollen Analysis; Blackwell: London, UK, 1991; pp. 1–216. [Google Scholar]
  42. Reille, M. Pollen et Spores d’Europe et d’Afrique Du Nord. France: Laboratoire de Botanique Historique et Palynologie, 2nd ed.; Boite 451: Marseille, France, 1999; pp. 1–535. [Google Scholar]
  43. Parsons, R.W.; Prentice, I.C. Statistical Approaches to R-Values and the Pollen—Vegetation Relationship. Rev. Palaeobot. Palynol. 1981, 32, 127–152. [Google Scholar] [CrossRef]
  44. Prentice, I.C. Pollen Representation, Source Area, and Basin Size: Toward a Unified Theory of Pollen Analysis. Quat. Res. 1985, 23, 76–86. [Google Scholar] [CrossRef]
  45. Kuparinen, A.; Markkanen, T.; Riikonen, H.; Vesala, T. Modeling Air-Mediated Dispersal of Spores, Pollen and Seeds in Forested Areas. Ecol. Model. 2007, 208, 177–188. [Google Scholar] [CrossRef]
  46. Borrell, J.S. Rapid Assessment Protocol for Pollen Settling Velocity: Implications for Habitat Fragmentation. Biosci. Horiz. 2012, 5, hzs002. [Google Scholar] [CrossRef]
  47. Gregory, P.H. The Microbiology of the Atmosphere; Leonard Hill: Aylesbury, UK, 1973. [Google Scholar]
  48. Li, F.; Gaillard, M.J.; Sugita, S.; Mazier, F.; Xu, Q.; Zhou, Z.; Zhang, Y.; Li, Y.; Laffly, D. Relative Pollen Productivity Estimates for Major Plant Taxa of Cultural Landscapes in Central Eastern China. Veg. Hist. Archaeobotany 2017, 26, 587–605. [Google Scholar] [CrossRef]
  49. Mazier, F.; Broström, A.; Gaillard, M.J.; Sugita, S.; Vittoz, P.; Buttler, A. Pollen Productivity Estimates and Relevant Source Area of Pollen for Selected Plant Taxa in a Pasture Woodland Landscape of the Jura Mountains (Switzerland). Veg. Hist. Archaeobotany 2008, 17, 479–495. [Google Scholar] [CrossRef]
  50. Bunting, M.J.; Schofield, J.E.; Edwards, K.J. Estimates of Relative Pollen Productivity (RPP) for Selected Taxa from Southern Greenland: A Pragmatic Solution. Rev. Palaeobot. Palynol. 2013, 190, 66–74. [Google Scholar] [CrossRef]
  51. Shumilovskikh, L.S.; Abdulmanova, I.F.; Efimik, E.G. Does Season Matter for Moss Surface Sample Collection? A Case Study from Kungur Forest-Steppe, Pre-Urals, Russia. Palynology 2021, 45, 191–199. [Google Scholar] [CrossRef]
  52. Vermoere, M.; Vanhecke, L.; Waelkens, M.; Smets, E. A Comparison between Modern Pollen Spectra of Moss Cushions and Cundill Pollen Traps: Implications for the Interpretation of Fossil Pollen Data from Southwest Turkey. Grana 2000, 39, 146–158. [Google Scholar] [CrossRef]
  53. Baker, A.G.; Zimny, M.; Keczyński, A.; Bhagwat, S.A.; Willis, K.J.; Latałowa, M. Pollen Productivity Estimates from Old-Growth Forest Strongly Differ from Those Obtained in Cultural Landscapes: Evidence from the Białowieża National Park, Poland. Holocene 2016, 26, 80–92. [Google Scholar] [CrossRef]
  54. Clymo, R.S.; Mackay, D. Upwash and Downwash of Pollen and Spores in the Unsaturated Surface Layer of Sphagnum-Dominated Peat. New Phytol. 1987, 105, 175–183. [Google Scholar] [CrossRef]
  55. Joosten, H.; De Klerk, P. In Search of Finiteness: The Limits of Fine-Resolution Palynology of Sphagnum Peat. Holocene 2007, 17, 1023–1031. [Google Scholar] [CrossRef]
  56. Sjögren, P.; Van Leeuwen, J.F.N.; Van Der Knaap, W.O.; Van Der Borg, K. The Effect of Climate Variability on Pollen Productivity, AD 1975-2000, Recorded in a Sphagnum Peat Hummock. Holocene 2006, 16, 277–286. [Google Scholar] [CrossRef]
  57. Broström, A.; Sugita, S.; Gaillard, M.J.; Pilesjö, P. Estimating the Spatial Scale of Pollen Dispersal in the Cultural Landscape of Southern Sweden. Holocene 2005, 15, 252–262. [Google Scholar] [CrossRef]
  58. Nielsen, A.B.; Sugita, S. Estimating Relevant Source Area of Pollen for Small Danish Lakes around AD 1800. Holocene 2005, 7, 1006–1020. [Google Scholar] [CrossRef]
  59. Twiddle, C.L. Pollen Analysis: Not Just a Qualitative Tool. In Geomorphological Techniques; British Society for Geomorphology: London, UK, 2012; Chapter 4, Section 1.4; pp. 1–11. [Google Scholar]
  60. Li, F.; Gaillard, M.J.; Xu, Q.; Bunting, M.J.; Li, Y.; Li, J.; Mu, H.; Lu, J.; Zhang, P.; Zhang, S.; et al. A Review of Relative Pollen Productivity Estimates from Temperate China for Pollen-Based Quantitative Reconstruction of Past Plant Cover. Front. Plant Sci. 2018, 9, 1214. [Google Scholar] [CrossRef] [PubMed]
  61. Von Stedingk, H.; Fyfe, R.M.; Allard, A. Pollen Productivity Estimates from the Forest-Tundra Ecotone in West-Central Sweden: Implications for Vegetation Reconstruction at the Limits of the Boreal Forest. Holocene 2008, 18, 323–332. [Google Scholar] [CrossRef]
  62. Eastwood, W.J.; Pearce, N.J.G.; Westgate, J.A.; Perkins, W.T. Recognition of Santorini (Minoan) Tephra in Lake Sediments from Gollhisar Golu, Southwest Turkey by Laser Ablation ICP-MS. J. Archaeol. Sci. 1998, 25, 677–687. [Google Scholar] [CrossRef]
Land 13 00591 g001
Figure 1. (a) Location of the study area and the schematic view of Oro-Mediterranean vegetation zone modified from Atalay I. (2014)) (b) Land cover types and distribution of the sampling sites within 50 km radius. (c) Land cover types in the study area, Quercus coccifera shrubland (left), Pinus nigra forest (right).
Figure 1. (a) Location of the study area and the schematic view of Oro-Mediterranean vegetation zone modified from Atalay I. (2014)) (b) Land cover types and distribution of the sampling sites within 50 km radius. (c) Land cover types in the study area, Quercus coccifera shrubland (left), Pinus nigra forest (right).
Land 13 00591 g001
Land 13 00591 g002
Figure 2. The climatic chart for the Gölhisar meteorological station, encompassing data from 2005 to 2022. Bars indicate precipitation levels in millimeters, while the lines in green, red, and orange represent the minimum, mean, and maximum temperatures in degrees Celsius, respectively.
Figure 2. The climatic chart for the Gölhisar meteorological station, encompassing data from 2005 to 2022. Bars indicate precipitation levels in millimeters, while the lines in green, red, and orange represent the minimum, mean, and maximum temperatures in degrees Celsius, respectively.
Land 13 00591 g002
Land 13 00591 g003
Figure 3. A visual representation of the vegetation survey conducted at each sample site, extending from the central point up to 5000 m. Original protocol can be found in Bunting et al. (2013).
Figure 3. A visual representation of the vegetation survey conducted at each sample site, extending from the central point up to 5000 m. Original protocol can be found in Bunting et al. (2013).
Land 13 00591 g003
Land 13 00591 g004
Figure 4. Pollen percentage diagram of the major pollen taxa found in the 21 moss pollster samples.
Figure 4. Pollen percentage diagram of the major pollen taxa found in the 21 moss pollster samples.
Land 13 00591 g004
Land 13 00591 g005
Figure 5. Land cover type categories percentages within a 5 km radius diagram in the 21 sampling sites.
Figure 5. Land cover type categories percentages within a 5 km radius diagram in the 21 sampling sites.
Land 13 00591 g005
Land 13 00591 g006aLand 13 00591 g006b
Figure 6. Plots displaying log-likelihood values as a function of distance are shown for a total of nine combinations, which involve three ERV sub-models and three distinct distance-weighting methods. These graphs are based on the analysis of pollen data obtained from moss samples collected at 21 randomly selected sites, coupled with the corresponding vegetation data encompassing a radius of 5000 m around each sampling point: (a) represents all nine model combinations, (b) illustrates the outcomes when ERV sub-model 2 is applied with the three available distance-weighting methods.
Figure 6. Plots displaying log-likelihood values as a function of distance are shown for a total of nine combinations, which involve three ERV sub-models and three distinct distance-weighting methods. These graphs are based on the analysis of pollen data obtained from moss samples collected at 21 randomly selected sites, coupled with the corresponding vegetation data encompassing a radius of 5000 m around each sampling point: (a) represents all nine model combinations, (b) illustrates the outcomes when ERV sub-model 2 is applied with the three available distance-weighting methods.
Land 13 00591 g006aLand 13 00591 g006b
Land 13 00591 g007
Figure 7. Scatter plots of the pollen–vegetation relationships at the distance of the relevant source area of pollen (RSAP = 102 m) as estimated using the ERV sub-model 2 and Prentice’s vegetation distance–weighting method: (a) original pollen proportion and vegetation absolute abundance, (b) relative pollen loading and absolute distance-weighted vegetation abundance.
Figure 7. Scatter plots of the pollen–vegetation relationships at the distance of the relevant source area of pollen (RSAP = 102 m) as estimated using the ERV sub-model 2 and Prentice’s vegetation distance–weighting method: (a) original pollen proportion and vegetation absolute abundance, (b) relative pollen loading and absolute distance-weighted vegetation abundance.
Land 13 00591 g007
Land 13 00591 g008
Figure 8. Relative pollen productivity estimates (RPPEs) along with their standard deviations are provided for five taxa found in the Mediterranean woodland ecosystem. Quercus coccifera is used as the reference with an RPP value set to 1. The RPP values for the other taxa are expressed in relation to Quercus coccifera. These RPP estimates are derived using ERV sub-models 1, 2, and 3, employing Prentice’s model, LSM, and the inverse distance (1/d) distance-weighting function.
Figure 8. Relative pollen productivity estimates (RPPEs) along with their standard deviations are provided for five taxa found in the Mediterranean woodland ecosystem. Quercus coccifera is used as the reference with an RPP value set to 1. The RPP values for the other taxa are expressed in relation to Quercus coccifera. These RPP estimates are derived using ERV sub-models 1, 2, and 3, employing Prentice’s model, LSM, and the inverse distance (1/d) distance-weighting function.
Land 13 00591 g008
Table 1. Location of the 21 sites and the major vegetation communities surveyed in the field within a 100 m radius area of the sampling sites. GPS coordinates are expressed as decimal degrees (DD).
Table 1. Location of the 21 sites and the major vegetation communities surveyed in the field within a 100 m radius area of the sampling sites. GPS coordinates are expressed as decimal degrees (DD).
Site NameLatitude (°)Longitude (°)Elevation (m)Major Vegetation Communities within 100 m
237.21126729.3220831071Juniperus excelsa-Quercus coccifera forest.
337.2417529.6065671089Pinus nigra mixed forest with Quercus coccifera and Juniperus oxcycedrus shrubs
637.08578329.3940331296Quercus coccifera vegetation with Juniperus oxycedrus and openland area with Astragalus and other herbs
937.12018329.7013171072Quercus coccifera shrubland with Juniperus oxycedrus and rocky area with herbaceous plants
1336.72198329.4128668Mediterranean Pinus brutia mixed forest with Quercus coccifera, Olea europea, and Phillyrea latifolia
1436.8764529.7069671666Juniperus foetidissima semi-forest and rocky area with herbaceous plants
1637.05008329.261051278Pinus nigra forest
1936.9775529.423751226Pinus nigra mixed forest with Quercus coccifera and Juniperus oxcycedrus shrubs
2236.95816729.9896831773High mountain steppe
2336.917729.5809331842Cedrus libani and Juniperus foetidissima semi-forest–High Mountain steppe
2936.946229.6396171412Pinus nigra-Juniperus oxycedrus semi forest
3137.11541729.269751330Pinus nigra-Juniperus oxycedrus forest
3237.47598329.6805671718Pinus nigra forest
3336.82646729.63751315Quercus coccifera vegetation with Juniperus foetidissima
3436.85526729.8981331468Steppe area with a few Juniperus and Quercus coccifera
3537.1158529.36251822Pinus nigra—Juniperus foetidissima semi-forest–High Mountain steppe
4137.51146729.6717671251Pinus nigra-Quercus pubescens mixed forest
4537.069129.467751183Pinus nigra mixed forest with Platanus orientalis
4736.96718329.2706331220Pinus nigra mixed forest with Quercus coccifera and Juniperus oxcycedrus shrubs
4936.9430529.8224331788High mountain steppe
5036.84228329.2821671105Pinus brutia mixed forest with deciduous Quercus
Table 2. Plant composition of the land cover types. Percentages are based on vegetation surveys and literature.
Table 2. Plant composition of the land cover types. Percentages are based on vegetation surveys and literature.
Land Cover TypesArable LandCedrus libani-Juniperus sp. ForestPinus brutia ForestPinus sp. ForestJuniperus sp. ForestQuercus coccifera MaquisPinus sp.—Deciduous ForestPinus sp.—Juniperus sp. ForestQuercus Deciduous ForestMaquisSteppeTransitional WoodlandSettlementBare Soil-Bare RockMoorWater
Apiaceae1
Asparagus 1
Asteraceae1 11 1112 1
Berberis 1 1
Boraginaceae 1 1
Brassicaceae 1 1 1 1 1
Caprifoliaceae 1
Cyperaceae 9 1 1 1
Caryophyllaceae 22 221 1
Cedrus 34 1
Chenopodiaceae 1 11 1
Cistaceae 1 1 1
Cornaceae 1
Cotinus 12
Crassula 1
Cupressaceae 2604302014775 2
Daphne 5
Ephedra 1
Ericaceae 5
Euphorbia 1
Fabaceae 7 1125 3 101512 1
Jasminum 0 1 2
Lamiaceae 1 542 3 441 1
Olea 1 5
Phillyrea 1 5
Pinus 8853 63366 1
Plumbaginaceae 2 2 1
Poaceae975161417571312155912
Potentilla 1
Decidious Quercus t.3 5 55 3
Evergreen Quercus t. 114 37 302
Rhamnus 1 1
Rosaceae4 1 1314
Rubiaceae 2
Salix 1
Sanguisorba 11
Saxifraga 1
Smilax 1
Styrax 3 11
Verbascum 1
Table 3. The plant species included in the vegetation survey were matched to pollen morphological categories and the fall speedestimates. The Olea FSP estimates was calculated from the reference taxon collection presented by Istanbul University for this study, the Cedrus estimate is from Borrel (2012) [46], and the other values are from Githumbi et al. (2022) [23].
Table 3. The plant species included in the vegetation survey were matched to pollen morphological categories and the fall speedestimates. The Olea FSP estimates was calculated from the reference taxon collection presented by Istanbul University for this study, the Cedrus estimate is from Borrel (2012) [46], and the other values are from Githumbi et al. (2022) [23].
Pollen Morphological TypeCorresponding Plant Taxa in VegetationFall Speed (m/s)
AsteraceaeCentaurea, Anthemis, Carduus, 0.051
Amaranthaceae/ChenopodiaceaeAtriplex0.019
CedrusCedrus libani0.097
JuniperusJuniperus excelsa, J. foetidissima, J. oxycedrus0.016
FabaceaeAstragalus, Calicotome, Dorycnium, Trifolium0.021
OleaOlea europaea0.011
PhillyreaPhillyrea latifolia0.015
PinusPinus brutia, Pinus nigra0.031
PoaceaePoaceae0.035
Deciduous Quercus tQuercus cerris, Q. ithaburensis, Q. pubescens0.035
Evergreen Quercus t.Quercus coccifera0.016
Table 4. Predictions of the RPP estimates with standard error and relevant source area of pollen (RSAP) in meters. These predictions are derived by exploring different combinations of ERV sub-models and distance-weighting methodologies, i.e., GPM (Prentice’s taxon-specific model), the inverse distance (1/d), and the Lagrangian stochastic model (LSM).
Table 4. Predictions of the RPP estimates with standard error and relevant source area of pollen (RSAP) in meters. These predictions are derived by exploring different combinations of ERV sub-models and distance-weighting methodologies, i.e., GPM (Prentice’s taxon-specific model), the inverse distance (1/d), and the Lagrangian stochastic model (LSM).
Sub-Model 1Sub-Model 2Sub-Model 3Sub-Model 1Sub-Model 2Sub-Model 3Sub-Model 1Sub-Model 2Sub-Model 3
Dispersal
Model		GPM (Prentice)GPM (Prentice)GPM (Prentice)1D1D1DLSMLSMLSM
RSAP18210266184106671455957
Taxonalphas.d.alphas.d.alphas.d.alphas.d.alphas.d.alphas.d.alphas.d.alphas.d.alphas.d.
Q.coccifera101010101010101010
Juniperus0.1120.0000.2790.0010.2530.0010.1200.0000.3010.0010.2750.0010.0953302.440.2577695.080.2390.001
Fabaceae0.0050.0000.0080.0000.0040.0000.0050.0000.0080.0000.0040.0000.0070.0000.0100.0000.0060.000
Pinus1.4240.0015.7820.0115.5890.0071.4380.0015.9190.0135.7820.0071.4500.0006.0660.0036.2550.009
Poaceae0.0600.0000.1120.0010.0570.0000.0590.0000.1160.0010.0610.0000.0590.0000.1190.0000.0840.000
Table 5. The RPP values were recalibrated using Poaceae as a reference taxon to enable comparison with previous studies. The Mediterranean region is a combination of the results obtained by the three local studies in the region. The average RPP values for the Mediterranean region used software developed by Sugita (unpublished), to calculate Standard Errors based on the Delta Method.
Table 5. The RPP values were recalibrated using Poaceae as a reference taxon to enable comparison with previous studies. The Mediterranean region is a combination of the results obtained by the three local studies in the region. The average RPP values for the Mediterranean region used software developed by Sugita (unpublished), to calculate Standard Errors based on the Delta Method.
RegionFrance, Medit. (ERV3)Romania (ERV3)Turkey, Medit. (ERV2)Mediterranean Region
Study referenceGithumbi et al. (2022) [23]Grindean et al. (2019) [25]This paperThis paper
RPPSDFSPRPPSDFSPRPPSDFSPRPPSDFSP
Herb taxa
Poaceae (reference taxon)100.035100.0351.000.000.041.0000.035
Apiaceae 5.911.230.042 5.911.230.042
Artemisia 5.893.160.014 5.893.160.014
Compositae (Asteroideae + Cichorioideae) 0.160.10.029 0.160.10.029
Comp. SF Cichorioideae1.1620.6750.061 1.1620.6750.061
Cerealia (Cerealia t. + Triticum t. + Secale + Zea) 0.220.120.06 0.220.120.06
Fabaceae 0.40.070.0210.0730.000.0210.160.0230.021
Plantago lanceolata 0.580.320.029 0.580.320.029
Ranunculaceae2.0380.3350.02 2.0380.3350.02
Rosaceae (Filipendula, Potentilla t., Sanguisorba) 0.290.120.018 0.290.120.018
Rubiaceae 0.40.070.019 0.40.070.019
Tree/shrub taxa
Buxus sempervirens1.890.0680.032 1.890.0680.032
Carpinus orientalis 0.240.070.042 0.240.070.042
Castanea sativa3.2580.0590.01 3.2580.0590.01
Corylus avellana3.440.890.025 3.440.890.025
Cupressaceae (Juniperus communis, J. phoenicea, J. oxycedrus)1.6180.1610.02 2.520.000.021.370.0540.02
Ericaceae (Arbutus unedo, Erica arborea, E. cinerea, E. multiflora)4.2650.0940.051 4.2650.0940.051
Fraxinus (F. excelsior, F. ornus) 2.990.880.022 2.990.880.022
Phillyrea0.5120.0760.015 0.5120.0760.015
Pistacia0.7550.2010.03 0.7550.2010.03
Evergreen Quercus t. (Q. ilex, Q. coccifera)11.0430.2610.015 8.970.000.0156.670.0870.015
Deciduous Quercus t. (Q. spp., Q. peduncularis dominant) 1.10.350.035 1.10.350.035
Pinus (P. sylvestris, P. brutia, P. nigra)6.0580.2370.031 51.850.010.03119.30.0790.031
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Ergin, E.; Marquer, L.; Mazier, F.; Bisson, U.; Dalfes, H.N. Relative Pollen Productivity Estimates for Mediterranean Plant Taxa: A New Study Region in Turkey. Land 2024, 13, 591. https://doi.org/10.3390/land13050591

AMA Style

Ergin E, Marquer L, Mazier F, Bisson U, Dalfes HN. Relative Pollen Productivity Estimates for Mediterranean Plant Taxa: A New Study Region in Turkey. Land. 2024; 13(5):591. https://doi.org/10.3390/land13050591

Chicago/Turabian Style

Ergin, Esra, Laurent Marquer, Florence Mazier, Ugo Bisson, and Hasan Nüzhet Dalfes. 2024. "Relative Pollen Productivity Estimates for Mediterranean Plant Taxa: A New Study Region in Turkey" Land 13, no. 5: 591. https://doi.org/10.3390/land13050591

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop