Next Article in Journal
Temporal and Spatial Variations in Landscape Habitat Quality under Multiple Land-Use/Land-Cover Scenarios Based on the PLUS-InVEST Model in the Yangtze River Basin, China
Previous Article in Journal
The Effects of Land Reallocation on Irrigation Collective Action: Moderating Effects of Informal Organizations and Leadership
Previous Article in Special Issue
Testing the Effect of Relative Pollen Productivity on the REVEALS Model: A Validated Reconstruction of Europe-Wide Holocene Vegetation
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Land
Volume 12
Issue 7
10.3390/land12071337
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 Major Plant Taxa in Middle Subtropical China

by
Qiuchi Wan
1,
Kangyou Huang
1,2,*,
Cong Chen
1,
Yongjie Tang
1,
Xiao Zhang
1,
Zhong Zhang
3 and
Zhuo Zheng
1,2
1
Guangdong Provincial Key Laboratory of Geodynamics and Geohazards, School of Earth Sciences and Engineering, Sun Yat-sen University, Zhuhai 519082, China
2
Southern Marine Science and Engineering Guangdong Laboratory (Zhuhai), Zhuhai 519082, China
3
Changguling Forest Farm, Jinggangshan National Nature Reserve, Jinggangshan 343600, China
*
Author to whom correspondence should be addressed.
Land 2023, 12(7), 1337; https://doi.org/10.3390/land12071337
Submission received: 15 May 2023 / Revised: 24 June 2023 / Accepted: 27 June 2023 / Published: 3 July 2023
(This article belongs to the Special Issue Pollen-Based Reconstruction of Holocene Land-Cover)
Browse Figures
Versions Notes

Abstract

:
Relative pollen productivity (RPP) is a key parameter for quantitative reconstruction of past vegetation cover. However, RPP estimates are rarely obtained in the subtropical and tropical regions. In this study, the extended R-value (ERV) model was used to estimate RPP for major plant taxa in the evergreen broadleaved and mixed forests in middle subtropical China based on soil samples and vegetation data from 23 sites. The best result was obtained with the combinations of ERV sub-model 3 and Prentice’s or 1/d vegetation distance-weighting method. The relevant source area of pollen (RSAP) of the soil samples was estimated to be ca. 500 m. RPP on the basis of ERV sub-model 3 and Prentice’s model was obtained for seven taxa and the ranking is as follows: Castanopsis (1 ± 0), Ilex (0.352 ± 0.031), Mallotus (0.221 ± 0.028), Liquidambar (0.115 ± 0.007), Cyclobalanopsis (0.107 ± 0.006), Camelia (0.033 ± 0.001), Symplocos (0.010 ± 0.002). RPPs for Cyclobalanopsis, Camelia, Ilex, and Symplocos which are dominant elements in the subtropical evergreen broadleaved forests were first obtained. Our result demonstrates a significant effect of pollen dispersal models on the estimates of RPPs. The RPPs obtained in this study provide an important basis for quantitative vegetation reconstruction in the subtropical region of China.

1. Introduction

The current state of the biosphere is not only affected by ongoing climate change but also by the history of (anthropogenic) land-cover change [1,2] and feedback between land-cover and the climate system [3]. Therefore, it is crucial to reconstruct past land-cover change and investigate its relationship with human activities and the climate to anticipate the impact of current and future anthropogenic climate change on the environment. Pollen analysis is one of the most important methods for past vegetation reconstruction. Several pollen-based methods have been developed including, e.g., modern analogue [4], and biomization [5]. The landscape reconstruction algorithm (LRA) is a theory-based approach to the quantitative reconstruction of land-cover changes. It includes two models, the REVEALS (Regional Estimates of VEgetation Abundance from Large Sites) model to estimate regional vegetation composition [6] and the LOVE (LOcal Vegetation Estimates) model to estimate local vegetation composition within the relevant source area of pollen (RASP) [7]. The LRA relies on one important parameter, relative pollen productivity (RPP), which can be estimated with the extended R-value (ERV) models [8,9,10]. The ERV models can reduce the effects of the non-linear nature of the pollen percentage-vegetation relationships and make it possible to objectively estimate pollen productivity. Considering that biases in pollen production and dispersal for a taxon are the main factors controlling pollen representation of plant abundance, the ERV models have a major assumption on pollen dispersal that the wind above the vegetation canopy is the dominant agent for pollen transport. The RPP has now been estimated for a wide range of plants in temperate regions where wind pollination is more common than in the tropics [11,12,13,14]. This has aided large-scale quantitative estimates of Holocene land-cover changes which in turn provides important information to assess the interactions between land-cover changes and climate in the past (e.g., [15,16]). However, attempts to estimate RPPs in tropical and subtropical regions of the world are rare, including only several studies from China and South Africa [17,18,19,20,21,22]. The difficulty of such attempts includes a wide variety of pollen productivity and dispersal mechanisms of plant species in the tropics with extremely high biodiversity [17].
Here we represent RPP results from a middle subtropical region of China, which is dominated by zonal evergreen broadleaved forests [23]. The available RPPs obtained in subtropical China focused on deciduous species [20,21]. Thus, our study is an important supplement to the estimates of RPPs for major evergreen plant taxa. This would enable an accurate quantitative reconstruction of the evergreen plant covers in addition to the deciduous plant covers, contributing to a comprehensive understanding of the past vegetation succession in subtropical China which is important for the study of the origin of human rice agriculture and land-cover-human-climate interactions. By comparing the results based on several combinations of models, we also highlight the impact of pollen dispersal models on the estimates of RPPs. In view of a majority of studies relying on Prentice’s model to estimate RPPs, we suggest validating the different RPP datasets produced with different dispersal models to obtain the most reliable estimates.

2. Study Area

Our study area is located within the Jinggangshan National Natural Reserve (Figure 1). The Jinggangshan Mountain region belongs to the middle subtropical climatic zone. The regional mean annual temperature is approximately 14 °C, and the mean annual precipitation is approximately 1800 mm [24]. The zonal vegetation in the Jinggangshan Mountains is a subtropical evergreen broadleaved forest. The vegetation composition in altitude is as follows [23]:
(1)
Evergreen broadleaved forest (<1000 m) dominated by Castanopsis sclerophylla (Lindl. et Paxton) Schottky, Castanopsis concinna (Champ. ex Benth.) A. DC., Castanopsis faberi Hance, Castanopsis tibetana Hance, Machilus thunbergii Sieb. et Zucc., Phoebe hunanensis Hand.-Mazz., and Elacocarpus japonicus Sieb.et Zucc.
(2)
Mixed evergreen and broadleaved forest (1000–1400 m) composed of deciduous elements including Castanopsis eyrei (Champ. ex Benth.) Tutch., Schima superba Gardn. et Champ., and Fagus lucida Rehd. et Wils.
(3)
Shrub-meadow (>1400 m) on the mountain summits dominated by Rhododendron simiarum Hance, Pieris formosa (Wall.) D. Don, and Enkianthus quinqueflorus Lour.
Due to human disturbance, the present-day vegetation in the mountainside is composed of many secondary forest tree species such as Pinus massoniana Lamb., as well as Phyllostachys edulis (Carrière) J. Houz. (bamboo) [25]. The lowlands are covered largely by rice paddy.
Land 12 01337 g001 550
Figure 1. Land cover map of the study area (modified from [26]) and the distribution of the 32 studied modern sites (white dots). Other RPP studies from subtropical China are also shown in the inset: 1, Daba Mountains [21]; 2, Ta-pieh Mountains [20]; 3, Hainan cultural landscape [22]. The vegetation zones in the inset are illustrated by different colors and numbers: I, temperate deciduous forest zone; II, northern subtropical evergreen broadleaved forest zone (EVBF); III, middle EVBF; IV, southern EVBF; V, tropical seasonal rainforest zone.
Figure 1. Land cover map of the study area (modified from [26]) and the distribution of the 32 studied modern sites (white dots). Other RPP studies from subtropical China are also shown in the inset: 1, Daba Mountains [21]; 2, Ta-pieh Mountains [20]; 3, Hainan cultural landscape [22]. The vegetation zones in the inset are illustrated by different colors and numbers: I, temperate deciduous forest zone; II, northern subtropical evergreen broadleaved forest zone (EVBF); III, middle EVBF; IV, southern EVBF; V, tropical seasonal rainforest zone.
Land 12 01337 g001

3. Materials and Methods

3.1. Field Work

The field work was undertaken in August 2019 within a 50 km × 50 km area. A total of 32 sites were randomly selected with a distance of at least 3000 m between them to avoid auto-correlation (Figure 1). Next, 32 surface soil samples were collected from the top ~2 cm of the soil layer and consisted of several random sub-samples within an area of a 0.5 m radius at each of the sampling points.
Following the standard protocol of [27], we conducted a vegetation survey within a 100 m radius of the sampling points for further RPP estimates using the ERV model. The major vegetation communities recognized in the field are listed in Table 1.

3.2. Pollen Analysis

Pollen samples were processed following the standard procedures of [28]. One piece of marker tablet with exotic Lycopodium (27,560 spores per piece) was added to each sample for calculation of pollen concentrations. Samples were treated with 10% HCl for carbonate removal and 10% KOH for humic acid removal. Coarse soil particles and other plant fragments were removed with a 0.125 mm mesh sieve. Heavy ZnBr2 liquid, 1.9–2.0 g/cm3, was used for gravity separation and the process was undertaken twice. Residues were mounted in glycerol on glass slides and sealed with Canada Balsam. Pollen grains were identified under 400× magnification with a Nikon and ZEISS light microscope and under 1000× magnification for more detailed examination. A minimum of 500 pollen grains were counted in each sample. Several references were used for pollen identification including the Angiosperm Pollen Flora of Tropic and Subtropic China [29], and the pollen floras by [30,31].
A minimum of 30 grains of each of the main pollen taxa identified in this study were measured on reference slides for their diameter (round grains) or their long and short axes. These measurements were used to calculate the fall speed of pollen (FSP) based on Stoke’s Law of particle settling velocity [32] (Table 2). FSP is used for the application of dispersion models used in ERV modeling.

3.3. Vegetation Analysis

ERV modeling requires vegetation data in concentric rings from the sampling sites to the maximum distance of the vegetation survey. For each concentric ring, the mean absolute cover (m2/m2) of the plant taxa was calculated based on land-cover types (vegetation communities) and taxa composition between 10 m and 1500 m at the sampling sites. Such data are available for a surveyed 100 m-radius area of every sampling point. Satellite images within a 1500 m radius from the sampling sites were downloaded from 91WeMap for each study site. Land-cover maps between 100 m and 1500 m at the study sites were then generated based on the satellite images and visual interpretation in ArcMap 10.2. The taxa composition of each land-cover type was assumed to be similar to the vegetation community surveyed in the field within 100 m at each of the study sites.
The distance increment from the sampling sites to 1500 m was set as follows: 0–0.5 m, 0.5–1.5 m, 1.5–3 m, 3–6 m, 6–10 m, and 100 m increments between 10 and 1500 m.

3.4. ERV Modeling

We used the ERV model as implemented in the ERV-Analysis software v 2.5.2 (Sugita, unpublished program) to calculate the RPPs of plant taxa with their standard errors (SEs) and the RSAP of the pollen sample sites. A summary of the theory behind the ERV model can be found in, e.g., [33]. Three ERV sub-models were developed which differ in the definition of the background pollen in the model equation relating pollen loading with vegetation [8]. The ERV sub-models 1 and 2 assume and estimate a species-specific constant background in pollen proportion and in the proportion of pollen loading to total plant abundance of all the taxa involved, respectively [8]. In addition, the relative cover (in percentage or proportion) of each plant taxon (harmonized with a pollen morphological type) is used in both sub-models. The ERV sub-model 3 does not have any additional assumption on the background pollen and utilizes the absolute cover of each plant taxon (in m2/m2). The three sub-models generally provide similar results. However, many cases have shown that one sub-model is more appropriate than the others, e.g., [34,35]. It might reflect that the vegetation data violate the assumption of the sub-models. In theory, pollen loading is linearly related to the distance-weighted plant abundance for each taxon [10]. Several methods have been developed to calculate the distance-weighted plant abundance, for example, the inverse distance (1/d), Prentice’s model [36], and the Lagrangian stochastic dispersal model (LSM) [37]. The LSM was suggested to be more powerful for the description of pollen dispersal when entomophilous plants dominate the vegetation [38]. However, it was not the case when the LSM was applied in the tropical Hainan Island [22]. Generally, different combinations of ERV sub-models and distance-weighting methods are used and their results are compared in one study to obtain the most suitable combination and most reliable estimates of the RPP dataset and RSAP.
For running the ERV model, plant taxa were first harmonized with pollen types (Table 2). The plant/pollen taxa were then selected for RPP estimates. The basic criterion for the selection is that the vegetation and pollen data exhibit a good spread of values from low to high in one taxon. A reference taxon (RPP = 1) was used to calculate the RPPs of all other taxa. Poaceae is commonly used as a reference taxon as it occurs frequently in the vegetation and pollen in temperate regions (e.g., [11,33,39]). In our study region, Poaceae occurs in different forms and is not appropriate for ERV modeling (see discussion). We finally chose Castanopsis as the reference taxon because it is abundant in both pollen and vegetation data.
Three datasets are required for ERV modeling, i.e., pollen counts, fall speed of pollen, and vegetation data. Apart from these, several parameters were set as follows: basin size (0.5 m), and wind speed (3 m/s). The ERV model was first run with the selected taxa (Table 2) and 32 sites. The ERV-adjusted pollen values and distance-weighted vegetation cover were plotted to select the taxa with the most linear pollen-vegetation relationships and remove some outlier sites. This procedure was repeated several times with a reduced number in taxa and sites (i.e., seven taxa and 23 sites in the final run) until the log-likelihood curve reached the theoretically best shape, i.e., increased with distance and reached an asymptote within the area of vegetation survey (1500 m).

4. Results and Discussion

4.1. Modern Vegetation and Surface Pollen Assemblages

The 32 sites are classified as evergreen woodland (EW) and mixed woodland (MW) (Figure 2), based on surveyed plant taxa from 93 families and 219 genera. The principal arboreal taxa in EW are Castanopsis, Cyclobalanopsis, Lauraceae Juss. (comprising mainly Machilus, Cinnamomum, and Sassafras), and Theaceae Mirb. (e.g., Eurya, Schima, and Camellia). The vegetation in MW comprises a mix of evergreen taxa similar to EW and coniferous and deciduous broadleaved trees and shrubs, such as Taxodiaceae Warming (comprising mainly Cunninghamia), Pinus, Liquidambar, Ericaceae Juss. (e.g., Rhododendron), Acer, and Fagus. Bamboo (e.g., Phyllostachys Siebold and Zucc., Indocalamus Nakai) also frequently occurs in these sites with high coverage. Herbs including mainly Poaceae are not abundant in our studied sites.
A total of 83 pollen types were identified at the genus-to-family levels in the 32 surface soil samples. Dominant pollen taxa in EW sites include arboreal types Castanopsis-Lithocarpus, Cyclobalanopsis, Camelia, Pinus, Ilex, Eurya, and non-arboreal Poaceae (Figure 2). Pollen assemblages of MW sites are similar to EW sites with differences in higher percentages of Pinus and Liquidambar.
Comparing percentages of vegetation cover and pollen, most evergreen broadleaved taxa have good pollen representation (e.g., Castanopsis, Cyclobalanopsis, Camelia, and Ilex); whereas, few or no pollen were found in the Lauraceae community. Deciduous taxa tend to be under-represented in surface pollen (Liquidambar as an exception). The coniferous tree Pinus is over-represented in surface pollen while Taxodiaceae is significantly under-represented. Considering this, seven taxa (Table 1) that have good spread values in both vegetation cover and pollen were selected for the final ERV analysis: Castanopsis, Liquidambar, Cyclobalanopsis, Camelia, Ilex, Mallotus, and Symplocos.

4.2. Relevant Source Area of Pollen and Affecting Factors

Figure 3 shows the log-likelihood curves of the nine combinations of ERV sub-models and distance-weighting models. For a given distance-weighting model, ERV sub-model 1 tends to have the poorest performance in terms of the lowest log-likelihood values among the three ERV sub-models and the increasing trend of the log-likelihood curve in comparison with an expected pattern, i.e., gradually increasing to reach an asymptote. The poor performance of ERV sub-model 1 is expected because ERV sub-model 1 assumes that the background pollen loading is a constant proportion of the total pollen loading [8]. Such assumption can be easily violated by variations in among-site vegetation composition and among-taxa pollen productivity as commonly observed in previous studies (e.g., [40]). ERV sub-models 2 and 3 have similar equations to linearize the pollen-vegetation relationship, but they differ in the way of expressing the vegetation (as percentages and projection area per unit area utilized by ERV sub-models 2 and 3, respectively) [10]. When non-pollen-producing areas (rocks, roads, water, etc.) are not included in the analysis, the input vegetation data are the same in both sub-models. As our study area is a fine mosaic of forests and croplands, there is a negligible proportion of non-pollen-producing areas. This could explain the similar log-likelihood values of ERV sub-models 2 and 3 with the latter being closer to an expected pattern of the log-likelihood curve. The combinations of ERV sub-model 3 with Prentice’s and 1/d vegetation distance-weighting methods show very similar values and trends of the log-likelihood. The combination of ERV sub-model 3 and LSM distance-weighting produces the highest log-likelihood values with a high variation. In general, ERV sub-model 3 with Prentice’s or 1/d vegetation distance-weighting method produces the most expected log-likelihood curve (i.e., gradually increasing to reach an asymptote), and thus is considered the best result in this study.
For the best combination of models (ERV sub-model 3 with Prentice’s or 1/d vegetation distance weighting method), the log-likelihood curve reaches an asymptote with relatively constant values beyond ca. 500 m (Figure 3c). Thus, the estimated RSAP is 500 m. This value is slightly higher than other values estimated in a subtropical forest landscape in China (e.g., 340 m in Ta-pieh Mountains [20], 360 m in Daba Mountains [21]), but the same as the value estimated in the tropical Hainan Forest landscape (i.e., 500 m [22]). RSAP might be influenced by many factors such as size and type of sediment basin and vegetation structure (e.g., patch size, the spatial distribution of patches, and plant taxa in the landscape). Previous studies have demonstrated that when the basin size was constant, the size of the vegetation patch would have a very significant effect on the RSAP [33]. In this study, the land-cover maps were created based on visual interpretation instead of the typically used supervision and non-supervision classification methods. As a result, the resolution of our land-cover maps is lower compared to previous studies. Thus, we expected a higher RSAP than previous studies because of the large vegetation patches being extracted. In addition, the pollen samples were collected from surface soil in both Hainan [22] and this study, while moss polster samples were studied in the two mentioned subtropical works [20,21]. It is suggested that soil pollen assemblages contain a larger background pollen component than the moss assemblages [41]. This could result in a higher RSAP when estimated with soil pollen samples compared to moss samples. Therefore, the higher RSAP estimate in this study might be influenced by both the resolution of the land cover maps and the basin type. It would be worth testing the contributions of the two factors to the estimation of RSAP.

4.3. Pollen-Vegetation Relationships

The scatter plots of ERV-adjusted pollen and vegetation values using ERV sub-model 3 and Prentice’s vegetation distance-weighting method at the RSAP of 500 m are shown in Figure 4. The pollen-vegetation relationship of Castanopsis shows a very good linear trend. For Cyclobalanopsis, Liquidambar, and Ilex, the spread of pollen and vegetation values are more or less close to a linear trend. Camelia and Symplocos exhibit a large spread of values with high plant cover that may correspond to low pollen value while low plant cover corresponds to high pollen value. There is a single high value in the pollen and vegetation data of Mallotus. In most cases, extremely low values of plant covers correspond to a range of pollen loading. Similar findings were obtained in a tropical study which is ascribed to entomophily for Mallotus [22]. This could also explain the poor relationships between pollen and vegetation observed in Camelia and Symplocos in this study. ERV model assumes pollen anemophily [8,9,10]. The violation of the basic assumption of the model already indicates that it is difficult to perform pollen-vegetation calibration for such taxa. Thus, the resulting RPPs for these taxa should be treated and evaluated with care.

4.4. Relative Pollen Productivity Estimates

We estimated RPPs for seven plant/pollen taxa, among which the RPPs for four taxa were first estimated, i.e., Cyclobalanopsis, Camelia, Ilex, and Symplocos. The nine combinations of the models yield some variation in the estimated RPPs which differ mainly from the applied pollen dispersal models (Table 3). Previous studies have also found a larger impact of the distance-weighting method on the estimated RPPs than the selection of ERV sub-models (e.g., [12]). In this study, the values (using Castanopsis as the reference) produced with Prentice’s model are consistently lower than those produced with 1/d and LSM model irrespective of the plant taxon and ERV models used. The 1/d and LSM models tend to produce similar values of RPPs. Similar result between the 1/d and LSM model is reasonable because the impact of taxon-specific fall speed is nearly negligible for fall speeds smaller than 0.04 m/s in the LSM [37] and 1/d does not consider the inter-taxonomic differences in pollen dispersal. However, the fall speed of pollen has a much larger impact on Prentice’s model which uses a Gaussian plume diffusion model (GPM) which was suggested to predict higher RPPs for taxa with large pollen with high fall speed [12,42,43]. In the case of this study, Castanopsis has the lowest FSP (0.0034 m/s). The resulting RPPs with the GPM (i.e., higher RPP for Castanopsis and lower for the other taxa compared to the result of 1/d) do not follow but rather contradict the suggested trend, i.e., higher RPPs for large pollen while lower RPPs for small pollen [42]. A similar result in this study was also obtained in other studies [14,38]. For example, in the study of [14], the 1/d method always produced lower RPPs for Castanea (a sister lineage to Castanopsis and usually used for comparison) than with Prentice’s distance-weighting method. The disagreement between the studies regarding the relationships between pollen size and RPP might be largely due to the Gaussian formulation-based calculation of the distance-weighted plant abundance (DWPA) in Prentice’s model. Such calculation can lead to a lower DWPA (thus a higher RPP) for a taxon with both too high and too low fall speed.
By observing the large difference in the estimated RPPs using different distance-weighting methods, a majority of studies preferred to use the GPM for RPP calculation [12]. Others demonstrated that the LSM is more realistic [38,42]. In this study, Prentice’s model apparently increases the RPP for Castanopsis and decreases the RPPs for the others compared to the other distance-weighting methods. Further study is needed to evaluate the appropriateness of these models.
Many studies use Poaceae as a reference taxon to estimate RPPs because it is one of the most common pollen types in the open and semi-open vegetation in the temperate zone. However, in the subtropical zone of China, Poaceae mainly occurs as bamboo in the mountainside. Different from other Poaceae herbs, most species of bamboo bloom only once in their lifetime and were demonstrated to have a significant effect on the estimated RPPs [20]. Our test runs with Poaceae included illustrate a similar result. Therefore, Poaceae was not included in our RPP estimates. Pinus was neither included in this study although it has large spread values in both vegetation and pollen data. The reason is that the average plant abundance of Pinus among sites shows a continuously increasing trend as distance increases, which violates the assumption of the ERV model of stationary vegetation (results not shown) [10].
There are two available subtropical works, one tropical work, and one synthesized temperate dataset from China that can be compared with the RPPs obtained in this study. All RPPs were converted to values relative to Castanopsis/Castanea (Table 4). Overall, our estimated RPPs based on ERV sub-model 3 and Prentice’s or 1/d vegetation distance-weighting method are generally comparable with other published values obtained in China. The RPP estimates for Liquidambar and Cyclobalanopsis produced with 1/d distance-weighting are more similar to previous studies than Prentice’s model whereas the RPP for Mallotus estimated with Prentice’s model is closer to the one published value than the 1/d method. Given that few RPP estimates are available for comparison, a further validation of these values using a REVEALS model-based method (see the application of this method in [44]) would enable a more meaningful comparison between studies.

5. Conclusions

A new RPP dataset from middle subtropical China was obtained which adds the first RPPs from several dominant elements in the evergreen broadleaved forest including Cyclobalanopsis, Camelia, Ilex, and Symplocos. The most important finding is that the application of different pollen dispersal models results in large different RPPs. Prentice’s model uses a Gaussian plume diffusion model apparently increases the RPP for Castanopsis with a very low fall speed as compared to 1/d and the LSM model. This result seems to contradict previous studies showing that the GPM would overestimate the RPPs for pollen with a high fall speed while underestimating for pollen with a low fall speed. Indeed, the Gaussian formulation-based Prentice’s model can lead to lower DWPA (thus a higher RPP) for a taxon with both too high and too low fall speed. By demonstrating the significant effect of pollen dispersal models on the RPP estimates and in view of a majority of studies relying on Prentice’s model to estimate RPPs, we suggest validating the different RPP datasets produced with different dispersal models in a real case, e.g., comparing the observed vegetation with an RPP dataset and REVEALS model-based land cover reconstruction.

Author Contributions

Conceptualization, Q.W.; funding acquisition, Q.W., K.H. and Z.Z. (Zhuo Zheng); investigation, Q.W., K.H., X.Z., Z.Z. (Zhong Zhang) and Z.Z. (Zhuo Zheng); methodology, Q.W.; software, Q.W.; writing—original draft, Q.W.; writing—review and editing, Q.W., K.H., C.C., Y.T., X.Z. and Z.Z. (Zhuo Zheng). All authors have read and agreed to the published version of the manuscript.

Funding

This research was funded by the National Natural Science Foundation of China (grant numbers 42077414, 42072205, 41630753, and 41902186), the Innovation Group Project of Southern Marine Science and Engineering Guangdong Laboratory (Zhuhai) (grant number 311020002), and the Project of the Influence of Quaternary Glacial Cycle on Biodiversity in Nanling Mountains (grant Number 2021GJGY001). We also acknowledge the financial support from the Swedish Foundation for International Cooperation in Research and Higher Education (STINT), and the NSFC [grant number 41611130050] for a Sweden-China Exchange Grant 2016–2019 (PIs Marie-José Gaillard and Qinghai Xu). This study was undertaken as part of the Past Global Changes (PAGES) project and its working group LandCover6k, which in turn received support from the U.S. National Science Foundation and the Swiss Academy of Sciences.

Data Availability Statement

The data presented in this study are available on request from the corresponding author.

Acknowledgments

The authors would like to thank Zhen Li, Qiang Fan and Wenbo Liao for providing the vegetation survey data. We also thank all members of the palynology team at Sun Yat-sen University for their help with field work.

Conflicts of Interest

The authors declare no conflict of interest.

References

  1. Zheng, Z.; Ma, T.; Roberts, P.; Li, Z.; Yue, Y.; Peng, H.; Huang, K.; Han, Z.; Wan, Q.; Zhang, Y.; et al. Anthropogenic impacts on Late Holocene land-cover change and floristic biodiversity loss in tropical southeastern Asia. Proc. Natl. Acad. Sci. USA 2021, 118, e2022210118. [Google Scholar] [CrossRef] [PubMed]
  2. Ellis, E.C.; Gauthier, N.; Goldewijk, K.K.; Bird, R.B.; Boivin, N.; Díaz, S.; Fuller, D.Q.; Gill, J.L.; Kaplan, J.O.; Kingston, N.; et al. People have shaped most of terrestrial nature for at least 12,000 years. Proc. Natl. Acad. Sci. USA 2021, 118, e2023483118. [Google Scholar] [CrossRef] [PubMed]
  3. Harrison, S.P.; Stocker, B.D.; Klein Goldewijk, K.; Kaplan, J.O.; Braconnot, P. Do we need to include anthropogenic land-use and land-cover changes in paleoclimate simulations? Past Glob. Chang. Mag. 2018, 26, 4–5. [Google Scholar] [CrossRef] [Green Version]
  4. Overpeck, J.T.; Webb, T., III; Prentice, I.C. Quantitative interpretation of fossil pollen spectra: Dissimilarity coefficients and the method of modern analogs. Quat. Res. 1985, 23, 87–108. [Google Scholar] [CrossRef]
  5. Prentice, I.C.; Guiot, J.; Huntley, B.; Jolly, D.; Cheddadi, R. Reconstructing biomes from palaeoecological data: A general method and its application to European pollen data at 0 and 6 ka. Clim. Dyn. 1996, 12, 185–194. [Google Scholar] [CrossRef]
  6. 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]
  7. Sugita, S. Theory of quantitative reconstruction of vegetation II: All you need is LOVE. Holocene 2007, 17, 243–257. [Google Scholar] [CrossRef]
  8. Prentice, I.C.; Parsons, R.W. Maximum likelihood linear calibration of pollen spectra in terms of forest composition. Biometrics 1983, 39, 1051–1057. [Google Scholar] [CrossRef]
  9. Sugita, S. A model of pollen source area for an entire lake surface. Quat. Res. 1993, 39, 239–244. [Google Scholar] [CrossRef]
  10. Sugita, S. Pollen representation of vegetation in Quaternary sediments: Theory and method in patchy vegetation. J. Ecol. 1994, 82, 881–897. [Google Scholar] [CrossRef]
  11. 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] [Green Version]
  12. Poska, A.; Meltsov, V.; Sugita, S.; Vassiljev, J. Relative pollen productivity estimates of major anemophilous taxa and relevant source area of pollen in a cultural landscape of the hemi-boreal forest zone (Estonia). Rev. Palaeobot. Palynol. 2011, 167, 30–39. [Google Scholar] [CrossRef]
  13. Xu, Q.; Cao, X.; Tian, F.; Zhang, S.; Li, Y.; Li, M.; Li, J.; Liu, Y.; Liang, J. Relative pollen productivities of typical steppe species in northern China and their potential in past vegetation reconstruction. Sci. China Earth Sci. 2013, 57, 1254–1266. [Google Scholar] [CrossRef]
  14. 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] [Green Version]
  15. Kleinen, T.; Tarasov, P.; Brovkin, V.; Andreev, A.; Stebich, M. Comparison of modeled and reconstructed changes in forest cover through the past 8000 years. Holocene 2011, 21, 723–734. [Google Scholar] [CrossRef] [Green Version]
  16. Kuosmanen, N.; Marquer, L.; Tallavaara, M.; Molinari, C.; Zhang, Y.; Alenius, T.; Edinborough, K.; Pesonen, P.; Reitalu, T.; Renssen, H.; et al. The role of climate, forest fires and human population size in Holocene vegetation dynamics in Fennoscandia. J. Veg. Sci. 2018, 29, 382–392. [Google Scholar] [CrossRef]
  17. Gaillard, M.-J.; Githumbi, E.; Achoundong, G.; Lézine, A.-M.; Hély, C.; Lebamba, J.; Marquer, L.; Mazier, F.; Li, F.; Sugita, S. The Challenge of Pollen-Based Quantitative Reconstruction of Holocene Plant Cover in Tropical Regions: A Pilot Study in Cameroon. In Quaternary Vegetation Dynamics, the African Pollen Database; Runge, J., Gosling, W.D., Lézine, A.-M., Scott, L., Eds.; CRC Press: Boca Raton, FL, USA, 2021; pp. 183–206. [Google Scholar]
  18. Tabares, X.; Ratzmann, G.; Kruse, S.; Theuerkauf, M.; Mapani, B.; Herzschuh, U. Relative pollen productivity estimates of savanna taxa from southern Africa and their application to reconstruct shrub encroachment during the last century. Holocene 2021, 31, 1100–1111. [Google Scholar] [CrossRef]
  19. Duffin, K.I.; Bunting, M.J. Relative pollen productivity and fall speed estimates for southern African savanna taxa. Veg. Hist. Archaeobotany 2007, 17, 507–525. [Google Scholar] [CrossRef]
  20. Chen, H.; Xu, Q.; Zhang, S.; Sun, Y.; Wang, M.; Zhou, Z. Relative pollen productivity estimates of subtropical evergreen and deciduous broadleaved mixed forest in Ta-pieh Mountains. Quat. Sci. 2019, 39, 469–482. [Google Scholar]
  21. Jiang, F.; Xu, Q.; Zhang, S.; Li, F.; Zhang, K.; Wang, M.; Shen, W.; Sun, Y.; Zhou, Z. Relative pollen productivities of the major plant taxa of subtropical evergreen–deciduous mixed woodland in China. J. Quat. Sci. 2020, 35, 526–538. [Google Scholar] [CrossRef]
  22. Wan, Q.; Zhang, Y.; Huang, K.; Sun, Q.; Zhang, X.; Gaillard, M.-J.; Xu, Q.; Li, F.; Zheng, Z. Evaluating quantitative pollen representation of vegetation in the tropics: A case study on the Hainan Island, tropical China. Ecol. Indic. 2020, 114, 106297. [Google Scholar] [CrossRef]
  23. Wu, Z.Y. Vegetation of China; Science Press: Beijing, China, 1980. [Google Scholar]
  24. He, Z.; Wu, Q.; Li, Y. Evaluation and Exploitation of the Meteorological Tourism Resources in Jinggangshan. Jiangxi Sci. 2018, 36, 262–296. [Google Scholar]
  25. Huang, K.; Zheng, Z.; Liao, W.; Cao, L.; Zheng, Y.; Zhang, H.; Zhu, G.; Zhang, Z.; Cheddadi, R. Reconstructing late Holocene vegetation and fire histories in monsoonal region of southeastern China. Palaeogeogr. Palaeoclimatol. Palaeoecol. 2014, 393, 102–110. [Google Scholar] [CrossRef]
  26. Liao, W.; Wang, Y.; Li, Z.; Peng, S.; Chen, C.; Fan, Q.; Jia, F.; Wang, L.; Liu, W.; Yin, G.; et al. Intergrated Study on Biodiversity of Mount Jinggangshan Regions in China; Science Press: Beijing, China, 2014. [Google Scholar]
  27. 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] [Green Version]
  28. Fægri, K.; Iversen, J. Textbook of Pollen Analysis; Wiley: New York, NY, USA, 1989. [Google Scholar]
  29. Institute of Botany and South China Institute of Botany, A.S. Angiosperm Pollen Flora of Tropic and Subtropic China; Science Press: Beijing, China, 1982. [Google Scholar]
  30. Wang, F.; Chien, N.; Zhang, Y.; Yang, H. Pollen Flora of China; Science Press: Beijing, China, 1995. [Google Scholar]
  31. Tang, L.; Mao, L.; Shu, J.; Li, C.; Shen, C.; Zhou, Z. An Illustrated Handbook of Quaternary Pollen and Spores in China; Science Press: Beijing, China, 2016. [Google Scholar]
  32. Gregory, P.H. The Microbiology of the Atmosphere; Leonard Hill: Aylesbury, UK, 1973. [Google Scholar]
  33. Gaillard, M.-J.; Sugita, S.; Bunting, M.J.; Middleton, R.; Broström, A.; Caseldine, C.; Giesecke, T.; Hellman, S.E.V.; Hicks, S.; Hjelle, K.; et al. The use of modelling and simulation approach in reconstructing past landscapes from fossil pollen data: A review and results from the POLLANDCAL network. Veg. Hist. Archaeobotany 2008, 17, 419–443. [Google Scholar] [CrossRef] [Green Version]
  34. Chaput, M.A.; Gajewski, K. Relative pollen productivity estimates and changes in Holocene vegetation cover in the deciduous forest of southeastern Quebec, Canada. Botany 2018, 96, 299–317. [Google Scholar] [CrossRef]
  35. Abraham, V.; Kozáková, R. Relative pollen productivity estimates in the modern agricultural landscape of Central Bohemia (Czech Republic). Rev. Palaeobot. Palynol. 2012, 179, 1–12. [Google Scholar] [CrossRef]
  36. 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]
  37. 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]
  38. Mariani, M.; Connor, S.E.; Theuerkauf, M.; Kuneš, P.; Fletcher, M.S. Testing quantitative pollen dispersal models in animal-pollinated vegetation mosaics: An example from temperate Tasmania, Australia. Quat. Sci. Rev. 2016, 154, 214–225. [Google Scholar] [CrossRef]
  39. 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]
  40. Räsänen, S.; Suutari, H.; Nielsen, A.B. A step further towards quantitative reconstruction of past vegetation in Fennoscandian boreal forests: Pollen productivity estimates for six dominant taxa. Rev. Palaeobot. Palynol. 2007, 146, 208–220. [Google Scholar] [CrossRef]
  41. Fang, Y.; Bunting, M.J.; Ma, C.; Yang, X. Are modern pollen assemblages from soils and mosses the same? A comparison of natural pollen traps from subtropical China. Catena 2022, 209, 105790. [Google Scholar] [CrossRef]
  42. Theuerkauf, M.; Couwenberg, J. Pollen productivity estimates strongly depend on assumed pollen dispersal II: Extending the ERV model. Holocene 2021, 32, 1233–1250. [Google Scholar] [CrossRef]
  43. 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] [Green Version]
  44. Wan, Q.; Zhang, X.; Huang, K.; Zheng, H.; Zhang, Y.; Yang, X.; Zheng, Z. Validation of relative pollen productivities for major tropical plant taxa: An empirical test using the REVEALS model in Hainan, China. Quat. Int. 2022, 641, 106–114. [Google Scholar] [CrossRef]
Land 12 01337 g002 550
Figure 2. Pollen percentage diagram of the major pollen taxa found in the 32 surface soil samples.
Figure 2. Pollen percentage diagram of the major pollen taxa found in the 32 surface soil samples.
Land 12 01337 g002
Land 12 01337 g003 550
Figure 3. Plots of the log-likelihood values with distance for nine combinations of three ERV sub-models and three distance-weighting methods (Prentice’s model, inverse distance (1/d), and Lagrangian Stochastic Model (LSM)), using pollen data from modern surface soil samples at 23 random sites and related vegetation data within 1500 m radius around each sample. (a) nine combinations of the models; (b) a combination of Prentice’s distance weighting method and three ERV sub-models; (c) a combination of ERV sub-model 3 and three distance-weighting methods.
Figure 3. Plots of the log-likelihood values with distance for nine combinations of three ERV sub-models and three distance-weighting methods (Prentice’s model, inverse distance (1/d), and Lagrangian Stochastic Model (LSM)), using pollen data from modern surface soil samples at 23 random sites and related vegetation data within 1500 m radius around each sample. (a) nine combinations of the models; (b) a combination of Prentice’s distance weighting method and three ERV sub-models; (c) a combination of ERV sub-model 3 and three distance-weighting methods.
Land 12 01337 g003
Land 12 01337 g004 550
Figure 4. Scatter plots of the pollen-vegetation relationships at the distance of the relevant source area of pollen (RSAP = 500 m) as estimated using the ERV sub-model 3 and Prentice’s vegetation distance-weighting method.
Figure 4. Scatter plots of the pollen-vegetation relationships at the distance of the relevant source area of pollen (RSAP = 500 m) as estimated using the ERV sub-model 3 and Prentice’s vegetation distance-weighting method.
Land 12 01337 g004
Table
Table 1. Location of the 32 studied sites and the major vegetation communities surveyed in the field within a 100 m radius area of the sampling sites.
Table 1. Location of the 32 studied sites and the major vegetation communities surveyed in the field within a 100 m radius area of the sampling sites.
SiteLatitudeLongitudeElevationMajor Vegetation Communities within 100 m
(°)(°)(m)
5114.2426.74314Woodland of Cunninghamia R. Br. ex A. Rich. and Cryptomeria D. Don; Bamboo
6114.1826.47520Woodland of Cyclobalanopsis Oerst. and Castanopsis Spach; crop field
14114.1226.35441Woodland of Castanopsis Spach and Cunninghamia; Bamboo
18114.1126.32480Woodland of Castanopsis and Castanea Mill.
22114.1626.39322Woodland of Castanopsis, Manglietia Blume, and Sterculia L.; Woodland of Cunninghamia; Bamboo
27114.1126.68320Woodland of Liquidambar L, Camellia L., and Schima Reinw. ex Blume; Bamboo; crop field
32114.2926.49275Woodland of Cyclobalanopsis, Sassafras J. Presl, and Machilus Nees; crop field
33114.2226.71380Woodland of Sapium Jacq., Fissistigma Griff., Eurya Thunb., and Myrtaceae Juss.; woodland of Pinus L.; Bamboo
52114.0326.67340Woodland of Machilus, Eurya, and Cinnamomum verum; Bamboo; crop field
222114.1926.53524Woodland of Castanopsis, Exbucklandia R. W. Brown, and Garcinia L.
223114.1726.65642Woodland of Fagus L. and Castanopsis; crop field
DSC113.926.72240Woodland of Cyclobalanopsis, Schima, and Taiwania Hayata; woodland of Camellia; crop field
GCZ114.0426.77221Woodland of Pinus and Cunninghamia; Bamboo; crop field
N1114.1126.61300Woodland of Daphniphyllum Bl.; woodland of Cryptomeria D. Don; woodland of Castanopsis
N2114.126.571330Woodland of Eurya and Acer L.; woodland of Crypotomeria
N3114.1126.541180Woodland of Machilus, Pinus, Castanopsis, and Schima; Bamboo
S01114.3126.53253Woodland of Castanopsis, Schima, and Cunninghamia; crop field
S05114.2226.51377Woodland of Castanopsis and Distylium Siebold and Zucc.; Bamboo; crop field
S15114.1726.5829Woodland of Castanopsis, Cunninghamia, and Cyclobalanopsis; woodland of Cunninghamia
S20114.1726.55594Woodland of Castanopsis
W1114.2626.63690Woodland of Castanopsis, Machilus, and Eurya
W10114.126.52910Woodland of Castanopsis, Liquidambar, Cyclobalanopsis, and Machilus; Bamboo; woodland of Pinus; crop field; woodland of Crypotomeria
W11114.1526.54880Woodland of Cunninghamia, Cyclobalanopsis, Castanopsis
W2114.2726.6370Woodland of Schima, Symplocos, and Castanopsis; woodland of Cunninghamia; Bamboo; crop field
W3114.2226.66370Woodland of Castanopsis and Ardisia Sw.; woodland of Cunninghamia; Bamboo; crop field
W4114.1726.57750Woodland of Castanopsis and Schima; woodland of Camellia
W5114.2226.42269Woodland of Castanopsis and Lophatherum Brongn.; woodland of Cunninghamia; Bamboo; crop field
W6114.1426.44776Woodland of Castanopsis and Michelia L.; Bamboo; crop field
W7114.126.45310Woodland of Liquidambar and Cyclobalanopsis; Bamboo; grassland
W8114.2126.46340Woodland of Castanopsis and Schima; Bamboo; crop field
W9114.0826.51200Woodland of Castanopsis and Cunninghamia; Bamboo
Y1114.0726.291076Woodland of Acer and Rhododendron L.; crop field
Table
Table 2. Constituent plant species and estimates of pollen fall speed for eight selected taxa based on measurement of 30 grains using Stoke’s law [32].
Table 2. Constituent plant species and estimates of pollen fall speed for eight selected taxa based on measurement of 30 grains using Stoke’s law [32].
Plant SpeciesPollen TaxaDiameter of Long and Short Axes (μm)Fall Speed (m/s)
Liquidambar formosana Hance, Liquidambar acalycina ChangLiquidambar310.0287
Castanopsis faberi, Castanea henryi (Skan) Rehd. et Wils., Castanopsis eyrei, Castanopsis tibetana, Castanopsis lamontii Hance, Castanopsis sclerophylla, Lithocarpus glaber (Thunb.) Nakai, Lithocarpus litseifolius (Hance) ChunCastanopsis16, 90.0034
Cyclobalanopsis glauca (Thunberg) Oersted, Cyclobalanopsis multinervis W.C.Cheng and T.Hong, Cyclobalanopsis gracilis (Rehder et E. H. Wilson) W. C. Cheng et T. Hong, Cyclobalanopsis sessilifolia (Blume) SchottkyCyclobalanopsis21, 160.0091
Camellia sinensis (L.) O. Ktze., Camellia japonica L., Camellia oleifera Abel., Schima argentea Pritz. ex Diels, Schima superbaCamelia34, 250.0229
Ilex asprella (Hook. et Arn.) Champ. ex Benth., Ilex ficoidea Hemsl., Ilex pubescens Hook. et Arn., Ilex triflora Bl., Ilex chinensis Sims, Ilex kwangtungensis Merr.Ilex L.24, 170.0110
Mallotus japonicus (Thunb.) Muell. Arg.Mallotus Lour.20, 190.0109
Symplocos sumuntia Buch.-Ham. ex D. Don, Symplocos lancifolia Sieb. et Zucc.Symplocos Jacq.36, 300.0390
Table
Table 3. Relative pollen productivity (RPP) with standard error for eight taxa estimated with the ERV sub-models and three vegetation distance-weighting methods, i.e., Prentice’s taxon-specific model, the inverse distance (1/d), and the Lagrangian stochastic model (LSM).
Table 3. Relative pollen productivity (RPP) with standard error for eight taxa estimated with the ERV sub-models and three vegetation distance-weighting methods, i.e., Prentice’s taxon-specific model, the inverse distance (1/d), and the Lagrangian stochastic model (LSM).
TaxaERV 1 PrenticeERV 2 PrenticeERV 3 PrenticeERV 1
1/d		ERV 2
1/d		ERV 3
1/d		ERV 1
LSM		ERV2
LSM		ERV 3
LSM
Castanopsis1 ± 01 ± 01 ± 01 ± 01 ± 01 ± 01 ± 01 ± 01 ± 0
Ilex0.354 ± 0.030.324 ± 0.0280.352 ± 0.0310.936 ± 0.0790.827 ± 0.0710.861 ± 0.0761.204 ± 0.1041.096 ± 0.0981.159 ± 0.101
Mallotus0.558 ± 0.0880.532 ± 0.0850.221 ± 0.0281.166 ± 0.2010.762 ± 0.1090.615 ± 0.0921.377 ± 0.2281.104 ± 0.1911.578 ± 0.283
Liquidambar0.116 ± 0.0060.106 ± 0.0060.115 ± 0.0070.421 ± 0.0230.388 ± 0.0220.398 ± 0.0230.855 ± 0.0420.823 ± 0.0440.815 ± 0.042
Cyclobalanopsis0.124 ± 0.0070.109 ± 0.0060.107 ± 0.0060.263 ± 0.0140.243 ± 0.0140.227 ± 0.0130.843 ± 0.0460.797 ± 0.0450.7 ± 0.039
Camelia0.034 ± 0.0010.032 ± 0.0010.033 ± 0.0010.12 ± 0.0050.114 ± 0.0050.113 ± 0.0050.214 ± 0.0090.207 ± 0.0080.205 ± 0.008
Symplocos0.011 ± 0.0020.009 ± 0.0020.01 ± 0.0020.044 ± 0.0060.038 ± 0.0060.04 ± 0.0060.052 ± 0.0080.043 ± 0.0080.036 ± 0.006
Table
Table 4. Comparison of relative pollen productivity (RPP) estimates between this study and other values obtained in China.
Table 4. Comparison of relative pollen productivity (RPP) estimates between this study and other values obtained in China.
Study AreaJinggangshan MountainsTemperate ChinaTa-Pieh MountainsDaba MountainsHainan Cultural Landscapes
Model combinationsERV3_Prentice/1/dAlt-1ERV1_PrenticeERV2_PrenticeERV2_Prentice
Castanopsis/Castanea11111
Liquidambar0.12/0.40 0.28
Cyclobalanopsis/Quercus0.11/0.230.381.89 2.70
Mallotus0.22/0.62 0.11
ReferenceThis studyLi et al.
(2018) [39]		Chen et al.
(2019) [20]		Jiang et al. (2020) [21]Wan et al.
(2020) [22]
The RPPs from other studies were all converted to values relative to Castanopsis/Castanea.
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

Wan, Q.; Huang, K.; Chen, C.; Tang, Y.; Zhang, X.; Zhang, Z.; Zheng, Z. Relative Pollen Productivity Estimates for Major Plant Taxa in Middle Subtropical China. Land 2023, 12, 1337. https://doi.org/10.3390/land12071337

AMA Style

Wan Q, Huang K, Chen C, Tang Y, Zhang X, Zhang Z, Zheng Z. Relative Pollen Productivity Estimates for Major Plant Taxa in Middle Subtropical China. Land. 2023; 12(7):1337. https://doi.org/10.3390/land12071337

Chicago/Turabian Style

Wan, Qiuchi, Kangyou Huang, Cong Chen, Yongjie Tang, Xiao Zhang, Zhong Zhang, and Zhuo Zheng. 2023. "Relative Pollen Productivity Estimates for Major Plant Taxa in Middle Subtropical China" Land 12, no. 7: 1337. https://doi.org/10.3390/land12071337

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