Chinese Journal of Oceanology and Limnology   2016, Vol. 34 issue(5): 937-951     PDF       
http://dx.doi.org/10.1007/s00343-016-5111-4
Institute of Oceanology, Chinese Academy of Sciences
0

Article Information

CHI Yuan(池源), SHI Honghua(石洪华), WANG Xiaoli(王晓丽), QIN Xuebo(覃雪波), ZHENG Wei(郑伟), PENG Shitao(彭士涛)
Impact factors identification of spatial heterogeneity of herbaceous plant diversity on five southern islands of Miaodao Archipelago in North China
Chinese Journal of Oceanology and Limnology, 34(5): 937-951
http://dx.doi.org/10.1007/s00343-016-5111-4

Article History

Received Apr. 7, 2015
accepted in principle Jun. 15, 2015
accepted for publication Jul. 13, 2015
Impact factors identification of spatial heterogeneity of herbaceous plant diversity on five southern islands of Miaodao Archipelago in North China
CHI Yuan(池源)1, SHI Honghua(石洪华)1, WANG Xiaoli(王晓丽)2, QIN Xuebo(覃雪波)3, ZHENG Wei(郑伟)1, PENG Shitao(彭士涛)1        
1 The First Institute of Oceanography, State Oceanic Administration, Qingdao 266061, China;
2 College of Environmental Science and Safety Engineering, Tianjin University of Technology, Tianjin 300384, China;
3 Tianjin Natural History Museum, Tianjin 300201, China
ABSTRACT: Herbaceous plants are widely distributed on islands and where they exhibit spatial heterogeneity. Accurately identifying the impact factors that drive spatial heterogeneity can reveal typical island biodiversity patterns. Five southern islands in the Miaodao Archipelago, North China were studied herein. The spatial distribution of herbaceous plant diversity on these islands was analyzed, and the impact factors and their degree of impact on spatial heterogeneity were identified using CCA ordination and ANOVA. The results reveal 114 herbaceous plant species, belonging to 94 genera from 34 families in the 50 plots sampled. The total species numbers on different islands were significantly positively correlated with island area, and the average α diversity was correlated with human activities, while the β diversity among islands was more affected by island area than mutual distances. Spatial heterogeneity within islands indicated that the diversities were generally high in areas with higher altitude, slope, total nitrogen, total carbon, and canopy density, and lower moisture content, pH, total phosphorus, total potassium, and aspect. Among the environmental factors, pH, canopy density, total K, total P, moisture content, altitude, and slope had significant gross effects, but only canopy density exhibited a significant net effect. Terrain affected diversity by restricting plantation, plantation in turn influenced soil properties and the two together affected diversity. Therefore, plantation was ultimately the fundamental driving factor for spatial heterogeneity in herbaceous plant diversity on the five islands.
Key words: Island ecology     herbaceous plant     biodiversity     spatial heterogeneity     impact factor identification     plantation    
1 INTRODUCTION

Plant diversity and impact factors have always been a research focus in ecology and phytogeography (Qian and Ma, 1994; Tilman, 2000). Islands are important platforms for protecting marine environments and maintaining the ecological balance. However, island ecosystems are vulnerable because they are isolated and their spaces are limited (Wang et al., 2014; Chi et al., 2015). This mainly manifests as simple regional structure and weak self-regulation and recovery capabilities (Shi et al., 2013); natural disasters, such as storms and drought, have a greater impact on island than on mainland ecosystems (Qie et al., 2011; Bustamante-Sánchez and Armesto, 2012; Katovai et al., 2012). Plant diversity is closely related to ecosystem structure and function (Schulze and Mooney, 1993; Heywood, 1995). Extensive studies have shown that plant diversity plays a fundamental role in maintaining and regulating ecosystem productivity, material recycling, and system stability (Tilman et al., 1997, 2001, 2006; Hooper et al., 2005; Cardinale et al., 2006; Ma, 2013). Plant diversity is especially important for island ecosystems. Herbaceous plants occupy the largest area and make up most of the plant species found on islands, and they are more sensitive to the environment. Moreover, herbaceous plants are the key protection focus of regional biodiversity conservation.

Herbaceous plant diversity exhibits spatial heterogeneity on islands, which includes heterogeneity between and within islands, and is affected by various factors due to the particular island environment. Heterogeneity between islands is closely related to both area differences and mutual distances (MacArchur and Wilson, 1963, 1967; Harrison et al., 1992; Soininen et al., 2007), and may be influenced by human activity (Halpern et al., 2008). Heterogeneity within islands is more complex, and is affected by environmental factors. Most islands in North China are bedrock islands, and eroded hills are the main topographic feature; therefore, terrain and soil are the basic impact factors. Native forests are always poor on islands as a result of atrocious weather, and plantation forestry has become an important mode of island ecosystem maintenance. However, plantation forestry essentially interferes with the native plants and may have an impact on native plant diversity (Rosoman, 1994; Michelsen et al., 2014). Islands exhibit both ocean and land ecosystem characteristics (Shi et al., 2009), whereby land-based life is influenced by marine hydrology and climate. Generally, sites close to the shore are more affected by the sea compared with those located in the island’s interior, which may result in differences in biodiversity. Terrain, soil, plantation, and the marine environment are all potential herbaceous plant diversity impact factors. Identifying impact factors and impact degree can reveal spatial distribution in herbaceous plant diversity on typical islands and provide a reference for biodiversity conservation. However, the existing research on island biodiversity has focused on descriptive statistics and model building, always taking the whole island as the research unit; thus, little attention has been paid to the spatial heterogeneity of diversity, especially heterogeneity within islands and its impact factors.

Therefore, we studied herbaceous plants on five southern islands in the Miaodao Archipelago that represent typical islands of North China. The spatial distribution of diversity was analyzed based on field surveys and sampling. The impact factors and impact degree of spatial heterogeneity on diversity were then identified using canonical correspondence analysis (CCA) ordination and ANOVA (analysis of variance) to elucidate the spatial characteristics of herbaceous plant diversity on typical islands and to provide a basis for island biodiversity conservation.

2 MATERIAL AND METHOD 2.1 Study area

The Miaodao Archipelago is located to the north of the Shandong Peninsula, at the juncture of the Yellow and the Bohai Seas (Fig. 1). We chose the following five southern islands, i.e., Nanchangshan Island, Beichangshan Island, Miao Island, Daheishan Island, and Xiaoheishan Island, because they are close to each other and to the mainland. They are the demographic, economic, and cultural center of Changdao County, Shandong Province, China. The five islands are in the East Asian monsoon region, with an average annual temperature of 12.0°C, average January temperature of -1.6°C, average July temperature of 24.5°C, and average annual rainfall of 537 mm (mostly concentrated between June–September). The region has ample sunshine, with an average annual sunshine time of 2 612 h. The terrain of all five islands is undulating, with mountains lying in a roughly south-north direction; the highest point is approximately 190 m. The soils mainly fall into to three categories: brown, cinnamon, and fluvo-aquic soils, with the brown soil occupying the largest area at a thickness of approximately 30 cm. The soil quality is poor with a lot of gravel (Shi et al., 2013). The native forests on the five islands are poor, and continuous plantation forestry brings the forest coverage to almost 60%, with Pinus thunbergii, Robinia pseudoacacia, and Platycladus orientalis as the dominant species. Woody plant species diversity is relatively low, and there are only several tree species and approximately ten shrub species, including Quercus acutissima and Vitex negundo. However, there is a variety of widely distributed native herbaceous plants. Moreover, the five islands vary in area and mutual distances. Therefore, these islands provided the necessary conditions for the present study. With increased population growth and the impact of urbanization in recent years, biodiversity conservation on these islands is facing difficult challenges (Zheng et al., 2014).

Figure 1 Sample plots of herbaceous plant on five southern islands in the Miaodao Archipelago, China Is 1: Nanchangshan Island; Is 2: Beichangshan Island; Is 3: Miao Island; Is 4: Daheishan Island; Is 5: Xiaoheishan Island.
2.2 Data collection

The field survey and sampling of herbaceous plants were carried out in the summer of 2012. Fifty sample plots were investigated based on island area, community type, and terrain factors (Fig. 1). The sample plots were 20 m×20 m in plantation communities and 10 m×10 m in natural herbaceous communities, which is consistent with previous research (Fang et al., 2009; Gao et al., 2014). The latitude and longitude, altitude, slope, and aspect of each plot were measured using a Trimble Geo Explorer GPS and SHEFFIELD compass. The plantation type was also recorded, and the canopy density (0–1) was documented based on the growth status of trees at the plot. Five quadrats were set up at the 4 corners and the center of each plot, each quadrat was 1 m×1 m (Gao et al., 2014). The species, abundance, coverage, and height of each species in five quadrats were recorded and averaged to obtain the data for each plot. Additionally, a method for mixing soil samples from multiple points was used; the soil samples collected from three points in each plot were evenly mixed to represent the soil property of the plot. Soil factors including moisture content, pH, total P, total K, total N, and total C were then measured in the laboratory.

Remote sensing images of the region covering the five islands taken in August, 2013 by LANDSAT8 (Chu et al., 2013) satellite (resolution 30 m) were used in this study. Based on these, contour vector maps for the five islands were drawn in ArcGIS 10.0 (Hillier, 2011) and the nearest distances between sample plots and island shore were extracted.

2.3 Data analysis 2.3.1 Species statistics

The species of herbaceous plants on the five islands were recorded and sorted into genera and families; the proportions of different families and genera were also calculated.

(1) Important value

Important value (Ⅳ), first propounded by Curtis and McIntosh (1951), is a comprehensive quantitative index to reflect the role and status of a species in the community and has become a key biodiversity measure (Qian and Ma, 1994; Zhang, 2004). Species features such as abundance, coverage, height, dominance, and frequency in a community decide the Ⅳ of the species, and there have been various Ⅳ calculation methods based on community type and existing data used. Abundance, coverage, and dominance are often selected to calculate the Ⅳ in tree layers and the classic formula proposed by Curtis and McIntosh used these three factors. As for shrub layer and herb layer, height was always adopted to replace dominance (Fang et al., 2009). Therefore, the Ⅳ of each species was calculated using the following Eq.1 (Zhang, 2004):

    (1)

where Ps, i is the importance value (Ⅳ), and Abs, i, Cos, i, and Hes, i are abundance, coverage, and height, respectively, of species i in sample plot s, while Abs, Cos, and Hes are the total abundance, total coverage, and total height, respectively, in sample plot s. Here, abundance directly represents the individual number of a species, while coverage and height reflect the growth status, i.e., coverage means the percentage of project area of a species in the sample plot, and height was the maximum vertical distance to the ground of blades, which are both important functional traits associated with ecological processes such as heat load, water retention, gas exchange, and light competition (Schulze et al., 1996; Anten and Hirose, 1999; Gallagher et al., 2011; Gao et al., 2014). According to the sum of the importance values of a species in each plot, the common (ranked top 20 in importance value) and dominant (ranked top three in importance value) species on the five islands were determined.

(2) Rarefaction curve

A rarefaction curve indicates the expected number of species from a collection of random sample plots and represents what is statistically expected from the accumulation curve (Gotelli and Colwell, 2001). It can help to judge whether or not the number of sample plots is adequate to represent the species in the study area, and to clarify the features of species richness in different regions. In this paper, rarefaction curves were created by repeatedly sampling all of the species collected at random using the software Estimate S 9.1 (Colwell, 1997).

2.3.2 Species diversity

(1) The α diversity

The α diversity is represented by the ShannonWiener index (H') and Pielou index (E) that are commonly used in relevant studies worldwide. While the former mainly reflects the species complexity of a community, and the latter indicates species evenness. They were respectively calculated using Eqs.2, 3 (Ma and Liu, 1994):

    (2)
    (3)

where H's and Es represent H' and E, respectively, of sample plot s; Pi, s is the importance value of species i at sample plot s; and Ns is the number of species at sample plot s.

(2) The β diversity

The β diversity is represented by the Jaccard index to reflect species differences between islands. This index has a wide application in diversity research with reliable results (Nekola and White, 1999; Mac Nally et al., 2004; Qian, 2009). It was calculated using Eq.4 (Qian, 2009):

    (4)

where J is the Jaccard index between two islands; a represents the number of species common to both islands; b is the number of species present in the first island only, and c indicates the number of species present in the second island only.

2.3.3 Impact factor identification

(1) Environmental factor statistics

Four types of environmental factors, i.e., terrain, soil, plantation, and marine, were analyzed. Terrain contained three quantitative factors, i.e., altitude, slope, and aspect; among them, the original aspect increased clockwise between 0–360°; 0° and 180° indicate north and south, respectively. We used south as the principle direction, and the aspect was standardized using Eq.5:

    (5)

where AS s is the standardized aspect, and As is the original aspect. A larger value of standardized aspect indicates a more southerly direction.

Soil included six quantitative factors, i.e., soil pH, moisture content, total P, total K, total N, and total C. Plantation condition was represented by two factors including one quantitative factor (canopy density) and one nominal factor (community type). The marine factor was evaluated by distance to the shore.

(2) CCA

The CCA was carried out in Canoco 4.5 (Jiang et al., 2007), with ‘sample plot × importance value’ and ‘sample plot × quantitative factors’ matrices as species and environmental data, respectively. The significance test of axes was conducted using the Monte Carlo permutation test.

The impact degrees of different environmental factors were identified. CCA was performed with each environmental factor as an independent variable to analyze the gross effect of each factor, and partial CCA was carried out with each environmental factor as an independent variable and the others as covariates to analyze the net effect of each factor. Significance tests were also performed using the Monte Carlo permutation test, and the impact degree of each environmental factor on species distribution was evaluated by its canonical characteristics (Ren et al., 2012).

(3) ANOVA

ANOVA was carried out in PASW Statistics 18 (Norušis, 2010) to further verify the impacts of plantation and marine factors, which represented the uniqueness of islands.

3 RESULT 3.1 Species statistics 3.1.1 Species composition

We recorded 114 species of herbaceous plant on the five islands, belonging to 94 genera from 34 families in the 50 plots sampled. At family level, Compositae had the most species with 29, followed by Gramineae and Labiatae with 11 and eight species, respectively. At the genus level, Artemisia had the highest number of species with seven representatives; most of the others had one species in each genus, and a few had two species in each genus. The dominant species on the five islands were Cleistogenes chinensis, Carex lanceolata, and Artemisia argyi. The common species in ranking order are shown in Table 1.

Table 1 Common species of herbaceous plant on five southern islands in the Miaodao Archipelago, China

As shown in Table 2, the islands ranked by the total number of species in descending order were as follows: Nanchangshan Island, Beichangshan Island, Daheishan Island, Xiaoheishan Island, and Miao Island. Cleistogenes chinensis was the dominant species on all the islands and also the one with highest importance value, and thus was considered representative of herbaceous plants on all five islands.

Table 2 Species statistics for herbaceous plants on five southern islands in the Miaodao Archipelago, China

The total number of species on different islands was associated with the number of sample plots. Since the sample plots were designed based mainly on island area, the association indicated that island area played a decisive role in the number of species (Fig. 2).

Figure 2 Relationship between the number of species and number of sample plots (left), number of species and island area (right)
3.1.2 Rarefaction curve

Rarefaction curves for number of species versus number of sample plots are shown in Fig. 3. The curve essentially plateaued with an increase in the number of sample plots, which indicated that more sample plots would likely only yield a few additional species. The number of species in the same number of sample plots on different islands in descending order were: Daheishan Island ≈ Nanchangshan Island > Xiaoheishan Island > Beichangshan Island > Miao Island.

Figure 3 Rarefaction curve for number of species versus number of sample plots on all islands (left) and different islands (right)

Rarefaction curves for number of species versus number of islands are shown in Fig. 4. The number of species rose with an increase in the number of islands.

Figure 4 Rarefaction curve for number of species versus number of islands
3.2 Species diversity 3.2.1 The α diversity

Figure 5 shows the results of the spatial distribution of α diversity in herbaceous plants on the five islands, in which the values were classified using approximately 25%, 50%, and 75% as dividing points. The ShannonWiener indices of the sample plots ranged from 1.39 for plot 24 to 2.46 for plot 4 (mean=2.02), and the Pielou indices ranged from 0.831 for plot 44 to 1 for plot 25 (mean=0.928).

Figure 5 Spatial distribution of α diversity in herbaceous plant on five southern islands in the Miaodao Archipelago, China

The box-plot of α diversity of different islands is shown in Fig. 6. The α diversity of Daheishan, Nanchangshan, and Xiaoheishan Islands were generally higher, followed by Beichangshan and Miao Islands.

Figure 6 Box-plot of α diversity in herbaceous plant on different islands
3.2.2 The β diversity

The Jaccard indices between any two islands are shown in Table 3; all were < 0.5, reflecting an evident difference in species composition between islands. Meanwhile, the Jaccard indices comparing Miao Island and Xiaoheishan Island with other islands were relatively low.

Table 3 Jaccard index between different islands

Among all of the islands, Nanchangshan and Beichangshan Islands were closest at approximately 0.9 km apart, while Nanchangshan and Daheishan Islands were furthest apart at 8.2 km. Meanwhile, island area differed significantly and that of Nanchangshan Island (the largest) was approximately 10 times of that of Xiaoheishan Island (the smallest) (Table 4). Mutual minimum distances did not affect Jaccard index, but Jaccard index exhibited an increasing trend with an increase in island area (Fig. 7).

Table 4 Island area and minimum distance between each island
Figure 7 Relationship between Jaccard index and minimum distance (left) and mean Jaccard index and island area (right)
3.3 CCA results 3.3.1 CCA diagram

The CCA results revealed that each ordination axis was statistically significant (P < 0.01) and the cumulative percentage variance of speciesenvironment relation was up to 56.5%, which contained a significant amount of ecological information, indicating the reliability of the CCA results (Table 5). CCA axis 1 was significantly correlated with all of the factors, except for distance to the shore, i.e., significantly positively correlated with aspect (standardized, the same below), moisture content, pH, total P, and total K, and significantly negatively correlated with the altitude, slope, total N, total C, and canopy density. CCA axis 2 was significantly positively correlated with distance to the shore, pH, total P, total K, and canopy density.

Table 5 Summary of CCA ordination of herbaceous plant sample plots on five southern islands in the Miaodao Archipelago, China

The two-dimensional CCA diagram was generated based on axis 1 and axis 2 (Fig. 8), and the projected position and length of each environmental factor on the axis represents the direction and strength of its correlation with the axis. From left to right on axis 1, moisture content, pH, total P, total K, and aspect increased gradually, while altitude, slope, total N, total C, and canopy density gradually decreased. From bottom to top on axis 2, distance to the shore, soil pH, total P, total K, and canopy density increased significantly. There were more sample plots distributed in places with higher altitude, slope, total N, total C, and canopy density and lower moisture content, pH, total P, total K, and aspect. To clearly interpret the relationship between α diversity (Shannon-Wiener and Pielou indices) and environmental factors, the sample plots in different ranges of diversity were labeled with different colors. In the H' diagram, the distribution of red and blue sample plots was dispersed, while pink plots were mostly distributed in the 3rd quadrant, and green plots were mostly distributed in the 2nd and 3rd quadrants. On the whole, the H′ values were not significantly correlated with environmental factors. However, the sample plots with higher altitude, slope, total N, total C, canopy density and lower moisture content, pH, total P, total K, aspect generally displayed slightly higher H′ values. In the E diagram, the blue and green plots were mostly distributed in the 2nd and 3rd quadrants, while the red and pink sample plots were dispersed throughout all four quadrants. The sample plots with higher altitude, slope, total N, total C, canopy density and lower moisture content, pH, total P, total K, and aspect generally had higher E values.

Figure 8 CCA ordination diagrams of herbaceous plant sample plots on five southern islands

As shown in Fig. 9, the species CCA results indicated that the characteristic distribution of species in the CCA diagram was generally consistent with that of the sample plots, i.e., the distribution of species showed characteristics of both concentration and dispersion, suggesting that most species had similar environmental requirements, while the environments of some species were unique.

Figure 9 CCA ordination diagram of all species of herbaceous plant sample plots on five southern islands

The CCA diagram of common species and the correspondence between numbers and species are shown in Fig. 10 and Table 1, respectively. The distribution characteristics exhibited a certain consistency with that of the sample plots, yet the common species were scarce in the 4th quadrant. We also observed that common species were often distributed in the center of the diagram, suggesting that they are more universally adapted to the environment. A closer distance between species indicated that they were under more similar environmental conditions and were more likely to appear in the same plot, i.e., their symbiosis was strong. The common species were more concentrated in the 2nd and 3rd quadrants, and an evident symbiosis was observed between species 1 and 9, between species 4 and 18, and among species 7, 14, 15, and 16. The distribution of common species in the 1st and 4th quadrants was somewhat dispersed, i.e., they had a relatively weak symbiosis with other common species.

Figure 10 CCA ordination diagram of common species of herbaceous plant sample plots on five southern islands
3.3.2 Impact of environmental factors

The results of the gross and net effects of different environmental factors on herbaceous plant diversity are shown in Table 6. With the exceptions of slope, total C, and distance to the shore, the other eight environmental factors exhibited significant gross effects, and they ranked by impact degree from highest to lowest as follows: pH, canopy density, total K, total P, moisture content, altitude, slope, and total N. In the assessment of net effects, only canopy density was significant, whereas the remaining factors did not significantly affect species diversity. The net effects of all factors were lower than those of the gross effects.

Table 6 Gross and net effects of environmental factors on biodiversity

The relationships among different factors (Table 7) indicated aspect was not significantly correlated with the other factors and distance to the shore only significantly correlated with altitude; however, a significant correlation was observed among the remaining factors.

Table 7 Correlation coefficients among different environmental factors
3.4 ANOVA 3.4.1 Plantation factors

Analysis of the diversity under different ranges of canopy density revealed that α diversity generally showed a ‘down-up-down’ trend with an increase in canopy density, with the maximum appearing between a canopy density range of 0.6–0.8 (Fig. 11). One-way ANOVA indicated that Pielou index did not change significantly under different ranges of canopy density (PE=0.368), whereas the Shannon-Wiener index varied significantly (PH=0.040).

Figure 11 Biodiversity in different canopy density

This study recorded four types of plantation community, i.e., Pinus thunbergii forest, Robinia pseudoacacia forest, Platycladus orientalis forest, and broadleaf/conifer mixed forest. With canopy density as a covariate, ANOVA was performed to reveal the diversity in different types of plantation communities. The results revealed that community type did not exert a significant effect on the Shannon Wiener index (PH=0.644), but it significantly affected the Pielou index (PE=0.050) under conditions excluding the impact of canopy density. The α diversity of the Pinus thunbergii and broadleaf/ conifer mixed forests were higher, followed by that of the Robinia pseudoacacia and Platycladus orientalis forests (Table 8).

Table 8 Biodiversity in different types of plantation
3.4.2 Marine factor

Analysis of α diversity at different distances to the shore showed that diversity fluctuated with an increase in distance (Fig. 12). One-way ANOVA indicated that there was no significant difference in diversity among sample plots at different distances to the shore (PH=0.465, PE=0.824).

Figure 12 Biodiversity at different distance to island shore
4 DISCUSSION 4.1 Species diversity

This study recoded 114 species in 50 sample plots, which is considerably higher than the number of herbaceous plant species (25) recorded on Putuoshan Island, China (Li et al., 2012). The rarefaction curve indicated that a reasonable number of sample plots was taken on the five islands (Fig. 3) and that each island has contributed to maintaining biodiversity (Fig. 4).

The α diversity values were generally high on Daheishan, Nanchangshan, and Xiaoheishan Islands, while those on Beichangshan and Miao Islands were relatively low. Among the five islands, Beichangshan and Miao Islands are famous tourist attractions, and thus experience relatively frequent and intense human disturbances, whereas Daheishan and Xiaoheishan Islands are less affected by human disturbance. Moreover, Nanchangshan Island is the capital of Changdao County; it experiences intense human disturbance but effective conservation measures have brought about higher α diversity. Human activities have the potential to adversely affect biodiversity on all five islands.

The β diversity results showed that the Jaccard indices between different islands were all < 0.5, indicating large differences in species composition. Since island area was the most basic environmental characteristic, our data suggested that ecological niche processes (island area) played an important role in the formation of β diversity on the five islands; however, species diffusion (mutual distance) had no obvious effect (Harrison et al., 1992; Soininen et al., 2007).

In recent years, zeta (ζ) diversity was proposed to represent the spatial structure of multispecies distributions (Hui and Mcgeoch, 2014). Zeta diversity is the number of species shared by multiple assemblages (sites, samples, or areas); as a concept and metric that unifies incidence-based diversity measures (Hui and Mcgeoch, 2014), it can provide a new way to clarify biodiversity patterns and processes. The preliminary application of ζ in our study is shown in Fig. 13. ζ declined with an increase in the number of islands and the decline followed the typical form of power law, which revealed the diversity pattern from a new perspective.

Figure 13 The relationship between ζ and number of islands i
4.2 Identification of impact factors

(1) Terrain factors

The gross effects of altitude and slope were significant, i.e., sample plots with higher altitude and slope values generally had a greater diversity. However, neither factor exhibited significant net effects, suggesting that neither altitude nor slope alone exerted a significant effect on diversity. The correlation between environmental factors indicated a significant positive correlation among altitude, slope, and canopy density, and significant gross and net effects of canopy density on diversity. In fact, the difference in altitude on the five islands was generally small, and the highest altitude was < 200 m; there was thus no significant vertical variation in microclimate, and the slope was significantly positively correlated with altitude. However, sample plots with higher altitude and slopes are often not conducive to urbanrural construction or farmland reclamation, and extensive plantation has resulted in a high canopy density, i.e., altitude and slope factors affected the herbaceous plant diversity mainly via changing the canopy density. Both the gross and net effects of aspect were statistically insignificant. In theory, more southerly aspects receive more solar radiation, thereby increasing species diversity. However, the southfacing sample plots experienced more severe drought stress, which may restrain plant growth and hence can explain the insignificant effect of aspect.

(2) Soil factors

The CCA results indicated that sample plots with lower moisture content, pH, total P, and total K, and higher total N and total C generally displayed a higher diversity. Five of the six factors (total C excluded) showed significant gross effects. Soil moisture was the main source of water needed for vegetation growth, and P and K were the necessary nutrients. The soil on the five islands is relatively dry and barren. A high biodiversity often represents higher productivity, and also implies that the community can sufficiently absorb the water and nutrient elements in soil, resulting in a reduction in moisture content and nutrients (Elton, 2000). Most soil nutrients are more effective in acidic environments (Khattak and Hussain, 2007). Our results also revealed higher diversity at the sample sites with lower pH. Furthermore, our study demonstrated that all of the above factors failed to exert a significant net effect on diversity. In fact, soil and other environmental factors interacted closely and exerted an integrated effect on herbaceous plant growth; single soil factor often exerted less of an effect when its interaction with other factors was not considered. Total N showed a significant gross effect, whereas both the gross and net effects of total C were insignificant; the above two factors had an extremely significant positive correlation with each other, and were significantly positively correlated with canopy density. The C and N levels in the soil are closely correlated with ecosystem condition (Feng et al., 1999); an increase in the number of trees often leads to increased litter, which is the major source of soil C and N elements.

(3) Plantation factors

The effect of plantation establishment on biodiversity is controversial. Some studies have suggested that plantation species are monotonous and mostly consist of highly invasive species, which is a threat to biodiversity maintenance (Li et al., 2009, 2013). Meanwhile, other studies have shown that there is no significant difference between native and plantation forests, which in some cases have a positive impact on biodiversity (Allen et al., 1995). In our study, the CCA ordination and ANOVA results both indicated that canopy density significantly influenced herbaceous plant diversity. Sample plots with greater canopy density generally displayed a higher diversity, and α diversity generally showed a ‘decreaseincrease-decrease’ trend with an increase in canopy density, suggesting that when plantations had minimal impact, the sample plots showed a higher diversity. With an increase in plantation intensity, diversity declined to its minimum, but with a further increase in plantation scale, herbaceous plant diversity increased gradually, reached its maximum, and then declined to certain extent with a further increase in plantation intensity.

In reality, the process of plantation impact on biodiversity is more complex, and the status of a plantation has a significantly different impact on the diversity of understory plants. The plantation process is also restricted by ecological conditions. Therefore, it is vital to carry out studies on the effects of plantations on native plant diversity in different regions, under different construction and ecological conditions. Although many researchers have conducted highly effective investigations, continued research is clearly needed. Understanding the effects of plantations on island biodiversity is of great importance because island ecosystems are vulnerable. Our results revealed that canopy density, which represents plantation status, was the only factor that had a significant net effect on species diversity; moreover, it was also significantly correlated with altitude, slope, and all of the soil factors. Altitude and slope are limiting factors for plantations. People are inclined to choose areas with high altitudes and slopes to develop plantations; such areas are not suitable for town construction and farmland reclamation. Additionally, plantations clearly affect soil properties. In our study, terrain factors affected species diversity by restricting plantation; plantation influenced soil properties and the two together affected diversity. Therefore, plantation was ultimately the fundamental driving factor of spatial heterogeneity in herbaceous plant diversity on the five islands.

(4) Marine factor

CCA ordination and ANOVA both indicated that distance to the shore did not significantly affect species diversity. In general, the sample plots closer to the shore are more affected by the sea. However, our study showed that the diversity of sample plots at the island edge did not differ significantly from that of plots in the island’s interior. This demonstrated that either marine conditions do not play a major role in the spatial distribution of diversity or their effects are cut off and weakened by hills and forest canopy, and thus made no difference.

5 CONCLUSION

This study investigated herbaceous plants on five southern islands in the Miaodao Archipelago. Based on the on-site survey, the spatial distribution of herbaceous plant diversity was analyzed, and impact factors and their impact degree on herbaceous plant diversity were further explored by CCA and ANOVA. Our conclusions are as follows:

(1) The field survey recorded 114 species of herbaceous plant, with Cleistogenes chinensis, Carex lanceolata, and Artemisia argyi as the dominant species. The rarefaction curve indicated that a reasonable number of sample plots was taken. The average Shannon-Wiener and Pielou indices of all samples plots were 2.02 and 0.93, respectively; the Jaccard indices between different islands were all lower than 0.5.

(2) The total number of species on different islands was significantly positively correlated with island area. The average α diversity was correlated with human activities, while the β diversity between islands was more affected by island area than mutual distances.

(3) All of the CCA axes were statistically significant. The cumulative percentage variance of species-environment relation reached 56.5%, providing rich ecological information; the CCA diagrams indicated that sample plots with higher altitude, slope, total N, total C, canopy density and lower moisture content, pH, total P, total K, aspect, generally exhibited higher diversity, and distance to the shore did not significantly effect diversity. Among the environmental factors, pH, canopy density, total K, total P, moisture content, altitude, and slope had significant gross effects, whereas only canopy density showed a significant net effect.

(4) Terrain, soil, and plantation factors all played important roles in the spatial heterogeneity of herbaceous plant diversity, while marine factor made no difference. Terrain factors affected diversity by restricting plantation, plantation influenced soil properties, and the two together affected diversity. Therefore, plantation was ultimately the fundamental driving factor of spatial heterogeneity in herbaceous plant diversity on all five islands.

References
Allen R B, Platt K H, Coker R E J, 1995. Understorey species composition patterns in a Pinus radiata D. Don plantation on the central North Island Volcanic Plateau, New Zealand.New Zealand Journal of Forestry Science, 25 : 301 –317.
Anten N P R, Hirose T, 1999. Interspecific differences in above-ground growth patterns result in spatial and temporal partitioning of light among species in a tall-grass meadow. Journal of Ecology, 87 (4) : 583 –597. Doi: 10.1046/j.1365-2745.1999.00365.x
Bustamante-Sánchez M A, Armesto J J, 2012. Seed limitation during early forest succession in a rural landscape on Chiloé Island, Chile:implications for temperate forest restoration. Journal of Applied Ecology, 49 (5) : 1103 –1112. Doi: 10.1111/jpe.2012.49.issue-5
Cardinale B J, Srivastava D S, Duffy J E, Wright J P, Downing A L, Sankaran M, Jouseau C, 2006. Effects of biodiversity on the functioning of trophic groups and ecosystems. Nature, 443 (7114) : 989 –992. Doi: 10.1038/nature05202
Chi Y, Shi H H, Wang X L, Li J, Feng A P. 2015. The spatial distribution and impact factors of net primary productivity in the island ecosystem of five southern islands of Miaodao Archipelago. Acta Ecologica Sinica, 35(23), http://dx.doi.org/10.5846/stxb201405040880. (in Chinese with English abstract)
Chu Q W, Zhang H Q, Wu Y W, Feng Z K, Chen B, 2013. Application research of Landsat-8. Remote Sensing Information, 28 (4) : 110 –114.
Colwell, R K. 1997. Estimate S:statistical estimation of species richness and shared species from samples. Version 5. User's Guide and application published at:http://viceroy.eeb.uconn.edu/estimates
Curtis J T, Mcintosh R P, 1951. An upland forest continuum in the prairie-forest border region of wisconsin. Ecology, 32 (3) : 476 –496. Doi: 10.2307/1931725
Elton C S, 2000. The Ecology of Invasions by Animals and Plants. University of Chicago Press, Chicago (IL).
Fang J Y, Wang X P, Shen Z H, Tang Z Y, He J S, Yu D, Jiang Y, Wang Z H, Zheng C Y, Zhu J L, Guo Z D, 2009. Methods and protocols for plant community inventory. Biodiversity Science, 17 (6) : 533 –548. Doi: 10.3724/SP.J.1003.2009.09253
Feng Z W, Wang X K, Wu G, 1999. Biomass and Productivity of Forest Ecosystem in China. Science Press, Beijing, China.
Gallagher R V, Leishman M R, Miller J T, Hui C, Richardson D M, Suda J, Trávníček P, 2011. Invasiveness in introduced Australian acacias:the role of species traits and genome size. Diversity and Distributions, 17 (5) : 884 –897. Doi: 10.1111/j.1472-4642.2011.00805.x
Gao M, Wang X X, Hui C, Yi H P, Zhang C Q, Wu X Q, Bi X L, Wang Y, Xiao L X, Wang D, 2014. Assembly of plant communities in coastal wetlands-the role of saltcedar Tamarix chinensis during early succession. Journal of Plant Ecology, 8 (5) : 539 –548.
Gotelli N J, Colwell R K, 2001. Quantifying biodiversity:procedures and pitfalls in the measurement and comparison of species richness. Ecology Letters, 4 (4) : 379 –391. Doi: 10.1046/j.1461-0248.2001.00230.x
Halpern B S, Walbridge S, Selkoe K A, Kappel C V, Micheli F, D'Agrosa C, Bruno J F, Casey K S, Ebert C, Fox H E, Fujita R, Heinemann D, Lenihan H S, Madin E M P, Perry M T, Selig E R, Spalding M, Steneck R, Watson R, 2008. A global map of human impact on marine ecosystems. Science, 319 (5865) : 948 –952. Doi: 10.1126/science.1149345
Harrison S, Ross S J, Lawton J H, 1992. Beta diversity on geographic gradients in Britain. The Journal of Animal Ecology, 61 (1) : 151 –158. Doi: 10.2307/5518
Heywood V H, 1995. Global Biodiversity Assessment. Cambridge University Press, Cambridge, UK.
Hillier A. 2011. Manual for Working with ArcGIS 10.University of Pennsylvania, Pennsylvania.
Hooper D U, Chapin F S, Ewel J J, Hector A, Inchausti P, Lavorel S, Lawton J H, Lodge D M, Loreau M, Naeem S, 2005. Effects of biodiversity on ecosystem functioning:a consensus of current knowledge. Ecological Monographs, 75 (1) : 3 –35. Doi: 10.1890/04-0922
Hui C, McGeoch M A, 2014. Zeta diversity as a concept and metric that unifies incidence-based biodiversity patterns. T he American Naturalist, 184 (5) : 684 –694. Doi: 10.1086/678125
Jiang Y, Kang M Y, Zhu Y, Xu G C, 2007. Plant biodiversity patterns on Helan Mountain, China. Acta Oecologica, 32 (2) : 125 –133. Doi: 10.1016/j.actao.2006.12.003
Katovai E, Burley A L, Mayfield M M, 2012. Understory plant species and functional diversity in the degraded wet tropical forests of Kolombangara Island, Solomon Islands. Biological Conservation, 145 (1) : 214 –224. Doi: 10.1016/j.biocon.2011.11.008
Khattak R A, Hussain Z, 2007. Evaluation of soil fertility status and nutrition of orchards. Soil & Environment, 26 (1) : 22 –32.
Li J L, Zhang J T, Zou C H, Liu Z Y, 2012. Type and sort of plant communities in Putuo Mountains under tourism development. Scientia Silvae Sinica, 48 (7) : 174 –181.
Li M Y, Zhang J J, Guo B N, Huang M, Ru H, 2013. Understory plant species diversity and hydrological effect of Pinus tabulaeformis plantations with different stand densities in Loess Plateau of Western Shanxi, China. Chinese Journal of Ecology, 32 (5) : 1083 –1090.
Li S, Luo Y Q, Wu J, Zong S X, Yao G L, Li Y, Liu Y M, Zhang Y R, 2009. Community structure and biodiversity in plantations and natural forests of seabuckthorn in southern Ningxia, China. Forestry Studies in China, 11 (1) : 49 –54.
Ma K P, 2013. Studies on biodiversity and ecosystem function via manipulation experiments. Biodiversity Science, 21 (3) : 247 –248. Doi: 10.3724/SP.J.1003.2013.02132
Ma K P, Liu Y M, 1994. Measurement of biotic community diversity I α diversity (Part 2). Biodiversity Science, 2 (4) : 231 –239.
Mac Nally R, Fleishman E, Bulluck L P, Betrus C J, 2004. Comparative influence of spatial scale on beta diversity within regional assemblages of birds and butterflies. Journal of Biogeography, 31 (6) : 917 –929. Doi: 10.1111/jbi.2004.31.issue-6
MacArchur R H, Wilson E O, 1963. An equilibrium theory of insular zoogeography. Evolution, 37 (4) : 373 –387.
MacArchur R H, Wilson E O, 1967. The Theory of Island Biogeography. Princeton University Press, Princeton, NJ.
Michelsen O, McDevitt J E, Coelho C R V, 2014. A comparison of three methods to assess land use impacts on biodiversity in a case study of forestry plantations in New Zealand. The International Journal of Life Cycle Assess ment, 19 (6) : 1214 –1225. Doi: 10.1007/s11367-014-0742-1
Nekola J C, White P S, 1999. The distance decay of similarity in biogeography and ecology. Journal of Biogeography, 26 (4) : 867 –878. Doi: 10.1046/j.1365-2699.1999.00305.x
Norušis M J. 2010. PASW Statistics 18 Guide to Data Analysis.Prentice Hall Press, Englewood Cliffs.
Qian H, 2009. Global comparisons of beta diversity among mammals, birds, reptiles, and amphibians across spatial scales and taxonomic ranks. Journal of Systematics and Evolution, 47 (5) : 509 –514. Doi: 10.1111/j.1759-6831.2009.00043.x
Qian Y Q, Ma K P, 1994. Principle and Methods of Biodiversity Research. Chinese Science & Technology Press, Beijing, China.
Qie L, Lee T M, Sodhi N S, Lim S L H, 2011. Dung beetle assemblages on tropical land-bridge islands:small island effect and vulnerable species. Journal of Biogeography, 38 (4) : 792 –804. Doi: 10.1111/jbi.2011.38.issue-4
Ren X M, Yang G H, Wang D X, Qin X W, Liu Z X, Zhao S X, Bai Y, 2012. Effects of environmental factors on species distribution and diversity in an Abies fargesii-Betula utilis mixed forest. Acta Ecologica Sinica, 32 (2) : 605 –613. Doi: 10.5846/stxb
Rosoman G. 1994. The Plantation Effect. Greenpeace New Zealand, Auckland. 48p.
Schulze E D, Mooney H A, Sala O E, Jobbagy E, Buchmann N, Bauer G, Canadell J, Jackson R B, Loreti J, Oesterheld M, Ehleringer J R, 1996. Rooting depth, water availability, and vegetation cover along an aridity gradient in patagonia. Oecologia, 108 (3) : 503 –511. Doi: 10.1007/BF00333727
Schulze E D, Mooney H A. 1993. Biodiversity and Ecosystem Function. Springer, Berlin Heidelberg.
Shi H H, Wang X L, Wang A, Liu Z Y, Ma D M, 2013. Storage and drivers of forests carbon on the Beichangshan Island of Miaodao Archipelago. Acta Ecologica Sinica, 33 (19) : 6362 –6372.
Shi H H, Zheng W, Ding D W, Lü J B, 2009. Valuation and ecosystem services of typical island. Marine Environmental Science, 28 (6) : 743 –748.
Soininen J, McDonald R, Hillebrand H, 2007. The distance decay of similarity in ecological communities. Ecography, 30 (1) : 3 –12. Doi: 10.1111/eco.2007.30.issue-1
Tilman D, Knops D, Wedin P B, Reich P, Ritchie M, Siemann E, 1997. The influence of functional diversity and composition on ecosystem processes. Science, 277 (5300) : 1300 –1302.
Tilman D, Reich P B, Knops J M H, 2006. Biodiversity and ecosystem stability in a decade-long grassland experiment. Nature, 441 (7093) : 629 –632. Doi: 10.1038/nature04742
Tilman D, Reich P B, Knops J, Wedin D, Mielke T, Lehman C, 2001. Diversity and productivity in a long-term grassland experiment. Science, 294 (5543) : 843 –845. Doi: 10.1126/science.1060391
Tilman D, 2000. Causes, consequences and ethics of biodiversity. Nature, 405 (6783) : 208 –211. Doi: 10.1038/35012217
Wang X L, Wang A, Shi H H, Zheng W, Zhou R, 2014. Discussion of carbon sequestration estimates in the island terrestrial ecosystems. Acta Ecologica Sinica, 34 (1) : 88 –96.
Zhang J T, 2004. Quantitative Ecology. Science Press, Beijing, China.
Zheng W, Shen C C, Qiao M Y, Shi H H, 2014. Contingent valuation of preserving ecosystem of Changdao Island Nature Reserve. Acta Ecologica Sinica, 34 (1) : 82 –87.