Annals of Forest Science

, 75:84 | Cite as

Demographic strategies of a dominant tree species in response to logging in a degraded subtropical forest in Southeast China

  • Xianyu Yang
  • Shouzhong LiEmail author
  • Baocheng Shen
  • Yuyan Wu
  • Suitao Sun
  • Rong Liu
  • Ruibo Zha
  • Shou-Li Li
Research Paper


Key message

The demography of pioneer tree species ( Pinus massoniana Lamb.) is significantly affected by logging in Southeast China. Logging negatively affects the population growth rate of P. massoniana , which facilitates the growth of individual trees but has no effect on reproduction probability. The survival and growth of seedlings contribute the most to population growth.


Subtropical forest degradation caused by unreasonable disturbances is closely related to anthropogenic activities in Southeast China, and the frequent small-scale logging activity by local people was the dominated disturbance regime in forests in this region over the past several decades.


The objective of this study is to evaluate the demographic consequences of logging on Pinus massoniana, a pioneer tree species, at individual level (survival, growth, and fecundity) and population level (the population growth rate and size distribution) over short-term period.


The size of tree individuals was combined with vital rates using various modeling approaches based on demographic data from three annual censuses. The integral projection model (IPM) was constructed and used to conduct comparative demographic analyses.


Logging negatively affected the population growth rate: from a slight expansion before logging to a moderate decline after logging. This study found a significant reduction in seedling recruitment after logging, and plant growth and mortality were slightly enhanced. The survival of seedlings greatly contributes to population growth rate compared to other life stages for both periods (before and after logging) while its relative importance decreases after logging. Seedling growth is also important to population growth, and its relative importance increased after logging. Shrinkage and fecundity have a minimal contribution effect on the population growth rate.


Growing plants in a nursery with a similar demography to P. massoniana could be beneficial for pioneer species regeneration in that this will improve the survival rate and growth of small individuals after logging.


Pinus massoniana Logging Integral projection model Elasticity Degraded subtropical forest Pioneer tree demography 

1 Introduction

Many plant species subsist in disturbed environments characterized by frequent anthropogenic activities or natural hazards (Benavides et al. 2013; Caswell 2000) Habitat disturbances, acting as a major selective force, may cause variation in vital rates (survival, growth, reproduction, and mortality) and, furthermore, influence the performance and persistence of plant populations (Easterling et al. 2000). While changes in vital rates before and after habitat disturbances have been well documented (Alyemeni and Sher 2010; Grogan et al. 2014; Mandle and Ticktin 2012; Schmidt et al. 2011; Yang et al. 2012), little is known about the demographic consequences of such variations at a population level. Forests are typically under tremendous pressure from human development, especially in degraded land (Ma et al. 2016). The mortality, growth, and recruitment of tree species are key factors that influence the population and successional dynamics of forests (Smith and Nichols 2005; Winkler et al. 2015). Identifying and understanding changes in plant demography are critical to understanding the numerous ecological processes that exist and provide guidelines for the sustainable management of forests (Grogan et al. 2014; Primack and Lee 1991).

Logging is one of the most common human disturbances of trees and is a commonly use practice in forest exploitation. It is increasingly being embraced to protect the integrity of forests (Dekker and de Graaf 2003; Egnell and Valinger 2003). The response of tree populations is closely associated with logging patterns (logging intensity, logging cycles, logging mode, logging number), species characteristics, and stand density. Previous studies have shown that the structure of populations could significantly change in the presence of dead trees caused by logging over a short time period, which, furthermore, could have an impact on plant demography over a long time period under a decrease in reproductive trees and seedling recruits (Carvalho et al. 2013; Villela et al. 2006). The resultant effects of logging differ among different species groups. For light-demanding tree species, studies have indicated that logging could promote plant growth and recruitment due to an increase in light availability after logging (Figueira et al. 2008; Smith and Nichols 2005). Generally, moderate logging practices have been shown to have the highest positive effects on stem density and basal area (Behjou and Mollabashi 2017; Qi and Zhao 2015). On the other hand, the proportion of pioneer species could increase in logged forests due to the creation of more suitable habitat conditions. It has been shown that the successful regeneration of pioneer species has a positive correlation to canopy gaps (Arets et al. 2003). However, for shade-tolerant species, logging could have a significant impact on population dynamics through a reduction in recruitment, which could even cause extinction if exposed to periodic regime (Yamada et al. 2016).

Subtropical evergreen broad-leaved forests are the most extensive forest type in China, especially in the southeast, which is largely deforested and fragmented due to human activities and economic development (Gong et al. 2013; Sun et al. 2006). In this region of China, vegetation degradation is closely related to the improper management practices associated with anthropogenic activities, including logging, fires, and severe reclamation (Hua et al. 2015; Jin et al. 2015). In combination with natural factors like fragmented topography, soil erodibility, and concentrated rainfall, the excessive exploitation of forests has further led to desertification over the last century (Dou et al. 2013; Xu et al. 2013; Vajpeyi and Ponomarenko 2001). To address these environmental problems, the Government of China instigated an extensive reforestation program, which has been widely adopted throughout the country over recent decades (He et al. 2011; Ma and Zha 2008). This has yielded some results, such as the restraint in large-scale irrational exploitation under the auspice of ecological protection policies.

Although forest cover has greatly increased after the implementation of these reforestation programs, most secondary forests are monoculture plantations, where a small variety of commercial species are selected, such as Pinus and Cunninghamia (Ren et al. 2007; Xu 2011). Pinus massoniana, a dominant pioneer tree species, occurs widely in southeast China at present due to its high tolerance of poor soil nutrient conditions, high sunlight demand, and longevity (Dou et al. 2013; Liu and Zeng 2017). The demography of P. massoniana and its population dynamics could signify the state of succession and forest restoration in degraded lands. However, frequent small-scale logging is still prevalent due to our demand for energy and economic well-being with increased human population (An-Ping et al. 2011; Zackey 2015). Previous studies have indicated that the performance of P. massoniana could be significantly influenced by logging intensity. The diameter growth increment and the canopy area in logged forests (where 60–80% individuals were removed) are higher than those in unlogged forests, but they decrease with an increase in logging intensity (Zeng et al. 2017). A 20–30% logging intensity had the greatest positive effect on the diameter at breast height (DBH) and height (Huang et al. 2013; Ma et al. 2017). The regeneration of pine trees could also be improved by low-intensity logging under higher light availability (Qi and Zhao 2015). However, unlike studies conducted in forests long after logging has occurred, demographic studies conducted in forests immediately after logging are inadequate in many regions, particularly in degraded subtropical forests. To fill this knowledge gap, research on the demographic response of P. massoniana populations under logging activities is essential to gain a complete understanding of sustainable forest management.

In this study, we investigate the effects of forest logging in Changting County, Fujian Province, southeast China. We examine the population dynamics of a dominant tree species P. massoniana and investigate the effects of logging on tree mortality, growth and fecundity. We attempt to test whether the population remained stable throughout the study period, as well as to identify and explain vital rate changes. We used the integral projection model (IPM) to provide insight into the response of the entire life cycle of P. massoniana by answering three main questions: (i) Do the vital rates of P. massoniana differ during periods before and after logging? (ii) How does logging affect population growth? (iii) Which demographic components are essential for population growth before and after logging? Following this, we provide recommendations for degraded subtropical forests management based on the implications of our findings.

2 Materials and methods

2.1 Experimental site and study species

This study was conducted in Changting County (25° 18′ 40″26° 02′ 05″ N, 116° 00′ 45″116° 39′ 20″ E) located in the southern region of the Wuyi Mountains in Southeast China (Ning et al. 2016). As shown in Appendix Fig. 5, the inter-annual average temperature in July and January is 27 and 8 °C, respectively. Summer is long and hot, where the maximum air temperature can reach 35 °C. Annual rainfall is 1710 mm (± 150 mm SD) with the most rainfall occurring March and June (Xu et al. 2013). The site is characterized by red tinged acidic soil with a high quartz content that is poorly resistant to corrosion and is lacking in fertility (Dou et al. 2013). As a typical soil erosion region example in Southeast China, the Government of China has allocated much effort to vegetating and re-vegetation this region since early 1980s (Xie et al. 2002). After three decades of silvicultural practices, secondary forests dominated by P. massoniana are widely distributed within the study area, where vegetation coverage has increased from 48% in 1985 to 86.5% in 2013 (Dou et al. 2013; Zhang et al. 2015).

Pinus massoniana (Pinaceae) is a coniferous tree species (Fan et al. 2008). Its height ranges from 15 to 45 m, and its DBH can reach 1.5 m, with a broad, rounded crown with long branches. The bark is thick and flakes higher up on the trunk. It grows in dry, poor, and acidic soils (Dong 1987). Annual DBH growth increases with age and reaches its peaks after approximately 15–20 years. It is a dominant species during the early succession stages of evergreen broad-leaved forests in the subtropical zone of Southeast China. It produces seeds at 5 to 6 years old. Pine cones ripen in October, and seeds germinate in the following March. The seeds have one wing and are readily dispersed by wind (Chen and Su 2001).

2.2 Data collection

In 2011, one permanent plot approximately 900 m2 (30 × 30 m) in size was established, where P. massoniana was the main dominant tree species. The total vegetation cover was 65 ± 3% (mean ± SE). To ease data collection, the plot was subdivided into 36 subplots of 5 × 5 m each. Measurements were taken within the entire plot in July 2011, 2012, and 2013. For each individual P. massoniana tree, basal diameter and total height were measured and labeled during the first census we conducted. For those individuals with heights greater than 1.5 m, DBH was also measured and recorded. In subsequent years, all labeled individuals were checked and re-measured, and new recruits (seedlings that germinated and established themselves between census periods) within the plot were searched for, labeled, and measured. The number of pine cones produced throughout the study period was also recorded. A total of 522 individuals were identified and measured throughout the study period, where the minimum size of basal diameter and height was less than 1 mm and 5 cm, respectively. The reason for calculating seedling recruitments based on measured plant size, instead of using measured pine cones or the number of pine nuts, was because of the following: (1) pinecones could remain attached to the tree until the second year after pine nuts have fallen off in the first year, thereby potentially causing measuring errors, where the same pine cones are counted twice in different years; (2) pine cones of P. massoniana usually produce an enormous amount of pine nuts, which makes it impossible to quantifiably count the number of pine nuts of each individual tree. Therefore, we assumed that the fecundity of a reproductive individual is positively associated with plant size, which is described by the basal diameter and the number of seedling produced per individual (fe(x)) and was calculated as fe(x) = x(n/sum(x)), where n is the number of established seedlings and x is the basal diameter of established seedlings in the plot.

We collected soil in the 0–10 cm layer, the 10–20 cm layer, and the 20–40 cm layer in the plot during the census period and measured the weight of wet soil and dry soil, which we died at 105 °C for 8 h. The soil water content was calculated by the weight of water in soil divided by the weight of wet soil. The ground surface temperature was also collected by means of surface thermometer during the census period.

The sample plot was logged in October 2012 after the second census, and 81% of large individuals (basal diameter > 50 mm) were removed, where 84% of the total number of individuals remained (Appendix Figs. 6 and 7). The number of P. massoniana individuals decreased from 1700 trees/ha before logging to 611 trees/ha after logging (calculated using trees with a measured DBH).

2.3 Statistical analyses

To test the differences in the vital rates during the different logging periods, the current plant size (x, at t) was used in conjunction with the future size (y, at t + 1), applying multiple regression models. We used basal diameter (BD) to characterize plant size and to construct the IPM, due to its higher R2 values compared to other size measurements (DBH and height) for all regression models. Paired t tests were used to test changes in temporal size (g(y, x)) between current and future basal diameters (sizet + 1-sizet), and liner models were used to examine the effects of logging on future basal diameter (μ(x)) and variance of growth (σ2(x)). At the same time, logistic regressions were used to fit the probability of survival s(x) and reproduction Pr(x) with the current basal diameter as explanatory variables.

The IPM describes the dynamics of a population structured by a continuous individual-level state variable as follows (Easterling et al. 2000; Ellner and Rees 2006):
$$ {\displaystyle \begin{array}{c}n\left(y,t+1\right)={\int}_{\Omega}\left[p\left(y,x\right)+f\left(y,x\right)\right]n\left(x,t\right) dx\\ {}={\int}_{\Omega}k\left(y,x\right)n\left(x,t\right) dx\end{array}} $$
where Ω is the range of all plant sizes used in the calculation. The kernel k(y, x) is a nonnegative surface and represents all possibilities of transitions from size x to size y throughout the range of time t to time t + 1, which is composed of p(y, x) and f(y, x). p(y, x)dx represents the probability of living individuals growing from size x to size y during the time interval (t, t + 1), which is calculated as p(y, x) = s(x) g(y, x). s(x) is the probability that a size x individual survives to the next time step. g(y, x) is a normal probability density function with a mean future plant size (μ(x)) and a growth variance (σ2(x)). Subsequently, f(y, x)dx represents the number of newborns at time t + 1 per size x individuals alive at time t, which is calculated as f(y, x) = Pr(x) fe(x) fd(y), where Pr(x) is the reproducing probability that a size x plant will reproduce in one time step, fe(x) is the number of seedlings produced on average per plant of size x, and fd(y) is the normal distribution of offspring size.

We converted the kernel k(y, x) into a large transition matrix K(y, x), using the midpoint rule (Ellner and Rees 2006). According to the midpoint rule, we divided the size range of individuals to m mesh points, which means the matrix also has m size classes. In theory, the fitting precision of the matrix increases with m. However, the category width could affect the population growth rate. For example, the population growth rate was considerably overestimated when using ≤ 10 classes and was slightly underestimated when using ≥ 1000 classes (Zuidema et al. 2010). Therefore, we used 100 size classes, which have also been commonly used in the literature. We constructed IPMs for the period before and after logging, separately. IPMs can yield outputs of population growth rates (λ), stable size distributions, sensitivity, and elasticity. Confidence intervals of λ were calculated by bootstrapping (Jongejans et al. 2010). We resampled with replacements from the dataset (n = 522 in our study), re-calculated regression coefficients, and established the kernel and calculated. This was repeated 5000 times, and 95% confidence intervals were obtained from the frequency distribution of these values (Li et al. 2011).

The stable population distribution that derived from IPM was compared to the observed population structure (means of before logging in 2011–2012 and after logging in 2013) using the percentage similarity index (PS). PS = ∑(min[obsi, ssdi]) × 100, where obsi and ssdi are vectors of observed size distributions and stable population structures, respectively (both vectors scaled to sum to 1) (Zuidema et al. 2007; Li et al. 2015). High values of this index are indicative of a high level of similarity and also a more stable population state.

Elasticity analysis quantifies the proportional change in λ resulting from infinitesimal relative changes in the matrix elements (de Kroon et al. 2000; Li et al. 2015), which indicates the relative importance of individuals and vital rates (survival, growth, shrinkage, and fecundity) to the population growth rate.

The survival rate would decrease with basal diameter if logged trees were added to the calculation, and, furthermore, the output of the IPM model may deviate from the trajectories of population dynamics in the long term. To avoid such errors, we therefore removed it from the data in 2012 when we constructed the IPM 2012–2013 model. All analyses were conducted using R version 3.4.1 software (R development core Team 2017).

Data availability

The datasets generated during and/or analyzed during the current study are available in the PANGAEA repository, (Yang et al. 2018). Datasets are not peer-reviewed.

3 Results

3.1 Vital rate variations

During the period before logging (2011–2012), the survival probability of P. massoniana seedlings (with a basal diameter of 0–10 mm) was very high (greater than 82%; Table 1; Fig. 1a). The probability of survival increased with stem size, which was above 99.9% for juvenile (with a basal diameter of 10–50 mm) and adult (with a basal diameter > 50 mm) plants. Survival probabilities were lower for seedlings and juveniles after logging (2012–2013). The survival of seedlings and juveniles was decreased by 20 and 7%, respectively.
Table 1

Statistical models and parameter estimates used to construct the kernels for the integral projection models of Pinus massoniana in a degraded subtropical forest during 2011–2012 (period 1) and 2012–2013 (period 2)

Demographic parameters



Survival probability (s)


Logit (s) = 1.296(0.638) + 0.573(0.240)x

n = 381, P < 0.05


Logit (s) = 0.850(0.159) + 0.057(0.017)x

n = 402, P < 0.01

Reproductive probability (Pr)


Logit (Pr) = − 4.220(0.399) + 0.058(0.006)x

n = 381, P < 0.001


Logit(Pr) = − 5.296(0.594) + 0.071(0.011)x

n = 402, P < 0.001

Future size (μ)


μ = 1.046(0.186) + 1.040(0.003)x

n = 369, R2 = 0.995, P < 0.001


μ = 1.064(0.120) + 1.066(0.005)x

n = 310, R2 = 0.994, P < 0.001

Variance of growth (σ2)


σ2 = 2.907(3.415) + 0.148(0.070)x

n = 339, R2 = 0.009, P < 0.05


σ2 = 1.247(0.494) + 0.140(0.020)x

n = 310, R2 = 0.137, P < 0.001

Size distribution of new seedlings


Gaussian with mean = 2.070

Variance = 1.739, n = 119


Gaussian with mean = 2.974

Variance = 4.372, n = 22

The models are functions of plant basal diameter (x, mm), treatment (logging), and interactions between basal diameter and treatment, using the unlogged period as a reference. Values in parentheses are standard errors of the parameter estimates. P < 0.01 indicates that slope and intercept in the regression models are significantly different from zero

Fig. 1

Relationships between vital rates and plant the basal diameter for Pinus massoniana in a degraded subtropical forest during 2011–2012 (a, c, e) and 2012–2013 (b, d, f). Regression functions are described in Table 1. BD basal diameter

Stem growth increased with basal diameter as indicated by the regressions with slopes > 1 both before and after logging (Table 1; Fig. 1c, d). Logging slightly stimulated the growth of the population: mean growth increments varied from 1.81 ± 0.12 mm/year before logging to 2.06 ± 0.13 mm/year after logging. The growth increment varied significantly between life stages. For example, adults grew fastest (6.46 ± 0.72 mm/year) after logging, which was higher than juveniles and seedlings (3.79 ± 0.33 and 1.12 ± 0.71 mm/year, respectively). At the same time, the proportion of steams with negative growth rates was 4.0% prior to logging and 4.5% following logging.

The reproductive probability increased with basal diameter, and it was not significantly affected by logging (Fig. 1e, f). The number of reproductive trees was 381 and 402 throughout the study period. Plants generally started to reproduce at a basal diameter greater than 70 mm (> 50% reproductive probability). The number of pine cones per basal area significantly decreased from 0.2280 before logging (a total of 1551 pine cones) to 0.1920 after logging (a total of 110 in total pine cones). Furthermore, new seedling recruits declined by 82% after logging (Table 1; Appendix Fig. 8), suggesting a strong logging influence on seedling recruitment.

3.2 Population growth rate and stable structures

In our study, the population growth rate (λ) of P. massoniana before logging was 1.065, which showed that the population was expanding. The number of individuals increased by 30%, where the maximum size changed from 153.66 to 164.5 mm (Table 2; Appendix Fig. 6). However, the P. massoniana population declined (λ = 0.946) after logging with numbers decreasing by 14% (Table 2; Appendix Fig. 6). Logging reduced the numbers of seedlings by 24%, where the maximum size changed to 127.9 mm (Appendix Fig. 6). While the number of seedlings decreased by 24% after logging, the proportion of seedlings in population increased from 59.7 to 65.9% throughout the study period.
Table 2

Population growth rates of Pinus massoniana in a degraded subtropical forest during 2011–2013



Population growth rate

Before logging


1.065 [1.056, 1.074]

After logging


0.946 [0.943, 0.949]

Values in brackets refer to 95% confidence intervals

The kernels showed a higher positive growth transition and a lower fecundity transition after logging (Fig. 2a, b). Before and after logging, both showed that the observed population structure of P. massoniana was characterized by a high abundance of seedlings and a low abundance of adults (Fig. 3). The similarity between the stable size distribution and the observed size distribution was 69.4% after logging, lower than the similarity (71.4%) before logging (Fig. 3).
Fig. 2

Fitted kernels of transitions for Pinus massoniana in a degraded subtropical forest during 2011–2012 (a) and 2012–2013 (b). Transition values are shown for transitions from present size (x-axis) to future size (y-axis). Gray tones indicate the magnitude of the transitions with values > 0.1 shown in white. Transitions along the diagonal represent survival-growth transitions, while those on the bottom correspond to recruitment

Fig. 3

Number of plants (Pinus massoniana) per square meter in degraded subtropical forest throughout the study period. The observed population structure (Observed) from study plot data in 2012 (prelogging) and 2013 (postlogging). Stable size structure (SSD) resulting from the integral projection model for 2011–2012 (prelogging) and 2012–2013 (postlogging)

3.3 Elasticity analysis

The survival of seedlings and juveniles yielded highest elasticity values both before and after logging (Fig. 4), which implied that seedlings and juveniles are critically important for the population maintenance of P. massoniana. Fecundity and shrinkage yielded much lower elasticity values, which suggested that the population growth rate was less sensitive to changes in these vital rates. Vital rate elasticity changed between periods (Fig. 4): the survival elasticity of seedlings was lower after logging compared to before logging, and the growth elasticity of seedlings was higher. There were minimal contributions of shrinkage and fecundity to the population growth rate.
Fig. 4

Vital rate elasticity for Pinus massoniana in a degraded subtropical forest before logging (a) and after logging (b)

4 Discussion

4.1 Effects of logging on vital rates

The quality of plant survival is the result of both biotic and abiotic factors (Connell and Green 2000; Zhang et al. 2009). Logging could both have a direct and indirect effect on trees survival, through adjustment of relationships among individuals and the change in habitat conditions in the forest (Carvalho et al. 2013). In our study, adult survival was not affected by logging, while there was a reduction in the survival rate of seedlings and juveniles after logging. In general, seedlings are highly sensitive to habitat variables and typically present nonrandom mortality during their immature development stage, and it is also the stage of the greatest change in numbers (Fangyuan et al. 2003; Han et al. 2012; Ma et al. 2016). After logging, when most large plants were removed, larger canopy gaps formed and higher temperature variations occurred in the forest. Soil water shortages could potentially be the principle factor that caused higher mortality of small sized individuals in the heat of the summer (Caspersen and Kobe 2001; Wang et al. 2001). On the other hand, due to poor soil erodibility, small-sized plants could be buried and their roots could also be exposed after the frequent rainstorms that occur during the rainy period. In addition, trampling by humans and soil compaction during logging processes could also lead to an increase in mortality of small-sized individuals (Roovers et al. 2004; Schweitzer and Dey 2013; Waltert et al. 2002).

In general, ecosystems are subject to limited habitat resources (Donovan et al. 2000). Stem density can decrease resource competition. Tree growth can accelerate in more favorable habitats (Duclos et al. 2013; Huang et al. 2013; Yang 2004). In our study, the growth rate of P. massoniana was promoted by logging, but it differed across life stages. Being a light-demanding tree species, sunlight availability, compared to other resources, was the key factor for P. massoniana growth (Zhang et al. 2005; Zhu 2015a, b). The fastest growth observed for larger sized individuals may have resulted from their larger crowns, where they were exposed to full sunlight and subsequently accumulated more photosynthetic products in the canopy layer (Win et al. 2012). On the other hand, being a common biological characteristic of pine trees, we frequently observed the exfoliation of rimose bark in our plots. We observed shrinkage in plant diameter when the growth rate was lower than bark exfoliation. Such shrinkage in plant size may help adult plants to maintain a high survival rate in degraded habitats (Salguero-Gómez and de Kroon 2010).

Recruitment limitations for tree species mainly include source limitations, dissemination limitations, and establishment limitations (Ning et al. 2011; Ma et al. 2016; Swamy et al. 2011). The seeds of P. massoniana are equipped with a wing and can be readily dispersed by wind; therefore, the reproduction rate of P. massoniana is mainly influenced by source limitations and establishment limitations. In our plot, the reproductive probability of P. massoniana was less affected by logging, but seedling recruits significantly decreased. When logging was not taking place, we calculated a total of 575 pine cones (data not shown) and 119 seedling recruits (Table 1). The average pine cone and seedling recruit of individuals was 7.7 and 1.6, respectively. Establishment limitations could be a main factor that influenced plant recruitment during this period. However, the number of pine cones and new seedlings were calculated at 195 (data not shown) and 22 (Table 1) during the postlogging period, and the average pinecone and seedling recruits of individuals were 4.1 and 0.5, respectively. Taking into account the changes in pinecones and seedling recruits during the census period, the significant decline in the number of pine cones could be one of the main reasons for a decrease in seedling recruits after logging.

4.2 Effects of logging on population dynamics

According to our results, unlogged populations showed a slight growth rate (λ = 1.065), while logged populations showed a moderate decline (λ = 0.956), both being composed of highly abundant seedlings and a small proportion of adults. The effect of logging on population dynamics depends on the sensitivity of population growth rates to vital rates, which change under logging. The elasticity analysis revealed that population growth before and after logging exhibited similar patterns. However, the relative importance of the survival of small-sized individuals was suppressed after logging while their growth increased slightly.

Generally, pine species take a long time to reach maturity and experience a loss during this process (Franco and Silvertown 2004; Sánchez-Velásquez and Pineda-López 2009). In our study, the survival rate of large adults was less affected by logging but, on individual survival rates, was positively associated with size, and the individual survival rate is closely associated with the population growth of P. massoniana. The importance of survival for population growth demonstrates the necessity of maintaining the survival of all individuals after logging, especially for seedlings and juveniles, which dominate the population structure. The higher mortality rates of seedlings and juveniles after logging could be a critical reason for the importance of the decline in the survival rate of small-sized plants during population development. Moreover, the increased importance of growth after logging also demonstrates the necessity in promoting the growth of seedlings and juveniles.

On the other hand, pioneer tree species generally achieve regeneration by means of massive seed production and low seedling establishment (Melo and Tabarelli 2003). Seed dispersal and seedling establishment of P. massoniana are typically significantly affected by habitat fluctuations in degraded subtropical regions (Peng et al. 2004; Tan et al. 2017). In order to reduce the influence of disturbance, the population growth rate of perennial trees are typically not sensitive to changes in the vital rates that are vulnerable to habitat fluctuations (Kolb et al. 2010; Winkler et al. 2015). Due to the low proportion of annual seedling recruitment and high mortality rates, fecundity has only a limited contribution to population growth both before and after logging.

4.3 Management implications

Although its long life span allows P. massoniana population to continue for a certain period after logging, this impediment to regeneration could result in a higher extinction risk to the remnant population over the long term (Vieira and Scariot 2008). Therefore, it is necessary to take some measures to promote the recovery of the population, especially after logging. Our study found that logging had different effects on vital rates, where survival and recruitment evidently decreased and growth was stimulated after logging. Management recommendations should be based on vital rates that are mostly affected by logging and those that are most important for population growth (Li et al. 2013).

An increase in vegetation cover is being increasingly recognized as an effective means of restoring ecosystem functions in degraded areas, especially in Southeast China (Zhu 2015a, b). Early successional species such as P. massoniana are typically selected for restoration due to their high survival rates and rapid growth, which can reduce the cost of restoration and provide forest products within a short period of time (Yang et al. 2017). Since P. massoniana population growth is highly sensitive to survival, especially for small-sized individuals, a slight decrease in seedling survival would result in a significant impact to the general population. Thus, the reduction in the number of small sized individuals by logging and trampling should be prevented. On the other hand, tending treatments are considered an efficient way in which to restore forests. For example, digging holes at the base of small plants could improve water condition by collecting available rainwater, which is in alleviating drought injuries caused by summer heat conditions. Regular fertilization could also have a positive effect on the survival and growth of plants (Hu et al. 2000; Li et al. 2014; Zhang et al. 2013). The light-demanding early successional species typically performed well with high survival rates and rapid growth in a degraded site, while some mid-successional and late-successional species also can be incorporated into restoration plantings (Yang et al. 2017). As it pertains to the restoration of degraded forest ecosystems, it is important to carefully choose native species (including Schima superba, Lespedeza bicolor, and Castanopsis sclerophylla) that are adapted to local conditions (Chen et al. 2005; Pan et al. 2016). Mixed planting could improve species richness, accelerate forest succession, and elevate the ecological services in degraded forests ecosystems (Lin 2005; Yang et al. 2017). On the other hand, the regeneration of P. massoniana was driven by the massive production of seeds and low seedling establishment. Previous studies showed that demographic shifts in some species populations, which, caused by low number of recruits, have trouble surviving in the community (Ma et al. 2016). While seedling recruits had a low contribution to population growth, they could also cause a reduction in the population growth rate, especially after logging. Protecting these individuals would potentially be favorable to population growth while promoting viable population restoration measures. Field planting of larger seedlings increased postplanting care, and planting in less degraded sites can be an effective way to improve the restoration of forests (Yang et al. 2017).

Being a pioneer tree species with high fecundity, P. massoniana could highly effective in reestablishing forests after disturbances (Dong 1987; Primack and Lee 1991). However, we observed a population declined during the study period. That is to say, if forests are exposed to under continuous logging management practices, pioneer species with a similar demography to P. massoniana may experience a population decline and even disapperar altogether due to low seedling recruitment (Hastings 2001; Yamada et al. 2016). On the other hand, previous studies have shown that it can take years or even decades for the influence of logging to be seen on long-lived tree populations (Bonnell et al. 2011; Egnell and Valinger 2003). Since our study is based solely on 3-year demographic data taken from a single plot, it can only capture a part of the consequences of logging. Therefore, long-term monitoring studies at multiple sites are necessary to obtain a more comprehensive understanding of the effects of logging on pioneer species.



We would like to thank Pei-Zhi Lin and Ying Zhao for their dedicated assistance during data collection in the field.


This study was supported by the National Natural Sciences Foundation of China (grant no 31470633) and the National Key Research and Development Program of China (grant no 2017YFC05054).

Compliance with ethical standards

Conflicts of interest

The authors declare that they have no conflict of interest.


  1. Alyemeni MN, Sher H (2010) Impact of human pressure on the population structure of Persicaria amplexicaule, Valeriana jatamansi and Viola serpens the naturally growing medicinal plants in Malam Jaba, Swat, Pakistan. J Med Plants Res 4:2080–2091Google Scholar
  2. An-Ping BI, Zhu HJ, Wang DG, Yue H, Tian-Hua MA (2011) Farmers’ fuel choices and economics in a soil and water loss region bulletin of soil and water conservation. 149–154Google Scholar
  3. Arets E, Van der Hout P, Zagt R (2003) Responses of tree populations and forest composition to selective logging in Guyana. Long-term changes in tropical tree diversity: studies from the Guiana Shield, Africa, Borneo and Melanesia Tropenbos Series 22:95–115Google Scholar
  4. Behjou FK, Mollabashi OG (2017) Impact of logging intensity on stem density, basal area and biodiversity indices five years after logging in a Caspian hardwood forest. J For Sci 63:167–172CrossRefGoogle Scholar
  5. Benavides R et al. (2013) Direct and indirect effects of climate on demography and early growth of Pinus sylvestris at the rear edge: changing roles of biotic and abiotic factors Plos One 8:e59824CrossRefPubMedPubMedCentralGoogle Scholar
  6. Bonnell TR, Reyna-Hurtado R, Chapman CA (2011) Post-logging recovery time is longer than expected in an East African tropical forest. For Ecol Manag 261:855–864CrossRefGoogle Scholar
  7. Carvalho JOP, Silva JNM, Silva MG, Gomes JM, Taffarel M, Nobre DNV (2013) Mortality of girdled trees and survival of seedlings in canopy gaps after logging in an upland forest in the Brazilian Amazon Revista De Ciencias Agrarias/Amazonian. J Agric Environ Sci 56:48–52Google Scholar
  8. Caspersen JP, Kobe RK (2001) Interspecific variation in sapling mortality in relation to growth and soil moisture. Oikos 92:160–168CrossRefGoogle Scholar
  9. Caswell H (2000) Prospective and retrospective perturbation analyses: their roles in conservation biology. Ecology 81:619–627CrossRefGoogle Scholar
  10. Chen JS, Su ZX (2001) Reproductive allocation of biomass in Pinus Massoniana at Mt. JinYun. Acta Phytoecologica Sinica 25:704–708Google Scholar
  11. Chen Z, Zhu H, Xiao H, Lin H (2005) The characteristics of plant communities in eroded granite hilly land after taking soil and water conservation measures Journal of Fujian Normal University (Natural Science) 21:97–102Google Scholar
  12. Connell JH, Green PT (2000) Seedling dynamics over thirty-two years in a tropical rain forest tree. Ecology 81:568–584CrossRefGoogle Scholar
  13. de Kroon H, Van Groenendael J, Ehrlén J (2000) Elasticities: a review of methods and model limitations. Ecology 81:607–618CrossRefGoogle Scholar
  14. Dekker M, de Graaf NR (2003) Pioneer and climax tree regeneration following selective logging in Suriname. For Ecol Manag 172:183–190CrossRefGoogle Scholar
  15. Dong M (1987) Population structure and dynamics of Pinus Massoniana Lamb. on Mount Jinyun, Sichuan, China. Plant Ecol 72:35–44CrossRefGoogle Scholar
  16. Donovan LA, West JB, Mcleod KW (2000) Quercus species differ in water and nutrient characteristics in a resource-limited fall-line sandhill habitat. Tree Physiol 20:929–936CrossRefPubMedGoogle Scholar
  17. Dou X, Deng Q, Li M, Wang W, Zhang Q, Cheng X (2013) Reforestation of Pinus massoniana alters soil organic carbon and nitrogen dynamics in eroded soil in south China Ecol Eng 52:154–160CrossRefGoogle Scholar
  18. Duclos V, Boudreau S, Chapman CA (2013) Shrub cover influence on seedling growth and survival following logging of a tropical forest. Biotropica 45:419–426CrossRefGoogle Scholar
  19. Easterling MR, Ellner SP, Dixon PM (2000) Size-specific sensitivity: applying a new structured population model. Ecology 81:694–708CrossRefGoogle Scholar
  20. Egnell G, Valinger E (2003) Survival, growth, and growth allocation of planted Scots pine trees after different levels of biomass removal in clear-felling. For Ecol Manag 177:65–74CrossRefGoogle Scholar
  21. Ellner SP, Rees M (2006) Integral projection models for species with complex demography. Am Nat 167:410–428CrossRefPubMedGoogle Scholar
  22. Fan XY, Liao CY, Xie Y, Gao BS (2008) Investigation and analysis on the growth of Pinus slyvestris var. momgolica in southeast Maowusu desert areas. Journal of Northwest Forestry University 23:112–116Google Scholar
  23. Fangyuan YU, Xizeng XU, Dguy R (2003) Effects of water and heat stress on the seedling height growth and biomass of five trees. Journal of Nanjing Forestry University 27:10–14Google Scholar
  24. Figueira AMES, Miller SD, De Sousa CAD, Menton MC, Maia AR, Rocha HRD, Goulden ML (2008) Effects of selective logging on tropical forest tree growth. J Geophys Res Biogeosci 113:243–250CrossRefGoogle Scholar
  25. Franco M, Silvertown J (2004) A comparative demography of plants based upon elasticities of vital rates. Ecology 85:531–538CrossRefGoogle Scholar
  26. Gong ZW, Gu L, Liu XS (2013) The distribution pattern of main coniferous resource for non-commercial forest in Beijing, P. R. China. International Journal of Digital Content Technology and its Applications 7:476–483Google Scholar
  27. Grogan J, Landis RM, Free CM, Schulze MD, Lentini M, Ashton MS (2014) Big-leaf mahogany Swietenia macrophylla population dynamics and implications for sustainable management. J Appl Ecol 51:664–674CrossRefGoogle Scholar
  28. Han WP, Ding GJ, Bao B (2012) Physiological and ecological responses of Pinus massoniana seedling from different provenances to drought stress. J Cent S Univ For Technol 32:25–29Google Scholar
  29. Hastings A (2001) Transient dynamics and persistence of ecological systems. Ecol Lett 4:215–220CrossRefGoogle Scholar
  30. He SJ, Xie JS, Yang ZJ, Yin YF, Li DC, Yang YS (2011) Status, causes and prevention of soil and water loss in Pinus massoniana woodland in hilly red soil region of southern China. Science of Soil and Water Conservation 9:65–70Google Scholar
  31. Hu BT, Wang XL, Cai HM, Hong SS, Feng JW, Yi JS (2000) Study on consecutive 8-year effect of fertilizer on a juvenile Masson pine plantation. For Res 13:286–289Google Scholar
  32. Hua P, Yan L, Zhi C, Simonson S, Luo G (2015) A preliminary study of desertification in red beds in the humid region of Southern China. Acta Geograph Sin 70:1699–1707Google Scholar
  33. Huang CB et al (2013) Effects of different sites and intermediate cutting intensities on growth of planted young Pinus massoniana forests. Journal of Northwest Forestry University 28:141–145Google Scholar
  34. Jin Y, Peng H, Yan L, Xiang Y, Wang X (2015) Discussion on desertification of humid region in southern China. Prog Geogr 34:772–780Google Scholar
  35. Jongejans E, Jorritsma-Wienk LD, Becker U, Dostal P, Milden M, de Kroon H (2010) Region versus site variation in the population dynamics of three short-lived perennials. J Ecol 98:279–289CrossRefGoogle Scholar
  36. Kolb A, Dahlgren JP, Ehrlén J (2010) Population size affects vital rates but not population growth rate of a perennial plant. Ecology 91:3210–3217CrossRefPubMedGoogle Scholar
  37. Li SL, Yu FH, Werger MJA, Dong M, Zuidema PA (2011) Habitat-specific demography across dune fixation stages in a semi-arid sandland: understanding the expansion, stabilization and decline of a dominant shrub. J Ecol 99:610–620Google Scholar
  38. Li SL, Yu FH, Werger MJA, Dong M, Ramula S, Zuidema PA (2013) Understanding the effects of a new grazing policy: the impact of seasonal grazing on shrub demography in the Inner Mongolian steppe. J Appl Ecol 50:1377–1386CrossRefGoogle Scholar
  39. Li M, Du M, Nie C (2014) Research advances in soil improvement of Pinus massoniana plantation. World Forestry Research 27:31–36Google Scholar
  40. Li SL, Yu FH, Werger MJA, Dong M, During HJ, Zuidema PA (2015) Mobile dune fixation by a fast-growing clonal plant: a full life-cycle analysis. Sci Rep 5:8395CrossRefGoogle Scholar
  41. Lin H (2005) Contrast analysis on the measures of Masson pine reclaim fertilization and mixed nitrogen fixation vegetation. Subtropical Soil and Water Conservation 17:59–60Google Scholar
  42. Liu S, Zeng W (2017) Large-scale individual tree growth models for Pinus massoniana in China. Forest Resources Management 2:28–33Google Scholar
  43. Ma Z, Zha X (2008) Research on ecological recovering of erosive degraded Pinus massoniana woodland in red soil region of Southern China. Research of Soil and Water Conservation 15:188–193Google Scholar
  44. Ma L, Lian J, Lin G, Cao H, Huang Z, Guan D (2016) Forest dynamics and its driving forces of sub-tropical forest in South China. Sci Rep 6:22561CrossRefPubMedPubMedCentralGoogle Scholar
  45. Ma ZR, Meng XJ, Wang L, Cheng BW (2017) Study on different intermediate cutting intensities experiment of Pinus massoniana plantations in Chongqing area. Journal of Fujian Forestry Science and Technology 44:33–36Google Scholar
  46. Mandle L, Ticktin T (2012) Interactions among fire, grazing, harvest and abiotic conditions shape palm demographic responses to disturbance. J Ecol 100:997–1008CrossRefGoogle Scholar
  47. Melo FPL, Tabarelli M (2003) Seed dispersal and demography of pioneer trees: the case of Hortia arborea. Plant Biol 5:359–365CrossRefGoogle Scholar
  48. Ning L, Bai B, Lu CH (2011) Recruitment limitation of plant population: from seed production to sapling establishment. Acta Ecol Sin 31:6624–6632CrossRefGoogle Scholar
  49. Ning QR, Li SZ, Jiang LC, Zhao Y, Liu R, Zhang XY (2016) Foliar nutrient content and resorption efficiency of Pinus massoniana in the subtropical red soil erosion region. Acta Ecol Sin 36:3510–3517Google Scholar
  50. Pan D, Zhang H, Pan G (2016) Community dynamics and stability of different aged Pinus massoniana plantations in southern Guangxi Chinese. J Ecol 35:1481–1490Google Scholar
  51. Peng S, Huang Z, Peng S, Ouyang X, Xu G (2004) Factors influencing mortality of seed and seedling in plant nature regeneration process. Guihaia 24:113–106Google Scholar
  52. Primack RB, Lee HS (1991) Population dynamics of pioneer (Macaranga) trees and understorey (Mallotus) trees (Euphorbiaceae) in primary and selectively logged bornean rain forests. J Trop Ecol 7:439–457CrossRefGoogle Scholar
  53. Qi L, Zhao FQ (2015) Impact of logging intensity on the spatial distribution and association of dominant tree species in a broadleaved Korean pine mixed forest. Acta Ecol Sin 35:46–55Google Scholar
  54. Ren H, Shen WJ, Lu HF, Wen XY, Jian SG (2007) Degraded ecosystems in China: status, causes, and restoration efforts. Landsc Ecol Eng 3:1-13CrossRefGoogle Scholar
  55. Roovers P, Verheyen K, Hermy M, Gulinck H (2004) Experimental trampling and vegetation recovery in some forest and heathland communities. Appl Veg Sci 7:111–118CrossRefGoogle Scholar
  56. Salguero-Gómez R, de Kroon H (2010) Matrix projection models meet variation in the real world. J Ecol 98:250–254CrossRefGoogle Scholar
  57. Sánchez-Velásquez LR, Pineda-López MDR (2009) Comparative demographic analysis in contrasting environments of Magnolia dealbata: an endangered species from Mexico. Popul Ecol 52:203–210CrossRefGoogle Scholar
  58. Schmidt IB, Mandle L, Ticktin T, Gaoue OG (2011) What do matrix population models reveal about the sustainability of non-timber forest product harvest? J Appl Ecol 48:815–826CrossRefGoogle Scholar
  59. Schweitzer CJ, Dey DC (2013) Logging intensity impact on small oak seedling survival and growth on the Cumberland Plateau in northeastern Alabama South. J Appl For 37:113–121CrossRefGoogle Scholar
  60. Smith RGB, Nichols JD (2005) Patterns of basal area increment, mortality and recruitment were related to logging intensity in subtropical rainforest in Australia over 35 years. For Ecol Manag 218:319–328CrossRefGoogle Scholar
  61. Sun B, Hu Z, Lü J, Zhou L, Xu C (2006) Nutrient leaching and acidification of Southern China coniferous forest red soil under stimulated N deposition. Chin J Appl Ecol 17:1820–1826Google Scholar
  62. Swamy V, Terborgh J, Dexter KG, Best BD, Alvarez P, Cornejo F (2011) Are all seeds equal? Spatially explicit comparisons of seed fall and sapling recruitment in a tropical forest. Ecol Lett 14:195–201CrossRefPubMedGoogle Scholar
  63. Tan JH, Tang SS, Chen H (2017) Advances of drought resistance in Pinus massoniana Guangxi. For Sci 46:1–7Google Scholar
  64. Vajpeyi D K, Ponomarenko A (2001) Deforestation in China. Deforestation environment, and sustainable development: a comparative analysis. Praeger Publishers, pp 91–110Google Scholar
  65. Vieira DLM, Scariot A (2008) Environmental variables and tree population structures in deciduous forests of central Brazil with different levels of logging. Braz Arch Biol Technol 51:419–431CrossRefGoogle Scholar
  66. Villela DM, Nascimento MT, Aragão LEOCD, DMD G (2006) Effect of selective logging on forest structure and nutrient cycling in a seasonally dry. Brazilian Atlantic forest Journal of Biogeography 33:506–516Google Scholar
  67. Waltert B, Wiemken V, Rusterholz HP, Boller T, Baur B (2002) Disturbance of forest by trampling: Effects on mycorrhizal roots of seedlings and mature trees of Fagus sylvatica. Plant Soil 243:143–154CrossRefGoogle Scholar
  68. Wang C, Han S, Huang M (2001) The ABA change and physiological responses of Pinus sylvestris L. var. mongolica Litv. to water shortage. Journal of Northeast Forestry University 29:40–43Google Scholar
  69. Win RN, Suzuki R, Takeda S (2012) Effects of selective logging on the regeneration of two commercial tree species in the Kabaung Reserved Forest, Bago Mountains, Myanmar. J Trop For Sci 24:312–321Google Scholar
  70. Winkler E, Marcante S, Erschbamer B (2015) Demography of the alpine pioneer species Saxifraga aizoides in different successional stages at the glacier foreland of the Rotmoosferner (Obergurgl, Otztal, Austria). Tuexenia 35:267–283Google Scholar
  71. Xie J, Yang Y, Chen G, Huang R, Zheng B (2002) Studies on the nutrient circulation and energy of the serious degraded community after closing of hillsides and management to facilitate afforestation. J Mt Res 20:325–330Google Scholar
  72. Xu J (2011) China’s new forests aren’t as green as they seem. Nature 477:371–371CrossRefPubMedGoogle Scholar
  73. Xu HQ, He H, Huang SL (2013) Analysis of fractional vegetation cover change and its impact on thermal environment in the Hetian basinal area of County Changting, Fujian Province, China. Acta Ecol Sin 33:2954–2963CrossRefGoogle Scholar
  74. Yamada T, Yu M, Okuda T, Kassim AR (2016) Long-term effects of selective logging on dipterocarp populations in the Pasoh Forest Reserve, Malaysia. Plant Ecology and Diversity 9:615–626CrossRefGoogle Scholar
  75. Yang M (2004) The effects of human-caused disturbances on the growth course of Pinus massiniana’ natural forest. Journal of Fujian Forestry Science and Technology 31:68–71Google Scholar
  76. Yang H, Jiang L, Li L, Li A, Wu M, Wan S (2012) Diversity-dependent stability under mowing and nutrient addition: evidence from a 7-year grassland experiment. Ecol Lett 15:619–626CrossRefPubMedGoogle Scholar
  77. Yang L, Ranjitkar S, Harrison RD, Xu J, Ou X, Ma X, He J (2017) Selection of native tree species for subtropical forest restoration in Southwest China. Plos One 12:e0170418CrossRefGoogle Scholar
  78. Yang X, Li S, Shen B, Wu Y, Sun S, Liu R, Zha R, Li S-L (2018) Plant demography data of Pinus massoniana measured from 2011-2013 in a degraded subtropical forest in Southeast China. V1. PANGAEA. [Dataset].
  79. Zackey J (2015) Peasant perspectives on deforestation in Southwest China: social discontent and environmental mismanagement. Mt Res Dev 27:153–161CrossRefGoogle Scholar
  80. Zeng J, Lei YC, Jia HY, Cai DX, Tang JX (2017) Dynamic growth response of Pinus massoniana plantation on intensive thinning in Southwestern Guangxi, China. For Res 30:335–341Google Scholar
  81. Zhang J, Lin G, Lin Z, Peng C (2005) Response of photosynthesis to growth light intensity in some south subtropical woody plants. Journal of Tropical and Subtropical Botany 13:511–518Google Scholar
  82. Zhang J, Li BH, Bai XJ, Yuan ZQ, Wang XG, Ye J, Hao ZQ (2009) Composition and interannual dynamics of tree seedlings in broad-leaved Korean pine (Pinus koraiensis) mixed forest in Changbai Mountain. Biodiversity Science 17:385–396CrossRefGoogle Scholar
  83. Zhang Y, Zhou Z, Yang Q (2013) Nitrogen (N) deposition impacts seedling growth of Pinus massoniana via N:P ratio rffects and the modulation of adaptive responses to low P (phosphorus). Plos One 8:e79229CrossRefPubMedPubMedCentralGoogle Scholar
  84. Zhang C, Hanqiu X, Zhang H, Tang F, Lin Z (2015) Fractional vegetation cover change and its ecological effect assessment in a typical reddish soil region of Southeastern China: Changting County, Fujian Province. J Nat Resour 30:917–928Google Scholar
  85. Zhu HJ (2015a) Strategies on eco-restoration in the subtropical mountain ecosystem fragility areas, China: based on the achievement of eleven years’ research in Changting County. J Nat Resour 28:1498–1506Google Scholar
  86. Zhu HJ (2015b) Ecological restoration strategies in fragile subtropical mountain ecosystems in China based on long-term research in Changting County. Journal of Resources and Ecology 6:139–145CrossRefGoogle Scholar
  87. Zuidema PA, De KH, Werger MJ (2007) Testing sustainability by prospective and retrospective demographic analyses: evaluation for palm leaf harvest. Ecol Appl 17:118–128CrossRefPubMedGoogle Scholar
  88. Zuidema PA, Jongejans E, Chien PD, During HJ, Schieving F (2010) Integral projection models for trees: a new parameterization method and a validation of model output. J Ecol 98:345–355CrossRefGoogle Scholar

Copyright information

© INRA and Springer-Verlag France SAS, part of Springer Nature 2018

Authors and Affiliations

  • Xianyu Yang
    • 1
  • Shouzhong Li
    • 1
    Email author
  • Baocheng Shen
    • 2
  • Yuyan Wu
    • 1
  • Suitao Sun
    • 1
  • Rong Liu
    • 3
  • Ruibo Zha
    • 4
  • Shou-Li Li
    • 5
  1. 1.State Key Laboratory of Subtropical Mountain Ecology (Funded by Ministry of Science and Technology and Fujian Province)Fujian Normal UniversityFuzhouChina
  2. 2.Section of Ecology, Department of BiologyUniversity of TurkuTurkuFinland
  3. 3.State Key Laboratory of Vegetation and Environmental Change, Institute of BotanyChinese Academy of SciencesBeijingChina
  4. 4.College of ToursimFujian Normal UniversityFuzhouChina
  5. 5.Department of Biology and Center for Infectious Disease DynamicsThe Pennsylvania State UniversityUniversity ParkUSA

Personalised recommendations