Introduction

Positive interactions, including mutualism, commensalism and facilitation, are traditionally defined as a type of interaction in which one species benefits from the other, without harm and potentially with benefits. These interactions can strongly shape the structure and function of plant communities [2], affecting population growth and fecundity, abundance, geographical range, soil nutrient supply, community diversity and stability [34–5]. For the positive effects of multi-species interactions [6], mixed afforestation was commonly used for forest conservation and restoration in degraded regions over the past two decades [7–8].

Coastal zones, as transition areas between terrestrial and marine ecosystems, are locations of intense exchange of matter and energy. Both waves and winds along the coast erode the ground and impose environmental constraints on forest succession, as well as threaten the restoration of original communities that have been destroyed. Historically, various species grew in the tropical coastal zone of southern China. However, because of the excessive exploitation of natural resources in the early 1900s, these coastal forests were destroyed, leaving a bald coastal landscape for nearly half a century [9]. This situation was not changed until the introduction of the exotic tree species Casuarina (C. equisetifolia).

C. equisetifolia is a she-oak species of the genus Casuarina, with a native range extending from Burma and Vietnam throughout Malaysia; east to French Polynesia, New Caledonia, and Vanuatu; and south to Australia [10]. It was introduced to Hainan Island to reduce coastal erosion in the 1950s and immediately became the dominant species due to its pioneer characteristics, including fast growth, adaptability to barren soils, and ability to resist wind [11]. Unlike normal broadleaf species, C. equisetifolia has morphological characteristics, such as a scaly leaf, lance-shaped branchlets, and a tower-shaped morphological structure, that increase its wind resistance and allow it to grow better in hostile coastal environments [12]. However, due to its high tannin content, the litterfall of C. equisetifolia decomposes slowly, decreasing the rate of nutrient return and influencing the growth of the forest. The litter beneath C. equisetifolia consists of fallen cladodes and can accumulate to form thick continuous deposits, which smother and prevent establishment of other species [13–14].

Because of its superior adaptability, Acacia (Acacia mangium) became a main plantation species that creates a green and attractive landscape and also contributes to soil improvement and environmental protection of the tropical coastal plain [15]. A large number of afforestation experiments showed that seedlings of A. mangium are vulnerable to an oceanic climate, and seedlings planted in a coastal open area will die within approximately two years because of unfavorable conditions if they are not mixed with other nurse plants. Therefore, A. mangium plantations were always established behind zones of C. equisetifolia to avoid destruction by erosion and winds. Considering the positive effects of mixed afforestation discussed above, we examined the interaction effects between the two tree species during 10 years of succession in the tropical coast. The objectives of this study were to determine: (1) how the vegetation dynamics and soil nutrients in monoculture stands of C. equisetifolia were influenced by A. mangium; and (2) whether C. equisetifolia protects A. mangium at the young stage.

Methods

Study area

The study was conducted in the northeast coastal zone of Wenchang city, Hainan Island, adjacent to the South China Sea, in the monsoon tropics of south China (Fig. 1). The tropical marine climate ensures a plentiful annual rainfall of 1,721.6 mm and mild temperatures averaging 22°C to 24°C. However, the area is also prone to severe typhoons and rain storms. Five to six typhoons occur each year, and some are fatal to replanted forests.

Fig. 1.

Study sites in the northeast coastal zone near Wenchang city, Hainan Province, China. X.Y. and Z indicate three experiment blocks (2 hm² each). Three planting treatment plots (a= Casuarina equisetifolia only, b= Acacia mangium I and c= Acacia mangium II) were placed in each block.

Experimental design

We examined the effects of planting A. mangium along with C. equisetifolia for tropical coastal forest succession and habitat restoration in an area of 2 hm². Both species' seedlings were selected in the nursery and planted systematically using the plant-to-row method on barren land when the plants were one year old in year 1998. Because A. mangium monocultures cannot survive in hostile coastal environments, we used three planting levels as follows: C. equisetifolia only, A. mangium I and A. mangium II (I indicates that the mixing proportion of C. equisetifolia to A. mangium is 1:1, II indicates that the mixing proportion of C. equisetifolia to A. mangium is 2:1). The initial density of afforestation was 2,500 stems per hectare (afforestation spacing=2 m × 2 m). The number of trees in each plot varied with factors such as typhoon, self-thinning, competition, and herbivory. The experimental treatments were determined using a randomized block design with three blocks. Within each block, three plots (plot area: 20 m × 20 m) were established for three planting levels independently, and in each plot, one subplot of 5 m × 5 m was established.

In each experimental plot, the number of trees per hectare was recorded biennially (2000 to 2010) to document survival in the various experimental plantings.

Biennially (2000 to 2010), the DBH (diameter at breast height) and height of trees were measured in each experimental plot and used to calculate the aboveground biomass using two biomass equations from the literature [16–17].

(1)

(2)

where Y is the individual aboveground biomass and H is the tree height.

The soil samples were collected for nutrient analysis from the different planting levels to evaluate the role of different afforestation types in the soil condition. The baseline soil data were collected in 2000 together with other surveys. Three punch-tube samples of the upper 15 cm of sandy soil were collected and mixed together in each plot. These bulk samples were then analyzed for total N using the micro-Kjeldahl method, and the total organic matter was measured using a modified Walkley-Black procedure [18]. In addition, we began the second survey in 2010.

In each subplot, we assessed the species and number of natural regenerations and shrubs, and also surveyed the grass coverage. All measurements were recorded in late November, biennially, from 2000 to 2010.

Seedlings are vulnerable to oceanic climate. To evaluate the nurse role of C. equisetifolia on A. mangium, we selected fifteen C. equisetifolia and A. mangium plants randomly in monocultures of C. equisetifolia, A. mangium I and A. mangium II in the periphery of these stands in year 2000 (three years old). The LAI was obtained from canopy analysis systems (Hemiview 2.1) by analyzing canopy photos of individual trees in the vertical direction. Crown lengths (W-E-N-S four directions) were measured to demonstrate the lopsided crown phenomenon under continuous environmental stress.

Results

Vegetation dynamics

The number of planted species in each experimental plot decreased with time because of constant interference from the tropical marine climate. Plots planted with C. equisetifolia only had a significantly greater number of plant species than other treatments in 2010 (Fig. 2A; Appendix 1). For the mixed treatments, the average plant number in the plots of A. mangium II (Number: C. equisetifolia: 1,424 /hm²; A. mangium 422/hm²) was greater than A. mangium I (Number: C. equisetifolia: 998 /hm²; A. mangium 742/hm²) (Table 1).

Fig. 2.

Plant species/hm² (A), aboveground biomass (B), recruitment of native tree species/hm² (C), and diversity of regeneration (D) and shrubs (E), grass coverage (%) (F), (mean + SE, n=3) in plots with 3 planting treatments for the years 2000–2010. (Within each graph, different letters to the right of each line indicate significant differences (P < 0.05) based on a Tukey's hsd test following the overall effect analysis of RANOVA. A. mangium is the abbreviation for Acacia mangium and C. equisetifolia is the abbreviation for Casuarina equisetifolia.

Table 1.

The effects of three planting treatments on community dynamics over a 10-yr period.

The aboveground biomass was, on average, highest in plots of A. mangium I (71.48 t/hm²). Next highest was the other mixing type of A. mangium II (67.53 t/hm²). The aboveground biomass in plots planted with C. equisetifolia was significantly lower (64.64 t/hm²) than the other two types (Fig. 2B).

The abundance and diversity of natural regenerating saplings differed significantly among the three treatments (Table 1). The average abundance (2,223 /hm²) and diversity (H′: 0.90) in plots of A. mangium I were significantly higher than the other two treatments. The next highest was A. mangium II, with a number and diversity (H′) of 1,833 /hm² and 0.74, respectively. The abundance (1,028 /hm²) and diversity (0.28) in pure C. equisetifolia plots were the lowest (Fig. 2 C and Fig. 2D).

The plots originally planted with A. mangium I had the highest shrub diversity (H′: 1.16) of the three treatments, followed by A. mangium II (Fig. 2E). The shrub diversity in plots planted only with C. equisetifolia was the lowest (0.63).

The grass cover was not significantly different between the two types of mixed plots (Fig. 2F). The pure plots had the lowest grass cover (18%), and much of the floor was covered by C. equisetifolia litter (Table 1).

Soil conditions

The 10 years of site occupancy by tree plantations significantly altered many soil properties, but the various treatments affected the soils differently.

The soil organic matter (SOM) content, on average, was significantly different among the three treatments (F_{2, 6}=85.23; P <0.01). Plots planted with A. mangium I had the highest SOM (0.73%) in all treatments (Table 2). This was followed by A. mangium II (0.65%) and the monoculture of C. equisetifolia (0.38%).

Table 2.

Soil organic matter and total N in the upper 15 cm of soil after 10 yrs of site occupancy by different plant species.

The variation in SOM was not the same at different treatments during the 10-yr period. The SOM increased significantly in the plots of A. mangium I (0.40 %) and II (0.27 %), but decreased significantly in the monoculture of C. equisetifolia (0.09%) (Table 2; Appendix 2-first).

The average totals of N in the plots planted with A. mangium (I [0.40 g/kg] and II [0.37 g/kg]) were significantly higher than the monoculture of C. equisetifolia. The variation of the total N in the plots planted with only C. equisetifolia (0.09 g/kg) was the lowest, which was significant lower than A. mangium I and II (Table 2 and Appendix 2-second).

Canopy condition

The LAI values were significantly different among the three treatments (LAI: P<0.01, F=15.832). The LAI value of C. equisetifolia in A. mangium II was higher than the other four, whereas the A. mangium in A. mangium I was the smallest (Table 3).

Table 3.

Leaf area index and crown length in different types of 3-year-old forests.

Generally, the crowns grew evenly along the four directions. Crowns formed the shape of an ellipse or were nearly tapered unless influenced by a continuous external force. The results showed that the crown lengths of A. mangium individuals in the four directions were highly significantly different (P=5.9 ×10⁻⁷, F=14.084) in A. mangium I (the ratio of A. mangium was high). The crown length in south and west directions was over three meters, which was significant longer than in the east and north directions (less than two m). However, the A. mangium crown length exhibited few differences in A. mangium II, and only the west crown showed a slight difference.

Discussion

Although the roles of interactions are recognized throughout the field of ecology, the effect of positive interactions between nonindigenous species is a controversial topic in restoration and conservation [19]. Our comparative trial revealed that nonindigenous A. mangium could improve the aboveground biomass, understory richness and diversity, and soil nutrients when combined with C. equisetifolia. C. equisetifolia could protect A. mangium effectively under variable maritime climate conditions.

Species respond differently to environmental stresses, and each species has a survival strategy for a hostile environment. Unlike C. equisetifolia, A. mangium is more likely to live in a protected environment. Compared to C. equisetifolia, A. mangium seedlings are vulnerable to death in the open, but their recruitment improved under C. equisetifolia, which provides shelter against a harsh maritime environment. Our study shows that the number of individuals in each treatment group depends significantly on the initial density of C. equisetifolia. A. mangium II benefitted from the protective effect from C. equisetifolia, and had higher survivorship than A. mangium I during the 10-yr period.

Beyond any maximum-yield density, interactions occur as component species strive to capture resources from a shared location [20]. Compared to a monoculture of C. equisetifolia, A. mangium has more capacity to accumulate biomass because it grows quickly. The biomass accumulation in the mixed forest was clearly improved, particularly when the proportion of A. mangium was high.

Natural regeneration plays a key role in vegetation restoration and development. In our experiment, A. mangium planted at different ratios to C. equisetifolia influenced the long-term understory characteristics, such as diversity and grass coverage. All of the understory layers (regeneration, shrubs, and grass) were improved in the mixed treatments (I and II). For the C. equisetifolia mixed forest, understory vegetation improved more under level I than II. However, we provide no direct evidence to explain the role of A. mangium on recruitment, regeneration and shrub development. Nevertheless, the long-term development of different species assemblages can profoundly affect understory vegetation dynamics and produce results similar to those of Bezemer et al. [21]. Many previous studies were performed on short time scales and used artificial management, such as weeding [22]. Therefore, data on the influence of mixed afforestation on understory vegetation restoration in the absence of human interference over 10 years are important for coastal vegetation management.

Soil organic matter is the main source of nutrients for plant growth and soil development. It can improve the soil physical properties and accelerate the activities of microorganisms. Because C. equisetifolia litter decomposes very slowly, the organic matter content is low and easily lost in coastal sandy soils [13]. In our experiment, the SOM increased significantly in plots planted with A. mangium I and II over 10 years. The A. mangium used in our study are nitrogen-fixing plants and effectively increase the soil total N when interplanted with the other nitrogen-fixing plant, C. equisetifolia.

Unlike most broadleaf species, C. equisetifolia has special canopy characteristics, such as degraded scaly leaves, lance-shaped branchlets, and tower-shaped morphological structures. These phenotypic traits increase wind resistance and allow for better growth in the hostile coastal environment, particularly during the young growth stages [11]. Because of the continual influence of the northeastern marine monsoon, the morphological characteristics of tree species generate differentiation, predominantly during the young stages. Lopsided crowns are a universal phenomenon for the majority of species and significantly affect growth. However, this phenomenon was not observed for C. equisetifolia. Its adaptive features allow C. equisetifolia to resist strong winds and protect other species from hurricane influences, decrease the crown damage from hurricanes, and maintain normal photosynthetic capacity. Evenly distributed crown growth, intact canopy conditions, and high LAI provide advantages for interacting with other species.

Implications for conservation

The complexity of the community structure, species richness and diversity, and soil fertility are the main factors in forest vegetation dynamics [23]. Plant facilitation plays an important role in these factors and may promote early succession in abandoned fields [8, 24]. Our 10-year experiment on the species configuration shows how different species change a community dominated by C. equisetifolia. As a fast-growing pioneer species, A. mangium showed better effects on the development of both stands (aboveground biomass and species diversity) and soil conditions (organic matter and total N) when combined with C. equisetifolia. Our results indicate that mixed afforestation could solve several problems caused by species such as C. equisetifolia, and may improve the structure and function of coastal forest ecosystems and promote the process of forest succession.