Use of exotic plants to control Spartina alterniflora invasion and promote mangrove restoration

In coastal China, the exotic invasive Spartina alterniflora is preventing the establishment of native mangroves. The use of exotic species, control of exotic plant invasion, and restoration of native plant communities are timely research issues. We used exotic Sonneratia apetala Buch.-Ham and S. caseolaris (L.) Engl. to control invasive Spartina alterniflora Loisel through replacement control for five years, which concurrently promoted the restoration of native mangroves. This process includes three stages. I: In a mangrove area invaded by S. alterniflora, exotic S. apetala and S. caseolaris grew rapidly due to their relatively fast-growing character and an allelopathic effect. II: Fast-growing S. apetala and S. caseolaris eradicate S. alterniflora through shading and allelopathy. III: The growth of native mangrove was promoted because exotic plant seedlings cannot regenerate in the understory shade, whereas native mesophytic mangrove plants seedlings can grow; when the area experiences extreme low temperatures in winter or at other times, S. apetala dies, and native mangrove species grow to restore the communities. This model has important implications for addressing the worldwide problems of “how to implement the ecological control of invasion using exotic species” and “how to concurrently promote native community restoration during the control of exotic invasion”.

The use of exotic species, control of exotic plant invasion, and restoration of native plant communities are timely research issues. However, they are associated with different research areas and such studies are very difficult to implement successfully. The problem of how to conduct a comprehensive study on these three topics has been undertaken for mangrove restoration in the coastal zone of South China, where the use of exotic Sonneratia apetala Buch.-Ham and S. caseolaris (L.) Engl. to control invasive Spartina alterniflora Loisel and concurrently promote the restoration of native mangroves has been investigated.
The ecological control of invasive species and the restoration of native communities have gained the attention of many researchers [1][2][3] . The interactions between invaders and competitors may occur directly through competitions for resources or indirectly by allelopathic inhibition 4 . Although certain non-invasive species exhibit stronger competitiveness than invasive species during a particular stage of the life cycle 5 , there are extremely limited successful cases in which noninvasive plants have been used to control invasion 6,7 . Thus, sufficiently strong noninvasive competitors and even less invasive exotics may be useful for biological control. A common method is to use a strong competitor for competitive control, thereby contributing to ecological control 6,8 . Replacement control is a control method based on the rules of interspecies competition and plant community succession. It uses a more valuable species to naturally replace a harmful plant species 9 and implement the goal of restoration at a higher level. In general, replacement control adopts native species or plants that have been proven non-harmful to native species after long-term growth. As a competitive plant, the selected species competes with exotic invasive plants and inhibits their growth, usually causing no danger to other useful native plants and also contributing to biodiversity. A few studies have assessed the possibility for the replacement control of invasive plants with native plants. In the Azores community, a native flame tree species has been used to control invasive Pittosporum undulatum in a replacement area up to 24% of the invaded area 10,11 . It is generally thought that the presence of an exotic plant species is likely to reduce the invasiveness of other invasive plants. In addition, Blank et al. (2015) 12 studied the exotic perennial grass Agropyron cristatum, which has been used extensively to control the exotic annual grass Bromus tectorum in the intermountain west of the United States, suggesting the co-opting of biological soil space by the perennial grass as another suppressive mechanism. However, the risk of invasion by exotic non-invasive species is unpredictable 13 , and few studies have successfully attempted to use exotic species for competitive replacement control in the field.
Due to their economic attributes, coastal zones are associated with overexploitation, resulting in the shrinkage of contiguous mangrove areas. The degradation and loss of mangrove habitats are most evident in Asia and the Atlantic region. In some places, nearly 70% of the original mangrove habitat has been lost 14,15 . The existing natural mangrove area in China is currently only 15,000 hm 2 , although it has historically reached 250,000 hm 2 16 . Mangrove restoration has become one of the most pressing issues in land improvement in China.
Lost mangrove areas are largely invaded by exotic S. alterniflora. Thus, exerting control over S. alterniflora is a prerequisite for mangrove restoration. S. alterniflora originates from the Atlantic coast. This species presents good performance in tidal-flat preservation, bank protection, wind breaking, and dyke strengthening. S. alterniflora was introduced into China in 1979. At present, S. alterniflora is the exotic species associated with the most serious invasion in China's coastal salt marshes 17,18 . To date, however, neither physical, chemical, nor biological control through natural-enemy introduction has been able to effectively control S. alterniflora [19][20][21][22] . Attempts have been made to "control grass with grass", i.e., to increase the competitiveness of the native reed (Phragmites communis Trin.) by changing the level of resources in the environment and thereby replacing S. alterniflora 18 . A previous study 23 used a native mangrove species (Kandelia obovata) for the replacement control of invasive S. alterniflora, which also played a certain role in preventing the spread of S. alterniflora and restoring the typical complex food web of mature mangrove ecosystems. However, none of the developed methods have been found to be effective for the control of S. alterniflora. Therefore, the use of replacement control, particularly the search for effective replacement plants, appears to be of great significance.
Exotic S. apetala and S. caseolaris have been shown to strongly inhibit S. alterniflora 17,22 . The present study attempted to use these two exotic mangrove species to control exotic invasive S. alterniflora and concurrently promote native mangrove restoration. S. apetala was introduced into China from Bangladesh. As an exotic species, S. apetala exhibits the characteristics of early successional species, such as heliophytic nutrition, fast growth and reproduction, and a short life cycle 17 . S. apetala is commonly used for the ecological restoration of mangroves in South China 18,23,24 . S. caseolaris, originating from Hainan Province, China, represents another exotic species in the study area. Under suitable conditions, their prominent fast-growing characteristics enable S. caseolaris and S. apetala to rapidly generate a closed canopy and results in forest establishment on bare tidal flats. The establishment of forest will increase soil fertility and improve the habitat, creating favorable environmental conditions for the settlement and growth of other native mangrove plants 25,26 . The mixed-species planting of S. caseolaris and S. apetala can effectively restore mangroves 27 . Additionally, a previous study has shown that S. apetala exerts a stronger allelopathic effect than S. alterniflora 28 . Because of its relatively low regeneration rate, S. apetala presents moderate invasiveness 23,24 . Based on these characteristics, S. apetala may be useful for the control of invasion by the exotic invasive S. alterniflora and for the promotion of the restoration of native mangrove communities.
The present study was designed to address the following research questions: 1) Can exotic species be used to control exotic invasive plants without causing a secondary invasion? 2) Can this ecological control method facilitate the restoration of native plant communities? The replacement model was further proposed, and the underlying mechanism was investigated to provide a paradigm for the comprehensive study of and the development of management practices regarding the use of exotic species to control exotic invasive species and concurrently promote native community restoration.

Materials and Methods
Study area description. The study area is located on Qi'ao Island, Zhuhai, Guangdong Province, . This area is situated in the Hengmen Estuary in the northeast of Zhuhai and covers an area of 2,400 ha. The area has a vegetation coverage of 85% under a southern subtropical monsoon climate. The mean annual precipitation is 1,964.4 mm, and the mean annual temperature is 22.4 °C. The tide is an irregular semidiurnal tide with a mean high tide level of 0.17 m and a mean low tide level of − 0.14 m. The water quality is relatively clean, and the mean annual seawater salinity is 18.2‰. The soil type is coastal saline meadow marsh soil, and the total salt content in the topsoil (0-13 cm) is 20.82‰ 29 . This area has been largely invaded and surrounded by S. alterniflora 30 .
The plot chosen was an approximately ~14.7-ha tidal flat with S. alterniflora in the west of Qi'ao Island. In early 2008, S. apetala single-species and S. apetala + S. caseolaris mixed-species plots were planted in the S. alterniflora tidal flat (planting density of ~35 trees/100 m 2 , Table 1). Additionally, a pure plot of S. alterniflora and a 6-8-year-old mature S. apetala plot without S. alterniflora growth were selected as controls (Fig. 1). Light intensity measurement. The understory light intensity was measured along two diagonals of the plot with an electronic illuminometer (BK-1332 A). Measurements were made at 10 points in each plot.
Biomass estimation and measurement. The biomass of S. caseolaris and S. apetala was calculated using an empirical formula for fast-growing mangrove plants 31 .
where D is the tree DBH (at 1.3-m height), and H is the tree height; S. alterniflora biomass measurement: The aboveground biomass of S. alterniflora was measured using a harvesting method. Plants were harvested from an area of 1 m × 1 m in each plot. All of the aboveground parts were cut and collected. The fresh plant material was weighed and then dried in an oven at 70 °C to obtain a constant weight.

RGR
Q dQ dt Biomass RGR: where Q is the quantity of the original material, and dQ/dt is the transient increment (unit: mg·g −1 ·d −1 ).
Soil survey in the sample plots. Surface soil (0-15 cm) samples were collected after the plot survey.
In each plot, soil samples were collected at nine points in a Z-shaped pattern, and the samples from every three points were pooled to obtain a composite sample by dichotomy. Three composite soil samples were obtained from each plot. The samples were coded and kept in sealable bags. The soil samples were naturally air-dried, and the plant residues were manually removed. After grinding, the samples were passed through 10-and 100-mesh sieves. The 10-mesh sieved samples were used for pH measurements, and the 100-mesh-sieved samples were used for assays of the soil physicochemical parameters.
The soil pH was measured by water extraction (soil: water = 1: 2.5, w: v) with a pH meter. The soil total carbon and nitrogen contents were determined with a Vario EL Cube elemental analyzer (Elementar, Germany). Data analysis. The statistical analyses were performed using SPSS 19.0 (SPSS Inc., Chicago, IL, USA).
One-way ANOVA was used to analyze the differences between samples at different times or subjected to different treatments. The least significant difference test was performed to test the significance of the differences. A linear regression analysis was performed to demonstrate the role of declining light intensity in the reduction of Spartina alterniflora. The significance level of the differences was set to P < 0.05. Plots were generated using Microsoft Excel 2010 and SPSS 19.

Results
Influence of different replacement control modes on S. alterniflora density. After one year of replacement control, the S. alterniflora density was 266 plants·m −2 in pure plot I compared with 82 and 64 plants·m −2 in plots II and III, respectively. The two replacement control measures significantly reduced the density of S. alterniflora. Similarly, after two years of replacement control, both the S. apetala planting and the S. apetala + S. caseolaris mixed-species planting showed a significant controlling effect on S. alterniflora (p < 0. 05). Compared with that in plot I, the S. alterniflora density was reduced by 66.82% in plot II and by 76.36% in plot III. After five years of replacement control, the S. alterniflora density was 72 plants·m −2 in plot I compared with 27 plants·m −2 in plot II and 21 plants·m −2 in plot III. The S. alterniflora density in plot III was reduced by 70.83% compared with that in the plot that was not subjected to replacement control. After five years of control with the single species S. apetala, the S. alterniflora density decreased by 62.5%. Compared with the pure plot of S. alterniflora, plots II and III showed control differences with significant effects (p < 0.05; Fig. 2).
Influence of different replacement control models on S. alterniflora biomass. After two years of replacement control, the aboveground biomass (dry matter) of S. alterniflora associated with a mixed-species planting of S. apetala + S. caseolaris was 0.11 ± 0.04 kg/m 2 , which is significantly less than that associated with a single-species planting of S. apetala (0.24 ± 0.07 kg/m 2 ). The aboveground biomass of S. alterniflora in the pure plot was 1.07 ± 0.19 kg/m 2 . Clearly, the two treatments both resulted in significant reductions in the S. alterniflora biomass compared with the control plot. After five years of replacement control, the S. alterniflora biomass showed no significant difference between plots II and III. However, the aboveground biomass of S. alterniflora had decreased to 0.12 ± 0.01 and 0.08 ± 0.03 kg/m 2 in the two treatment plots, which are significantly lower than that in the control plot of pure S. alterniflora that was not subjected to any replacement control (Fig. 3).

Differences in the RGR of various plants under different replacement control models.
After one, two, and five years of replacement control, S. alterniflora presented significantly lower RGR values in plots II and III than in plot I (Fig. 4a), which indicated that the growth of S. alterniflora was inhibited under the control of S. apetala and S. caseolaris. In plot II, S. apetala had a significantly higher RGR than S. alterniflora (Fig. 4b), and in plots III, both S. apetala and S. caseolaris had significantly higher RGR  values than S. alterniflora (Fig. 4c). This finding showed that both S. apetala and S. caseolaris exhibit stronger competitiveness than invasive S. alterniflora under coexisting conditions.

Differences in the understory light intensity and soil properties under different replacement control models.
Influence of different replacement control models on the understory light intensity. The understory light intensity was reduced after five years of replacement control (Fig. 5). In the control plot of pure S. alterniflora, the understory light intensity reached 636.00 Lux. In plots II and III, the understory light intensities were 457.60 and 420.27 Lux, respectively, i.e., 71.94% and 66.08% of that in the S. alterniflora tidal flat without afforestation. The reductions in understory light intensity were 28.05% in plot II and 33.8% in plot III, showing significant differences (p < 0.05). Among the different plots, the lowest light intensity of 161.2 Lux was found in plot I; this minimal value amounts to 25.34% of that in plot I, showing a significant difference (p < 0.05). The regression analysis indicated the declining light intensity is a major factor in the reduction of S. alterniflora biomass (Fig. 6).
Influence of different replacement control models on soil physicochemical properties. A comparative analysis of soil physicochemical properties showed that the total carbon content, the total carbon nitrogen and the carbon/nitrogen ratio of the soil increased over time in all three types of plots (Fig. 7). In 2013, the soil total carbon, total nitrogen, and carbon/nitrogen ratio all significantly increased in plots II and III compared with plot I, which indicated that the planting of S. apetala and S. caseolaris markedly enhanced  the nutritional status of the soil in some respects in S. alterniflora-invaded areas, resulting in significant improvements in soil properties after five years of replacement control.
Restoration of native mangrove plants. Four native mangrove species, including Acanthus ilicifolius, Kandelia candel, Aegiceras corniculatum, and Derris trifoliata, were found in the plots (Fig. 8). In (a) One-way ANOVA was used to test for differences in the RGR of Spartina alterniflora among plots I, II and III; (b) One-way ANOVA was used to test for differences in the RGR between two species in plot II; (c) One-way ANOVA was used to test for differences in the RGR among three species in plot III. (mg·g −1 ·d −1 , mean ± S.E., p < 0.05): (I) S. alterniflora; (II) S. apetala controls S. alterniflora; and (III) S. apetala + S. caseolaris control S. alterniflora. particular, the types and numbers of native mangrove species in plot III were higher than those in plots I and II, indicating that the mixed-species planting of S. caseolaris and S. apetala can effectively restore native mangrove. In addition, native mangrove plants were found in plot IV, showing that S. apetala will not occupy the habitat forever.

Replacement control model and mechanisms for the control of invasive S. alterniflora by exotic S. apetala and native mangrove restoration. Our study successfully used S. apetala and
S. caseolaris to control S. alterniflora (Figs 2 and 3) and concurrently promote the restoration of native mangrove plants (Fig. 8). The entire replacement and restoration model is summarized in Fig. 9. Changes in the dominant species during different stages reflect the model of exotic species control and native community restoration. At the first stage (invasion process), the native plant dominance declines in parallel with exotic species invasion due to interference resulting from human activities. At the second stage (replacement control), the planted transitional exotic species grows fast and thus reduces the dominance of the exotic invasive species. At the following stage (native community restoration succession), the transitional exotic species cannot regenerate and gradually degenerates; however, the shaded habitat built by the transitional exotic species (Figs 4 and 5) provides favorable conditions for the restoration of native plants, thus improving soil properties (Fig. 7) and restoring native communities. The colonization of plants occurs in multiple stages. Because various factors work in different stages, each stage should be studied 32 . In the model of exotic species replacement and native community restoration (Fig. 9), different mechanisms are involved in the various stages (Fig. 10).
Stage I (Fig. 10a,b): Due to the destruction of native mangrove communities, S. alterniflora grows rapidly after invasion. The fast-growing characteristic of S. alterniflora enables it to rapidly close the canopy and establish a forest to prevent the growth of native mangrove plants. This study planted exotic S. apetala    in an S. alterniflora-invaded area for ecological control. As a pioneer species, S. apetala showed a higher growth rate than the invasive species S. alterniflora (Fig. 4). Species with higher resource competitiveness and growth rate can overcome their competitors 33,34 . For example, a number of invasive species compete with native species due to their high growth rate [34][35][36] . The ability to grow in S. alterniflora-invaded areas is the mechanism of S. apetala colonization. Moreover, during the colonization stage of pioneer species, S. apetala gains an advantage in its competition with S. alterniflora through allelopathy (Fig. 10b). Previous research has shown that S. apetala secretes more volatile allelopathic substances than S. alterniflora, and the former can also increase the leaf malondialdehyde content and inhibit growth in the latter, thereby achieving the goal of the replacement control of S. alterniflora 37 . Species exerting allelopathic effects can obtain more competitive advantages 38 . This mechanism may promote the colonization of S. apetala seedlings in high-density communities of S. alterniflora. However, a high allelopathic effect suppresses the regeneration of native mangrove plants 39 . This explains why native mangrove plants could not grow in S. alterniflora invaded areas.
Stage II (Fig. 10c,d): Once established, on the one hand, an S. apetala plot exerts an allelopathic effect to prevent S. alterniflora growth, and on the other hand, when the biomass or height of S. apetala exceeds that of an invasive species, the closed canopy reduces the understory light intensity (Fig. 5), preventing S. alterniflora growth (Fig. 10d). For example, it has been shown that plant communities compete for light resources through asymmetrical competition 40 . Species that win this competition are often higher and can create a shaded habitat for competitors within the community 17,41 . Thus, asymmetrical competition between trees (S. apetala) and grass (S. alterniflora) is conducive to tree growth. Ultimately, the exotic mangrove S. apetala replaces the exotic invasive grass S. alterniflora.
Can exotic S. apetala cause a secondary invasion? Once S. apetala effectively inhibits S. alterniflora, will S. apetala undergo large-scale expansion and thus lead to a secondary invasion? The use of exotic species in ecological control and native community restoration may lead to a new invasion-this approach appears to be a paradox. In this context, the regeneration features of moderately invasive species are of particular importance to the restoration of native communities. The present study showed that S. apetala substantially and almost completely replaced S. alterniflora after five years of control treatment and that native mangrove plants began to grow in the S. apetala understory thereafter (Fig. 8). In fact, the growth promotion of late-stage species by this type of plot is fairly common during forest community succession. For example, the tree canopy indirectly promotes the growth of Quercus suber seedlings by influencing the herbaceous layer 42 . Thus, "nurse trees" are commonly used in ecological restoration 25 .
In the present study, S. apetala seedlings were unable to regenerate because of their inherent growth characteristics. S. caseolaris and S. apetala are heliophytic fast-growing tree species that can rapidly close the canopy and establish a forest. The resultant low-light conditions inhibit the seed germination and seedling growth of S. caseolaris 43 , resulting in a higher mortality of S. apetala seedlings. Research shows that with low light resources, S. apetala no longer displays its fast-growing characteristic 17 . Consequently, S. caseolaris and S. apetala, as successional pioneer species, eventually die out naturally. This is another factor limiting the natural regeneration of S. caseolaris and S. apetala. Additionally, extreme low temperature is the main cause of death for 1-year-old S. apetala seedlings. For example, the extreme low temperature over the years on Qi'ao Island is approximately 2.5 °C. On the island, S. apetala features a high fruit yield but few perennial seedlings. This is because the seeds have a significantly low natural germination rate in intertidal flats, and seedlings are largely removed due to high tides in winter and autumn 30 . Moreover, the shaded habitat created by S. apetala and S. caseolaris is conducive to the growth of native mangrove plants (Fig. 10e). Unlike the two exotic species, the native mangrove A. corniculatum exhibits strong tolerance to low-light conditions, and its seedlings can grow without being affected by light intensity changes. In shaded environments, S. apetala no longer competitively inhibits A. corniculatum 17 , allowing seedlings of the native mangrove plants to begin to grow. The low preservation rate of S. apetala seedlings and their weak resistance to low temperature indicate that this exotic species will ultimately lose its dominance and be replaced by native mangrove communities.
When invasive plants outcompete native plants, the former often compete with the latter for resources to take over the ecological niche or inhibit the latter through allelopathy. However, S. apetala exerts a stronger allelopathic effect on itself than on native mangrove plants, and the autotoxicity to S. apetala seed germination is particularly notable 37 . Thus, S. apetala is unlikely to harm native mangrove plants through allelopathy. Both the Sonneratia species are heliophytes and have no ecological niche overlap with the major native constructive mangrove species 26 . When resources are limited, S. apetala competes intensely with S. caseolaris but only weakly with native species. Thus, S. apetala will not replace native plants and is unlikely to harm native mangrove species. This is the mechanism underlying the fourth stage through which the exotic species ultimately degenerates while native mangrove communities gradually recover to establish forest (Fig. 10f).
Ecological demonstration of the significance of using exotic species for the control of invasive species in order to address worldwide problems. This study provides an example of the use of exotic species to control invasive species and to further restore native plants. An introductory experiment over 20 years has shown that S. apetala grows well in river estuaries in South China; as the pioneer species for artificial mangrove restoration, S. apetala has not caused significant ecological invasion 23 . Moreover, mature S. apetala forest exhibits strong resistance to cold, i.e., it endures a mean monthly temperature as low as 14.1 °C and an extreme low temperature of 0.2 °C. Presently, S. apetala is widespread in the provinces of Hainan, Guangxi, Guangdong, and Fujian, with its northern boundary extending to 28°52′ N 31 . In China, S. apetala is distributed in almost all of the latitudinal zones of the native mangrove plant distribution. As long as the introduced area has extreme low temperatures to constrain seedling regeneration, the possibility of secondary invasion can almost be eliminated. Theoretically, the use of S. apetala to control S. alterniflora can be extended to a relatively large range.
The difference between replacement control and biological control is that the latter directly and specifically kills pests through predators or parasites, whereas the former replaces harmful plants through the natural process of secondary succession, i.e., plant competition in the short term or more complex processes of secondary succession involving a series of plant communities in the long term 9 . Replacement control is a long-term approach that provides ecological benefits, such as soil and water conservation. When conventional biological control methods failed to control S. alterniflora, we broke through the  conceptual bottleneck to find a strong competitor for S. alterniflora, i.e., another invasive species, S. apetala, which successfully controls the growth of S. alterniflora.
A number of studies have been conducted to "control grass with grass". However, the majority of the existing research has focused on how to change resources in the environment and use native plants to control invasive plants 44 . For example, a previous study 28 successfully replaced invasive Flanueria bidentis with native foliage. Almost no studies have explored the potential value of invasive plants. Although one study assessed the possibility of using S. apetala to control S. alterniflora 18 , it involved no long-term field verification. The present study is the first to successfully implement the goal of the "economical, effective, and long-lasting control of S. alterniflora with exotic invasive species". Thus, we believe that "generalized" biological control should be the practice or process through which an undesirable organism is controlled by means of another (beneficial) organism, including the use of not only natural enemies in the narrow sense of biological control but also native species with enhanced competitiveness after environmental resource improvement and the "exotic invasive plants" applied in the present study.
Estuarine wetlands and coastal tidal wetlands are types of ecosystems that provide the highest value of ecosystem services per unit area 45 , but they can be easily invaded 46 . The ecological role provided by mangroves is worth up to $1.6 billion a year. S. alterniflora invasion often results in substantial reductions in the distribution area and number of other native populations 47 . How to effectively control S. alterniflora and concurrently restore mangrove communities is a worldwide problem requiring attention. Since the late 1970 s, a series of measures have been taken in different regions of the world to mitigate mangrove degradation and loss and to restore mangrove forests 16 . The success rate of the previous measures has been fairly low, and a few conventional control methods are likely to harm native species 48,49 .
This study is the first to take advantage of growth competition with an exotic invasive plant (S. apetala) for achieving control over S. alterniflora. This measure takes full advantage of the ecological characteristics of S. apetala and leads to the natural eradication of this species after the completion of its mission, thereby preventing a new invasion. The new approach saves time and money compared with manual control; it avoids pollution and has higher safety than chemical herbicides; additionally, replacement control presents better efficacy, higher safety, and longer-lasting control than conventional biological controls. Even more exciting, during the control process, native plants regain their dominance and are slowly restored. This study provides a new concept for addressing the global problem regarding serious S. alterniflora invasions and the difficulty of mangrove restoration: learn from enemies to compete with enemies, i.e., take full advantage of the growth competition of exotic plants for the control of S. alterniflora. The proposed method has achieved remarkable results and greatly reduces the cost of exotic plant prevention and native plant restoration. More importantly, the restoration of mangrove communities has brought about ecological benefits that cannot be underestimated.