Introduction

Zonal distribution of plant communities is an obvious characteristic of salt marsh ecosystems1,2,3,4. The plant zonation pattern is typically correlated with a combination of multiple factors, such as the saturated or unsaturated flow in the soil2 and sediment accretion5,6. In fact, plant community spatial distribution and abundance of species are both an indication of the relative physiological tolerance of an individual species to abiotic conditions and competition among species1,4,7,8,9.

Plant competition can be affected not only by resource availability10, but also by environmental factors within a salt marsh4,11,12,13, such as sediment9,14,15, sea level4,5 and salinity7,16. It has been shown that competition between species, especially between exotic and native species, often depend on nutrient supply9,17,18, flooding19, elevation4 and other factors20. Elevation, which varies with plant zonation, is highly correlated with the redox potential in salt marsh sediment21. Also, interspecific competition has also identified as always disproportionate, resulting in the replacement of one species by another16. However, there is typically little information available about competition between exotic and native species in different sediments types.

Salt marsh sediments are mainly influenced by plant activity, elevation and tidal flooding14,22,23,24,25. Because much of the organic matter in tidal flooding originates from external sources or from benthic microalgae, sediments are usually rich in organic matter and phosphorus26,27,28. Also, sediments of salt marshes are characterized with high concentrations of sulfides and ammonium originating from the high rates of anaerobic reduction5,23, which can affect plant root respiration29. Research has shown that poor water drainage of marsh sediments can be a critical factor limiting plant root growth and causing decay23. In addition, moderate levels of sediment slurry enrichment can have beneficial effects on the soil by increasing elevation and soil bulk density30, which could assist with the restoration of the macro-invertebrate community and its related habitat31. Thus supplementary sediments from tides ameliorate environmental factors for plant growth for some species27, including Spartina anglica32,33.

Spartina anglica C. E. Hubbard (hereafter called Spartina) is an exotic species originating in England that was first introduced in China in 196334,35. It developed populations successfully for 30 years on Chinese salt marsh36 and provided significant economic benefits, including protection for dams and feed for livestock. Yet, in recent years, Spartina populations have been in decline on the Chinese coast and the cover has decreased to less than 50 ha34,36. Spartina has been pushed to a higher elevation gradient due to the recovery of the salt marsh and the expansion of Spartina alterniflora, which is another exotic and invasive species36. Also there are differences in physical and chemical characters between sediments at low and high tide zones.

Scirpus triqueter L. (hereafter called Scirpus) is species native to China37. It grows on low-lying moist sites, where an overlapping niche exists with Spartina and it likely results in competition between the two plant species. Although research has been conducted regarding the ecological interaction between the two plant species with different nitrogen levels38, little is known of competition between the two plant species under different types of sediment.

In this experiment we used three kinds of sediment, i.e. clay, sand and a mixture of those two sediments, to test competition outcomes between the native species, Scirpus and the exotic species, Spartina. As sediment structure is known to vary within saltmarshes, we investigated how such variation might influence competitive outcomes between Spartina and Scirpus. Here, we quantified competition using various growth measurements when plants were grown with and without their competitor present.

Results

Competition intensity

The more negative the values of the log response ratio (LogRR) were, the greater was competition intensity. Sediment type significantly affected competition intensity, as measured by LogRR, in both species (Table 1). For Spartina, both inter- and intraspecific competitions were greater when it grew in clay than when it grew in sand or in the sand-clay mixture (Figure 1A). For Scirpus, competition in sand and clay did not differ significantly, but it was greater than that in the sand-clay mixture (Figure 1B).

Table 1 Effects of sediment type and competition type (inter- vs. intraspecific competition) on interaction intensity (LogRR) of the two species
Full size table
Figure 1
figure 1

Interspecific and intraspecific log response ratio (LogRR) of Spartina anglica (A) and Scirpus triqueter (B).

Means ± SE are presented.

Full size image

The competition type significantly affected LogRR of Spartina, but not that of Scirpus (Table 1B). For Spartina, interspecific competition was significantly greater than intraspecific competition (Figure 1A). For Scirpus, intraspecific competition did not differ significantly from interspecific competition (Figure 1B).

Growth measures

Sediment type and competition between species affected significantly (P < 0.05) or tended to affect (P < 0.1) all growth measures of both Spartina and Scirpus (Table 2). The interaction between sediment type and competition type was statistically significant for all measures of growth for Spartina (except rhizome length), however the interaction term was never statistically significant for Scirpus (Table 2).

Table 2 Effects of sediment type and species competition on the growth of the two species
Full size table

For Spartina, all growth measures were significantly larger when the sediment was clay than when it was sand or the mixture of sand and clay; they did not differ significantly between the sand and sand-clay mixture treatments (Table 2A, Figure 2, SNK tests). Overall, growth measures of Spartina were the largest in SA2, smallest in SA2 + ST2 and intermediate in SA4 (Table 2A, Figure 2). Such effects were larger when the sediment was clay than when it was sand or the sand-clay mixture (Table 2A, Figure 2).

Figure 2
figure 2

Biomass and asexual characteristics of Spartina anglica in different sediment types and species competition experiments.

Means ± SE are presented.

Full size image

For Scirpus, all growth measures except number of ramets were significantly larger when the sediment was clay than when it was sand or the sand-clay mixture (Table 2B, Figure 3, SNK tests). The number of ramets of Scirpus in clay was significantly higher than that in the sand-clay mixture, but did not differ from that in sand (Figure 3B). All growth measures except number of ramets were significantly larger in ST2 than in SA2 + ST2 and ST4, but they did not differ between SA2 + ST2 and ST4 (Table 2B, Figure 3).

Figure 3
figure 3

Biomass and asexual characteristics of Scirpus triqueter in different sediment types and species competition experiments.

Means ± SE are presented.

Full size image

Discussion

Our results suggest that sediment type has significant effects on the growth of Spartina. Although Spartina could adapt to different sediments1,24,29, it still performs better in more nutrient-rich sediments such as clay39, likely because the ability of clay to preserve moisture and nutrients is greater than sand. Scirpus had good growth performance in all sediments, perhaps because of its low nutrient needs38.

The sediment type in salt marsh depends on factors such as plant species27, sea level40, tide32, tidal creeks15 and elevation21. Tidal flooding gives salt marsh increased vigor because of the sediment component of the substrate, which increases the soil mineral matter and decreases nutrient deficiency30,32. Tidal creeks shape sequential geomorphic features, which receive different types of sediment (coarse or fine) from tides and channels15. The clay is a fine-type sediment driven by hydro-geomorphic processes with low bulk density, which yields poor drainage conditions and limits plant root growth and causing decay15,23. The clay can also change the redox conditions and sulphide concentrations, which can affect plant growth39. For instance, the health of plants shows a sharp decline when the redox potential falls below −50 mv21 and the sulphide concentration can reach about 320 mg kg−1 in the sediment water interface of Spartina alterniflora39. Simply, abiotic factors including the sediment electric conductivity, sediment oxygenation and salinity affect seed germination and subsequent plant growth41. These physical and chemical sediment traits may greatly influence the growth and adaption of plant species30,41,42, such as Spartina and Scirpus, thus changing their competition.

The intensities of interspecific and intraspecific competition for both species were significantly different with sediment types. For Spartina, interspecific competition from Scirpus was larger than intraspecific competition and competition intensity was greatest in clay (Figure 1A). On the other hand, for Scirpus, overall there was no significant difference between interspecific and intraspecific competition and intraspecific competition tended to be higher than interspecific competition in clay (Figure 1B). These results suggest that, when the two species grow together, Scirpus had a stronger competitive effect on Spartina in clay.

Competition between plant species can be changed by environmental factors4,8,11,17,43,44 and our previous study has also shown that nitrogen level could change competition between Spartina and Scirpus38. In this study, we found that competition between Spartina and Scirpus became stronger when the sediment was clay than when they were sand or a clay-sand mixture. Similarly, sediment type has been found to significantly alter competition between Puccinellia maritima and Spartina1 and between S. alterniflora and P. australis43. Changes in competition between Spartina and Scirpus under different sediment type suggest that in the upper zone salt marsh where Spartina is currently distributed and where Scirpus is abundant replacement of Spartina by Scirpus may happen. Therefore, we predict that further declines in Spartina are likely to take place in the upper zone salt marsh. Because S. alterniflora invasion drives up45,46, there are some differences in sediment types among low and high zone salt marsh47,48,49,50. However, for a more accurate prediction, effects of other environmental factors such as tide action and salinity on competition between two species should be taken into consideration. Our findings might facilitate the development of schemes to control the Spartina invasion that is occurring in some countries in the world.

Methods

The species

Spartina is a rhizomatous perennial grass that is highly invasive in German, Australian, Irish estuarine mudflats, sand flats and salt marshes35,51,52. The plant may reach a height of 50–100 cm and its leaf blades are flat or in-rolled and 5–12 mm wide. Spartina is mainly wind pollinated and produces abundant flowers and viable seeds in Europe36. However, on the coastal areas of China, Spartina rarely produces viable seeds due to its poor pollen quality and abnormal pollen tube; it spreads mainly by clonal growth36,53. The growth form of Spartina varies in different habitats35.

Scirpus is a rhizomatous perennial sedge native to China37. The plant may reach a height of about 20–100 cm. Its stem is trigonous, with leaves 1.5–5.5 cm in length and 1.5–2.5 mm in width. Scirpus flowers and produces seeds from June to September. It occurs in different habitats in tidal wetlands, ranging from brackish to freshwater along the coast37.

Experimental design

Plants of both Spartina and Scirpus were collected from the mash zone of Xinyang Harbor in Yancheng Wetland National Nature Reserve in Jiangsu Province, China. The plant collection was authorized by the management of Yancheng Wetland National Nature Reserve because the field studies did not involve endangered or protected species. Similar-sized plants (ramets), each consisting of a single tiller with attached roots, were selected and used for this experiment. All the ramets of both species were collected within about 500 m, which should reduce the chance of different genotypes. All selected ramets were about 6 cm tall and were cultivated in pots (28 cm in diameter and 20 cm in height) containing a 1:1 (v:v) mixture of sand and clay. For the experiment, we used similar-sized ramets with a height of 12.6 ± 0.5 cm (mean ± SE) for Spartina and 15.2 ± 0.5 cm for Scirpus. The ramets were selected at random from the cultivated stock population to reduce the possible influence of clonal variation and plant history.

The experiment had three types of sediment and five species combinations in a factorial design, resulting in 15 treatments. The three sediment types were (1) sand, (2) clay and (3) a mixture of sand and clay at a volume ratio of 1:1. The amounts of available nitrogen and phosphorus were measured before the experiment. The amount of available nitrogen was 43.65 ± 2.49 (mean ± SE, n = 3) mg kg−1 in sand, 76.34 ± 1.23 mg kg−1 in clay and 52.5 ± 3.59 mg kg−1 in the sand-clay mixture. The amount of available phosphorus was 1.42 ± 0.25 mg kg−1 in sand, 8.31 ± 0.76 mg kg−1 in clay and 5.09 ± 0.18 mg kg−1 in the sand-clay mixture. There were five species combination treatments, i.e. each pot was planted with (1) two ramets of Spartina (coded as SA2), (2) two ramets of Scirpus (ST2), (3) four ramets of Spartina (SA4), (4) four ramets of Scirpus (ST4) and (5) two ramets of Spartina and two ramets of Scirpus (SA2 + ST2).

We sampled the salt water in the natural habit of Spartina and Scirpus to measure the salinity, which was about 1.48 ± 0.02% (mean ± SE, n = 3). To simulate the marsh conditions that these two species commonly experienced, salt water containing 1.5% NaCl was added to the pots and water was maintained 2 cm above the soil surface level. The salt water was produced by dissolving crude salt into tap water. The crude salt had been directly extracted from seawater in Jiangsu Province. The salinity content in the water in the pots was monitored weekly and adjusted to the initial conditions (1.5% NaCl) when it was below 1.4% or above 1.6%. The experiments lasted 26 weeks (from 20 May to 4 December) and were carried out in a greenhouse at the Pukou campus of Nanjing University. There were six replicates for each treatment.

Data collection

At harvest, we measured total biomass, the number of ramets and rhizome length of each plant for the two plant species. Biomass is a typical measure of plant growth, the number of ramets a measure of asexual (clonal) reproduction and rhizome length a measure of clonal expansion (the distance of lateral clonal spreading). We counted the number of ramets and measured the rhizome length of Spartina and Scirpus separately. Plants of Spartina and Scirpus were dried at 80°C for 72 h and weighed.

No plants produced flowers or set seeds during the experiment. One plant of Spartina in SA2 + ST2 grown in the sand-clay mixture died during the experiment and this replicate was excluded from harvest and subsequent analysis for the two species.

Data analysis

Before analysis, we normalized the data from each plant. For each replicate, we calculated biomass, number of ramets and rhizome length per initial ramet. For instance, for SA2 the final biomass was divided by two and for SA4 it was divided by four. Thus all these measures were standardized to per initial plant level. The subsequent analyses were based these derived data.

We used two-way ANOVA to test the effects of sediment type and species competition on all growth measures of each species. Sediment type and species competition were treated as two fixed factors. Sediment type had three levels (sand, sand-clay mixture and clay) and species competition had also three levels (no competition, with intraspecific competition and with interspecific competition). For species competition of Sparina, SA2, SA4 and SA2 + ST2 were used, treating SA2 as no competition, SA4 as with intraspecific competition and SA2 + ST2 as with interspecific competition. Similarly, for competition of Scirpus, ST2, ST4 and SA2 + ST2 were used, treating ST2 as no competition, ST4 as with intraspecific competition and SA2 + ST2 as with interspecific competition. We used the Student-Newman-Keuls (SNK) test to compare the overall means among the three sediment treatments and among the three competition treatments.

To measure the intensity of interspecific competition, we calculated the log response ratio (LogRR) as LogRR = log (B+/B0), where B+ is biomass of plants in the presence of the interspecific neighbors and B0 is the mean biomass of plants in the absence of neighbors across the six replicates55,56. Similarly, we also measured the intensity of intraspecific competition by calculating LogRR, where B+ is biomass of plants in the presence of the intraspecific neighbors and B0 is the mean biomass of plants in the absence of neighbors across the six replicates54,55. A positive value of LogRR indicates facilitation and a negative value indicates competition54,55,56.

In this study, the absence of neighbors (i.e. no competition) meant the treatments in which one pot was planted with two ramets (either Spartina or Scirpus), the presence of intraspecific neighbors (with intraspecific competition) referred to the treatments in which one pot was planted with four ramets of the same species and the presence of interspecific neighbors (with interspecific competition) referred to the mixture treatment in which each pot was planted with two ramets of both Spartina and Scirpus. In other words, intraspecific competition encompassed competition mainly from the same species, while interspecific competition encompassed competition mainly from other species. We used two-way ANOVA to test the effects of sediment type (three levels: sand vs. sand-clay mixture vs. clay) and interaction type (two levels: intraspecific competition vs. interspecific competition) on LogRR.

Statistical analyses were conducted with SPSS 18.0 for Windows (SPSS Inc., USA). The effects were considered significant if P < 0.05. Measures of biomass were log transformed to improve homogeneity of variance prior to ANOVA.