Introduction

Tidal saltmarshes are critical transition zones that perform an essential function in maintaining the health and ecology of estuaries (Teal and Howes 2000; Hughes 2004; Rosso et al. 2006). Normal saltmarsh functioning, especially throughout the vegetation component, is closely related to a series of natural factors such as competition, salinity, degree of waterlogging, and nutrient status. All these factors play extremely important roles in regulating productivity (De Leeuw et al. 1990; Gross et al. 1990; Pennings and Callaway 1992; Pont et al. 2002), and therefore affect regulation of processes throughout the ecosystem.

Invasion by exotic species has been recognized as a major threat to the conservation of these ecosystems (Kriwoken and Hedge 2000). The genus Spartina includes 17 species with main native ranges in coastal America, Europe, and North Africa and some of these species have been intentionally or accidentally introduced outside their native ranges in numerous coastal regions (Ranwell 1967; Callaway and Josselyn 1992; Baumel et al. 2003; Hedge et al. 2003; An et al. 2007). Most Spartina species show high competitive ability and are able to exclude native species in coastal saltmarsh and other intertidal, and sometimes subtidal, environments (Ranwell 1967; Hedge et al. 2003; Grevstad 2005; Rosso et al. 2006; Wang et al. 2006; An et al. 2007). Hence, Spartina invasions have significant impact on native flora and fauna communities and on the invaded ecosystem processes, such as carbon and nitrogen cycling and dynamics of sediment organic matter (Callaway and Josselyn 1992; Kriwoken and Hedge 2000; Rosso et al. 2006; Neira et al. 2007), causing considerable economic loss (Daehler and Strong 1994, 1996; Wang et al. 2006). Nowadays, invasion by exotic Spartina is a complex coastal zone management problem that challenges the ability of managers to develop and implement an effective and timely invasion management response (Hedge et al. 2003; Grevstad 2005; An et al. 2007).

S. densiflora Brong., one of the most important invasive species of the genus, is invading Europe (Castillo et al. 2000, 2005) and North America (Kittelson and Boyd 1997) from South American marshes where it is native (Bortolus 2006). In the Iberian Peninsula it was accidentally introduced in the Gulf of Cádiz (south-western Spain), probably in the 16th century (Nieva 1996), and colonized a wide range of habitats including dunes, high marsh and levees, salt pan, and intertidal flats, mainly on the southwest coast of Spain (Odiel marshes), often colonizing entire habitats and converting them into cordgrass meadows (Nieva et al. 2001, 2003; Costa et al. 2003; Ayres et al. 2004; Castillo et al. 2005). In Portugal its occurrence is only documented in the southeast coast, in the estuary of river Guadiana, inside the saltmarsh of the Natural Reserve of Castro Marim (INAG 2001). However, it is progressively spreading and it is already in close contact with Arthrocnemum macrostachyum, a native evergreen perennial stem-succulent, which is one of the most widespread native species in the middle area of the intertidal zone of the Castro Marim saltmarsh. Due to its competitive characteristics, like tall canopy, dense tussocks, and abundant seed production and germination (Nieva 1996), it is expected that S. densiflora might replace the native species, eventually leading to the disruption of the ecosystem functioning.

Several studies have assessed above and belowground biomass and net annual primary productivity in native Spartina species (Gallagher et al. 1980; Cranford et al. 1989; Gross et al. 1990, 1991). However, to our knowledge, no comparison of productivity and nutrient dynamics has been made between invasive Spartina species and native species of a different genus. Also, despite halophytic species differ widely in the extent to which they accumulate nutrients, and in their capacity to adjust nutrient concentrations to fluctuations of salinity (Blumwald 2000; Khan et al. 2000), no information is available regarding S. densiflora and A. machrostachyum. Therefore, it is crucial to get deeper understanding of structural and functional details of S. densiflora to assess its establishment mechanisms and to control its continuing expansion. In this context, a study was developed to (a) evaluate and compare temporal patterns of above and belowground biomass; (b) assess and compare net aboveground primary productivity by two different methods; and (c) determine the monthly variation of nutrient (N, P, K, Ca, Mg and Mn) concentrations in above and belowground components of Spartina densiflora and Arthrocnemum macrostachyum, in the Castro Marim saltmarsh (Portugal).

Material and Methods

Study Site

The study was carried out in the “Reserva Natural do Sapal de Castro Marim e Vila Real de Santo António”, in southeastern Portugal (lat.37º13´N; long.1º40´W). This reserve covers about 2,087 ha, of which 28% are saltmarsh between the Guadiana River and the Atlantic Ocean. This area has a Mediterranean climate, characterized by a long dry period in summer, with a mean annual rainfall of 492 mm and air temperature of 17.2°C. During the study period (2001/2002), the monthly maximum rainfall was in September (100 mm), and the total was 593 mm. The maximum absolute temperature was 35.9°C in July and the minimum was 5.8°C in December (Vila Real de Santo António Meteorological Station, unpublished data; www.meteo.pt).

The topography of the study area is mostly flat with an altitude ranging from 0 to 5 m above the mean level of water. The soils (fine texture), developed on alluvial sediments, are mainly halomorphic with low or negligible content of lime, and a pH that ranges from 6.0 to 6.7 (Lousã 1986). The soil salinity level is high and electrical conductivity ranges from 20 to 360 dS m-1 in dense vegetation areas, reaching 675 dS m-1 in unvegetated areas (Lousã 1986). Soils adjacent to the alluvial influenced area are mostly developed on shales and limestone.

The vegetation shows a clear vertical zonation according to the frequency of inundation and soil characteristics (Lousã 1986). The low marsh vegetation is dominated by Spartina maritima (Curtis) Fernald. In the middle marsh, the vegetation is dominated by A. macrostachyum (Moric.) Moris, Atriplex portulacoides L., Arthrocnemum perenne (Miller) Moss., A. fruticosum (L.) Moq., in similar proportions (from 10 to 20% ground cover), and to a less extent the invasive S. densiflora (about 10%). The high marsh is mainly composed by Limoniastrum monopetalum (L.) Boiss and Suaeda vera Forssk. ex Gmelin.

Sampling

An area of approximately 1 ha of mid-saltmarsh (near the Esteiro das Lezírias), where S. densiflora (SD) and A. macrostachyum (AM) occurred in similar proportions, was chosen for sampling purposes. Sampling was conducted in five months during 2001/2002, in the middle of each season (October-01, January, April, July and October-02). In the beginning of the study, 15 individuals of each species were randomly selected and labelled. At each sampling date, aboveground biomass of three plants per species was individually harvested from 50 × 50 cm quadrats. Total alive and dead biomass collected from each quadrat was separated into live leaves, live stems, reproductive organs, belowground biomass and necromass (dead leaves and dead stems) for SD; live photosynthetic stems, live woody stems, belowground biomass and necromass (dead stems) for AD. After drying (80°C), biomass components were weighed.

Leaf area was measured in a subsample of each plant leaf or photosynthetic stem fraction, using a LI-3000A Portable Area Meter.

At the same sampling dates and beneath the same individuals used for aboveground biomass estimates, belowground biomass was sampled through the extraction of three soil cores (8 cm diameter) per plant (nine cores per species), to a depth of 15 cm, with a thin-wall, sharpened stainless steel tube. After washing to separate from soil material, belowground biomass of each core was dried and weighed as for aboveground biomass.

Samples of each biomass component collected from the same species and sampling date were pooled, resulting in one sample per organ and collection date, and a subsample was taken for chemical analysis.

Laboratory Procedures

Samples of plant material were dried and ground to pass through a 1 mm mesh screen, and analysed for nutrient (N, P, K, Ca, Mg and Mn) concentrations. Total N was determined using Kjeldhal digestion (Digestion System 40, Kjeltec Auto 1030 Analyzer). The mineral elements (Ca, Mg, K and P) were determined after ashing (6 h at 450°C) and taken up in HCl. Ca, Mg, K and Mn were determined by atomic absorption spectrophotometry. Concentration of P was measured colorimetrically by the molybdenum blue method.

Data and Statistical Analysis

Specific leaf area (SLA) was calculated by the ratio between leaf area and leaf dry weight. Leaf area index (LAI) was calculated as the ratio between total plant leaf area and the canopy cover projected on the ground (2,500 cm2).

Seasonal aboveground relative growth rate (RGR) was calculated for each sampled individual as:

$$ RGR = {{{\left( {\ln D{W_s} - \ln D{W_{s - 1}}} \right)}} \left/ {t} \right.} $$

where DWs is the dry weight of aboveground biomass in the middle of a particular season; DWs-1 is the dry weight of aboveground biomass in the middle of the previous season; and t is the time period in days (90). For each species and growth period, the three individuals sampled were averaged.

Aboveground production was estimated for both species using two different methods: the Smalley method (described in Groenendijk 1984), which is based solely on biomass changes, and the Wiegert and Evans method (Wiegert and Evans 1964), which also takes in account the disappearance of dead material.

Two sample Student’s t-tests were performed to assess differences between the two species regarding peak aboveground biomass, relative growth rates in each period, net above primary productivity and the turnover of the aboveground live biomass. Differences in all the other biomass data between species and months and their interaction were determined with two-way ANOVA (GLM). If significant differences occurred, Scheffé tests for multiple comparisons were used to identify significant differences between months in each species. Some of the data were transformed (natural log or square root) for homogeneity of variances. Analyses were performed using SPSS 16.0 (Apache Software Found., USA).

Results

Biomass Across Seasons and Species

The total aboveground standing biomass of S. densiflora (SD) was significantly higher (P < 0.001) than that of A. macrostachyum (AM) (3,779 ± 455 and 1,215 ± 91 g m-2, respectively). The difference was even more pronounced for the total belowground biomass of the two species (Table 1). However, live aboveground biomass showed no significant differences between AM and SD (Table 1).

Table 1 Results of two-way ANOVA analysis for the effects of species, month and their interaction on total aboveground biomass (TAGB), live aboveground biomass (LAGB), belowground biomass (BGB) and root/shoot ratio (RSR) of Arthrocnemum macrostachyum and Spartina densiflora, from Oct-01 to Oct-02
Full size table

Peak above and belowground biomass was observed in April for SD and in July and January for AM (Fig. 1). Significant among-month differences (P < 0.001) were detected in SD aboveground biomass, while AM aboveground and both species belowground biomass showed no marked oscillations.

Fig. 1
figure 1

Mean monthly above (top) and belowground (bottom) biomass of Arthrocnemum macrostachyum (AM) and Spartina densiflora (SD) (LPS/LL=life photosynthetic stems/live leaves, LWS/LS=live woody stems/live stems, respectively for AM and SD; NM=necromass; BGB=belowground biomass). Bars represent the standard error of the mean total above and belowground biomass (n = 3). Different letters on the top of columns indicate significant differences (P < 0.001) in total aboveground biomass between sampling months; belowground biomass did not differ significantly across months

Full size image

The root/shoot ratio (RSR) was significantly higher (P < 0.001) in SD (1.7 ± 0.2) than in AM (0.9 ± 0.1) (Table 1). Despite the pronounced decrease observed in this ratio in July, for both species, this trend was consistently observed across all seasons of the study period.

The aboveground standing biomass of SD primarily consisted of necromass (dead standing components), which accounted for 70% of total. In contrast, most of the aboveground biomass of AM consisted of live components (77%). Although the biomass of photosynthetic components (live leafy stems for AM and live stems and leaves for SD) was higher (P < 0.001) in SD, they represented similar proportions of total aboveground biomass in both species (25 and 28%, respectively). Woody stems comprised 52% of total in AM.

Necromass peaked in October (76-77% of the total) for SD and in January (32%) for AM (Fig. 1). The proportions of SD live leaves and stems increased from October-01 (23–24% of total) to April (33%), decreasing afterwards, while proportion of AM live photosynthetic stems varied from 22% in January to 25–28% in October. Thus, monthly variation in LAI (Fig. 2) was more pronounced (P < 0.01) for SD (ranging from 0.6–0.9 in October to 2.0 in April) than for AM (0.7–0.9). SLA of both study species (P = 0.573) showed weak variations (Fig. 2) with time (28–32 cm2 g-1), except for a minimum (21-22 cm2 g-1) in July.

Fig. 2
figure 2

Variation of specific leaf area (SLA) and leaf area index (LAI) of Arthrocnemum macrostachyum (AM) and Spartina densiflora (SD) across sampling months (values are means±S.E. for three plants) and results of two-way ANOVA analysis for the effects of species, months and their interaction on SLA (Species: F = 0.33, P = 0.57; Month: F = 11.74, P < 0.001; Interaction: F = 1.51, P = 0.24) and LAI (Species: F = 11.40, P < 0.01; Month: F = 6.65, P < 0.01; Interaction: F = 5.38, P < 0.01). Different letters indicate significant differences between sampling months; asterisks indicate significant differences between species

Full size image

Relative Growth Rate and Primary Productivity

Maximum relative growth rate was observed in the April/July period for AM (3.37 ± 1.26 mg g-1 day-1) and in October/January for SD (4.92 ± 0.36 mg g-1 day-1). Both species showed negligible growth between July and October (Table 2).

Table 2 Mean relative growth rate (RGR) of Arthrocnemum macrostachyum (AM) and Spartina densiflora (SD) from Oct-01 to Oct-02
Full size table

Whatever the method used, net aboveground primary productivity (NAPP) was higher in SD than in AM (Table 3). The difference was four-fold following the Smalley method and three-fold by the Wiegert and Evans method.

Table 3 Mean live aboveground biomass (LAGB), peak live aboveground biomass (PLAB), mean belowground biomass (BGB) and net aboveground primary productivity (NAPP) of Arthrocnemum macrostachyum (AM) and Spartina densiflora (SD) from Oct-01 to Oct-02
Full size table

The turnover of the aboveground live biomass (primary production/peak live biomass) of AM was about half that of SD (Table 3).

Nutrient Concentrations in Biomass Components

Live aboveground biomass of AM showed higher weighted K (8.13 mg g-1) and Mg (5.08 mg g-1) concentrations than belowground biomass (3.70 and 5.01 mg g-1, respectively). In contrast, N (11.72 mg g-1), P (0.90 mg g-1), Ca (6.90 mg g-1) and Mn (0.40 mg g-1) concentrations were higher in belowground than in aboveground biomass (8.32, 0.79, 3.05 and 0.02 mg g-1, respectively). For SD, P and K concentrations were higher in live aboveground (0.80 and 8.82 mg g-1) than in belowground biomass (0.65 and 3.02 mg g-1), whereas those of N, Ca, Mg and Mn were higher in belowground (6.62, 3.86, 3.78 and 0.09 mg g-1) than in aboveground biomass (5.77, 1.15, 2.17 and 0.03 mg g-1).

Concentrations of N, P, K, Ca and Mg in photosynthetic structures were higher for AM than for SD (Figs. 3 and 4). The highest N concentration in live biomass was determined in July and the lowest in January and April, for both species. Maximum P and K values were observed in October and January for live leaves and in April for belowground biomass (Fig. 3). Concentrations of Ca and Mg in live leaves reached a maximum in AM and, in contrast, a minimum in SD during April (Fig. 4). In belowground biomass Ca concentration of AM was about twice that of SD. Mn minimum values were observed in July for both species (Fig. 4).

Fig. 3
figure 3

Concentrations of N, P and K (mg g-1) in the biomass components of Arthrocnemum macrostachyum (AM) and Spartina densiflora (SD) from Oct-01 to Oct-02 (LPS=life photosynthetic stems; LWS=live woody stems; LL=live leaves; LS=live stems; NM=necromass; BGB=belowground biomass)

Full size image
Fig. 4
figure 4

Concentrations of Ca, Mg and Mn (mg g-1) in the biomass components of Arthrocnemum macrostachyum (AM) and Spartina densiflora (SD) from Oct-01 to Oct-02 (LPS=life photosynthetic stems; LWS=live woody stems; LL=live leaves; LS=live stems; NM=necromass; BGB=belowground biomass)

Full size image

Discussion

Seasonal Growth and RGR

Study species showed similar live aboveground biomass, but they differed significantly in amounts of standing dead material and in the root/shoot ratio. Major differences were also observed in seasonal patterns of growth between the exotic and the native species, suggesting that the exotic may be a successful invader of these marshes and alter patterns of nutrient cycling.

The high amounts found for total biomass of both study species are in the range observed by Curcó et al. (2002) for middle marsh of other Mediterranean-type climate saltmarshes. According to these authors, the high aboveground biomass in the middle marsh may be associated with the moderate waterlogging and hypersalinity, which are stress factors in the low and upper marsh, respectively. The higher biomass and RGR of SD as compared to AM suggests that the former has a high potential to spread within and among local marshes, as reported for other invasive species (Minchinton and Bertness 2003). This makes SD an aggressive competitor with the native saltmarsh species (Wang et al. 2006).

The high SD root/shoot ratio, also reported for other saltmarsh plants, including Spartina species (Lana et al. 1991), which suffer periodic waterlogging, indicates a high investment in the root system, which may be valuable under stressful conditions, even to colonize the lower marsh (Scarton et al. 1999; Pont et al. 2002; Neves et al. 2007). However, the success of any invasive species can only be fully evaluated in the context of the invasive species and the invaded habitat interactions (Callaway and Josselyn 1992). Castro Marim is a highly disturbed saltmarsh, as a result of land reclamation, water diversion, waste disposal, and other factors, and the filling, diking, and dredging of the area have left a small percentage of the original tidal saltmarsh and many intertidal sites unvegetated (ERENA 2007). Also, the reflooding of the area for marsh restoration has opened new areas to colonization. Since native species are generally slow to colonize new marsh areas (Callaway and Josselyn 1992) and disturbances are reported to decrease their biomass and, in contrast, to increase the density, height, and biomass of saltmarsh invasive species (Minchinton and Bertness 2003), it is predictable that SD may spread within Castro Marim saltmarsh. Moreover, based on its invasive history in Europe (Castillo et al. 2000, 2005) and North America (Kittelson and Boyd 1997), it is clear that once established, SD might quickly spread within a local area, displacing the matrix vegetation.

Study species differed in the degree of senescence, as shown by the much greater amounts of necromass in SD than in AM, following the pattern observed by Callaway and Josselyn (1992) in S. alterniflora. High amounts of standing dead material decrease flow energy by increasing frictional effects and may create dieback areas rich in mud and organic matter (Neira et al. 2007).

Both study species showed high biomass production in April, a period of favorable radiation and temperatures. However, major differences between the invasive and the native species were observed in seasonal growth, since SD initiated growth earlier and also had greater biomass than AM throughout the year. For SD peak aboveground biomass occurred in April, as a result of maximum RGR in the October/January period, which corroborates previous studies on S. alterniflora (Gallagher et al. 1980). For AM maximum RGR was achieved later, in April/July, and thus the highest live aboveground biomass was recorded in July. The trend observed in the present study for the native species agrees with studies on Atriplex portulacoides (Groenendijk 1984; Bouchard et al. 1998). According to Curcó et al. (2002), the difference found between the study species might be associated with the greater tolerance of AM to high salinity and higher sensitivity to flooding, whereas SD grows better in anaerobic soils, but is more sensitive to extreme salinity. In addition to greater biomass, differences in the initiation of growth allow SD to shade out AM along the mid marsh where they grow together and outcompete for light early in the growing season. Because almost all saltmarsh plants are long-lived, vegetatively-reproductive perennials, the initial occupation of space by roots and rhizomes is important in determining which plants dominate the saltmarsh (Callaway and Josselyn 1992). The 2-fold increase in LAI of SD, followed by a pronounced reduction in July (Fig. 2), represents an adaptive trait to the Mediterranean climatic constraints (Larcher 1995; Neves et al. 2007). During spring favorable temperature and moisture conditions plants made a greater investment in photosynthetic components, decreasing the transpiratory surface under the summer high temperatures and intense drought.

NAPP and Turnover Rates

In accordance with RGR, NAPP was much higher in SD than in AM, for either estimation method. The values obtained for AM (692–1,012 g m-2 yr-1), although higher than those (189–835 g m-2 yr-1) reported by Curcó et al. (2002) for the same species in Spain, are similar to those (678–1,123 g m-2 yr-1) obtained by Ibàñez et al. (1999) and Scarton et al. (2002) for Arthrocnemum fruticosum, in France and Italy. NAPP values estimated for SD (2,603–2,923 g m-2 yr-1) are close to that (2,391 g m-2 yr-1) reported by Costa and Peixoto (2004) for the same species and in the range (1,300–3,700 g m-2 yr-1) reported by Gallagher et al. (1980) for S. alterniflora. They are, however, much higher than reported by other authors (Cranford et al. 1989; Lana et al. 1991) for Spartina alterniflora (113–370 g m-2 yr-1) or S. anglica (1,162–1,649 g m-2 yr-1, Groenendijk 1984). The values obtained in the present study for the invasive species are in the maximum range (2,917–3,635 g m-2 yr-1, for L. monopetalum) reported for Mediterranean saltmarsh species (Neves et al. 2007). According to Gray et al. (1991), the NAPP of Spartina species is expected to increase with the ongoing climatic changes, either by increased conversion efficiency in high CO2 environment through reduced photorespiratory losses, or by the effect of elevated temperatures which enable plants to increase early spring growth. The increase in the size and photosynthetic capacity of SD canopy under elevated atmospheric CO2 indicates the possibility of a long-term C sink, which could partially mitigate anthropogenic increases in atmospheric CO2. Nevertheless, the fluctuation of other environmental variables, like salinity, flooding, and drought periodicity, as well as competition, can moderate plant productivity. Moreover, increased atmospheric CO2 has been shown (Mayor and Hicks 2009) to stimulate sustained increases of above and belowground biomass rates of Scirpus olneyi, a C3 plant species, whereas no growth enhancement was observed in neighbouring C4 plants of Spartina patens. Therefore, further long-term studies must be done on SD and AM, C4 and C3 plant species, respectively, to a more comprehensive understanding of their competitive interactions and their role on the net C balance in these ecosystems.

The NAPP values based on the Smalley method were lower than those estimated by the W-E method. The difference may be related to the fact that the former method is based solely on biomass changes, whereas the latter also takes into account the disappearance of the dead material. Several authors consider that the W-E method provides the best estimates (Linthurst and Reimold 1978; Gallagher et al. 1980; Groenendijk 1984), while others assert that this method overestimates NAPP (Hopkinson et al. 1980). Estimates by the W-E may differ from actual NAPP because they are influenced by transport (by wind or water) of litter out of the study area (Linthurst and Reimold 1978). The actual value of NAPP should lie between the overestimate from the W-E method and the underestimate from the Smalley method.

Turnover rate for AM doubled that reported by Curcó et al. (2002) for the same species and A. fruticosum in Spain, which may be associated with specific hydrodynamic and soil characteristics of the studied saltmarsh. The estimated AM turnover rate is in the range reported for other European marsh species, such as Atriplex portulacoides (0.7–1.0 yr-1) in France (Bouchard et al. 1998; Bouchard and Lefeuvre 2000) and Portugal (Neves et al. 2007), Elytrigia aetherica, Festuca rubra and Puccinellia maritima (0.7, 0.8 and 1.0 yr-1, respectively) in France (Bouchard and Lefeuvre 2000), and L. monopetalum (0.9 yr-1) in Portugal (Neves et al. 2007). The turnover rate estimated for SD (1.7 yr-1) is higher not only than that for AM but also than the rates reported for most European saltmarsh species. According to Nieva et al. (1999), high NAPP and biomass turnover rate for SD indicate a high carbon uptake capacity, which may represent competitive advantage over AM and other native saltmarsh species, in response to the varying environmental conditions of disturbed saltmarshes.

Nutrient Concentrations

Saltmarshes often have salt saturated soils, where Na and Cl concentrations frequently exceed by one or two orders of magnitude those of most macronutrients. Sodic soil solutions can, therefore, depress nutrient-ion activities and create extreme ratios of Na/Ca, Na/K and Mg/Ca (Qadir and Schubert 2002). As a result, plants, which vary not only in the rate at which they absorb an available nutrient, but also in the manner by which they distribute that element spatially within their organs, become susceptible to high osmotic stress, specific ion toxicity and nutritional disorders (Qadir and Schubert 2002).

The high N concentrations found in living tissues of both study species in July might be related to the high N requirement by halophytes for cytoplasmic osmoregulation, in response to the high levels of salinity during the dry season (Gorham and Wyn Jones 1983). The lowest N concentrations at the end of the growing period may be ascribed to N dilution while plants accumulate biomass, as reported by other authors (Gallagher et al. 1980; Boyer et al. 2001; Neves et al. 2007) for several saltmarsh species.

In close relation to the efficient functioning and utilization of N, concentration of P can limit the productivity of Spartina, as reported by Mendelssohn and Morris (2000), in a South Carolina saltmarsh. In Castro Marim saltmarsh, P does not seem to be a limiting factor to SD growth. The decrease of P concentration in leaves of both species in April, reaching the minimum during the drought period, may be in relation with translocation, during senescence, to stems (live stems in SD and woody stems in AM). Because samples were pooled for nutrient analysis, there are some limitations in comparison of nutrient patterns between species. Even so, the trend of higher P concentration (see Fig. 3) and lower N/P ratio in photosynthetic active components and belowground biomass of SD (7 and 10), as compared to AM (11 and 13), suggests that SD may have high ability to absorb P and compete with other species. Also, SD may have a more efficient N use.

The tendency for lower K concentrations in photosynthetic tissues of SD than of AM suggests that the former may be more suited to limitations of K availability; also, its greater biomass and lower K concentration may be associated with a more efficient use of this nutrient.

We may emphasize that Ca and Mg concentrations in photosynthetic tissues were also lower in SD than in AM; moreover, Mn concentrations in belowground biomass of AM were higher than in that of SD, suggesting different root Mn exclusion and capacity to oxidize the rizosphere sediments (Rozema et al. 1985). These trends suggest that SD differs from AM regarding the use of available nutrients, with implications for its competitive ability. Nevertheless, further studies should be developed to get deeper insights on the processes and mechanisms of nutrient use by both study species.

Conclusions

Our results show that in the Castro Marim saltmarsh S. densiflora is more productive than a native species (A. macrostachyum), which is associated with its higher biomass, productivity and turnover rate. As biomass production by S. densiflora occurs earlier (winter-spring period) than by A. macrostachyum (spring-summer), differences in productivity may increase within the expected scenario of climate changes. Therefore, S. densiflora seems to have a high potential to spread within the saltmarsh and displace the matrix vegetation. The different trends observed in nutrient concentrations of biomass components in S. densiflora and A. macrostachyum suggest that the exotic species has a more efficient use of N and K, and a higher ability to absorb P. These traits, which may represent competitive advantage over native species, suggest that S. densiflora may be a successful invader of these marshes and alter the patterns of organic matter and nutrient cycling. Further studies should be developed to understand such ability and competition capacity, and to lay down guidelines for saltmarsh management and conservation purposes.