Geochemical factors promoting die-back gap formation in colonizing patches of Spartina densiflora in an irregularly flooded marsh

Estuarine, Coastal and Shelf Science 189 (2017) 104e114

Contents lists available at ScienceDirect

Estuarine, Coastal and Shelf Science journal homepage: www.elsevier.com/locate/ecss

Geochemical factors promoting die-back gap formation in colonizing patches of Spartina densiflora in an irregularly flooded marsh Nicolai Mirlean*, Cesar S.B. Costa
lia km 08 Campus Carreiros, 96201-900, Rio Grande, RS, Brazil Federal University of Rio Grande, Oceanography Institute, Av. Ita

a r t i c l e i n f o

a b s t r a c t

Article history: Received 15 June 2016 Received in revised form 24 February 2017 Accepted 3 March 2017 Available online 6 March 2017

Circular (RP) and ring-shape (RP) patches of vegetation in intertidal flats have been associated with the radial expansion of tussock growth forms and die-back gap in older central stands, respectively. RP formation has not yet been sufficiently explained. We accomplished a comparative geochemical study of CP and RP structures of Spartina densiflora within a single saltmarsh in a microtidal estuary (<0.5 m). The pore water under these structures demonstrated distinctive physical-chemical properties by marked seasonal changing in water level and salinity. During high-water period dissolved H2S was frequently low in pore waters of S. densiflora structures due to reactive-Fe, which scavenge the sulfide from solution and form solid sulfides. During less flooded-brackish water period, pore water pH goes down below 4 inside the vegetated bordering areas of RP. In these locations the concentration of soluble sulfides dramatically increases up to 140 mM L
1. The high concentration of protons in pore water is the result of solid sulfides atmospheric oxidation to sulfuric acid. High dissolution of H2S, along with the low pH, creates a toxic environment for S. densiflora and die-back central gap formation in RP. CP structure was 5 cm higher in the intertidal than RP but shows frequent presence of a water layer, less severe oxidation of sulfides and limited building-up of toxic condition to plants. Development of S. densiflora RP probably indicates the uplift of sediment by this bioengineer grass and/or periodic lowering of the water surface below a certain critical level. © 2017 Elsevier Ltd. All rights reserved.

Keywords: Estuary Saltmarsh Sediment Diagenesis Spartina dieback Patch configuration

1. Introduction The development of vegetation in saltmarshes is a complex process involving a number of biotic and abiotic factors. The spatial expansion of plant stands could have irregular shape as well as symmetrical shape in form of circles and rings. The circular patch corresponds to the normal radial expansion of plant species, especially vegetative propagation by rhizomes with tussock formation, on a homogeneous substrate in the absence of a strong competitive species (Feist and Simenstad, 2000; Lewis et al., 2001; Perillo and Iribarne, 2003; Dennis et al., 2011; Marangoni and Costa, 2012). Widely separated circular patches at the leading edge of the mud flat invasions grow until coming in contact with another patch, however very frequently after years and even decades of growth, the central part of the patches show signs of die-back, leaving a depleted plant cover or empty central gap and finally a ring-shape patch (Lewis et al., 1990, 2001; Castillo et al., 2003;

* Corresponding author. E-mail address: [email protected] (N. Mirlean). http://dx.doi.org/10.1016/j.ecss.2017.03.006 0272-7714/© 2017 Elsevier Ltd. All rights reserved.

Perillo and Iribarne, 2003; Minkoff et al., 2006; Escapa et al., 2015). Meanwhile these central gaps may be colonized by other successional plant species (Lewis et al., 1990; Castellanos et al., 1994; Alberti et al., 2008). The formation of vegetation ring patches in estuarine and coastal areas has not been sufficiently explained yet, in fact, it has been almost overlooked in scientific literature. Fonseca and Kenworthy (1987) explained the appearance of round patches of the sea grass Zostera marina L. by the action of waves and currents, contributing to alluvium deposits and raising the bottom in the outermost parts of meadows. With increasing current velocities, the shape of sea grass meadow becomes more ellipsoid, and tends to develop perpendicular to the water flow (Fonseca et al., 1983). The sea grass die-off in the center of the patches is probably caused by excessive sand accumulating inside the patches during storm events. The ring structure of Z. marina, the so-called 'fairy rings', appearing on the chalk plates in shallow water outside the calcium carbonate cliffs of the island of Møn, Denmark were well described by Borum et al. (2014). The authors found that neither the clonal growth pattern of this sea grass, sediment burial of shoots,

N. Mirlean, C.S.B. Costa / Estuarine, Coastal and Shelf Science 189 (2017) 104e114

hydrodynamic forcing nor nutrient limitation could explain the ring-shaped pattern. They conclude that the most likely explanation must be the accumulation of toxic sulfide in the sediment due to low iron availability in the carbonate-rich environment invaded by Z. marina shoots. Circular and ring-shaped patches were described to several species of the genus Spartina. For example, Spartina maritima colonization can alter intertidal physiography. Over time, sediment accretion, accelerated by colonizing S. maritima and this biogenic process, can rapidly transform unvegetated littoral flats into monotypic circular patches. Castellanos et al. (1994) described the process of ring patch formation in SW Spain saltmarshes as successional replacement between S. maritima and Arthrocnemum perenne, i.e. the central area of Spartina tussocks are invaded by A. perenne leaving, over time, only a fringe of Spartina around the edge of the tussocks. Areas invaded by A. perenne were characterized by oxidizing sediment, while the Spartina-dominated areas remained highly reducing, even in the surface layers. However, the authors did not identify specific inhibitory agents which caused plant die-off in the centers of Spartina patches. According to Lewis et al. (1990) Spartina argentinensis (sin. Spartina spartinae) is a dominant species of saline soils in the Great Chaco region, at Santa
Province (northern Argentina), and their dense tussocks with Fe round shapes prevent other plants settlement in saltmarshes by light interference. According to these authors, as the tussocks age, gaps like a “monk's tonsure” develop at their center and they are later colonized by Solidago chilensis and Neptunia pubescens. Lewis et al. (2001) found that the soil of S. argentinensis gap is richer in organic matter and phosphate and it has lower pH than that of the soil outside the gap. Alberti et al. (2008) pointed out that the colonization of central part of Sarcocornia perennis colonizing patches in NE Argentina saltmarshes by Spartina densiflora resulted in out-competition of this forb and probably not due to differences in desiccation or salt stress between evaluated health circular patches and depleted ring patches of S. perennis. Contrariwise, salt pan formation inside ring patches of S. perennis has been related to physical processes (water-logging, ice-scouring, sub-surface drainage, surface erosion and tidal wrack deposition) and bioturbation by activity of the burrow crab Neohelice granulata (Perillo and Iribarne, 2003; Minkoff et al., 2006; Escapa et al., 2015). N. granulata is widespread between southern Brazil and northeast Argentinean coast, being saltmarsh productivity and physicalchemical conditions influenced by its borrowing activities (Fanjul et al., 2007; Martinetto et al., 2016). Finally, Castillo et al. (2003) found a distinct central die-back phenomenon in S. densiflora tussocks with radius larger than 20 cm at SW Spain marshes and that areas remained occupied by high amount of plant debris. At southern Brazil, the estuary of Patos Lagoon is characterized by the development of saltmarshes largely occupied by two Spartina species (S. alterniflora and S. densiflora). The colonization patches of Spartina species on intertidal flats have a distinctive circular shape, but specially tussocks of S. densiflora. This circular patches are very common and the long-term monitoring (56 years) pointed to a steady lateral spread over mud flats for either species but at different spread rates (Costa and Marangoni, 2010; Marangoni and Costa, 2012). Organic matter and sediments accumulate inside the S. densiflora tussocks'crowns lifting the soils several centimeters above general ground level (Castillo et al., 2003; Costa and Marangoni, 2010). The central part of old ring structures is free of vegetation, or may have signs of incipient recolonization of S. densiflora and other types of vegetation. The hydrological regime of this microtidal estuary is controlled by the dominance of the river flow but particularly during rainy winter/ spring; sea water enters the estuary as a result of wind surge in a weakened river flow (Costa et al., 1988; Marangoni and Costa,

105

2012). The changing fresh and brackish water periods have a long-term character and specifically impacts on the development of vegetation and geochemical processes in sediments of Patos Lagoon estuary (Costa, 1997; Costa et al., 2003; Marangoni and Costa, 2012). The simultaneous presence of ring and circular patches of S. densiflora indicates the existence of non-biological specific factors causing the death of plants in the centers of the ring structures. Furthermore, similar occurrence of S. densiflora tussocks with and without central gaps was observed in the Argentinean Chaco region (Lewis et al., 2001) and SW Spain marshes (Castillo et al., 2003). We conducted the present study based on the assumption that these factors may be geochemical processes accompanying diagenetic changes in sediments of saltmarshes. The aim of the study was to compare the geochemical parameters of sediments in the rhizome horizon along transects across the ring and circular patches of S. densiflora in contrasting hydrological periods of the estuary, and to disclose possible factors of suppression or death of shoots inside its tussocks while they expand over the intertidal flat and alter the bathymetry. 2. Material and methods 2.1. Study area The study was carried out on an intertidal mudflat 15e25 cm below the mean water level of the Patos Lagoon estuary, which is
lvora Island (Rio Grande, Brazil, 32  010 S, 52  060 W). located in Po The island is situated in the center of the estuary, and at a distance of about 25 km from its mouth (Fig. 1). This site is characterized by a warm temperate climate and by a microtidal regime (<0.5 m) with an irregular flooding pattern driven primarily by winds and freshwater runoff from a 200 000 km2 watershed (Costa et al., 1988, 2003; Marangoni and Costa, 2012). The hydrologic pattern shows marked seasonal variation from high water levels and low salinities (0e5) during a rainy winter/spring to low water levels and high salinities (20e30) during summer/fall (Costa et al., 2003; Vaz et al., €ller et al., 2009). This hydrologic pattern can be disrupted 2006; Mo by inter-annual variability associated with the quasiperiodic El ~ o Southern Oscillation phenomenon. El Nin ~ o (warm phase) Nin events in the tropical Pacific promote excessive rainfall and a high discharge of rivers in southern Brazil during the austral spring of the event year and summerefall of the year following the start of El ~ o (Vaz et al., 2006; Mo €ller et al., 2009; Marangoni and Costa, Nin 2012). 2.2. Sampling The sampling was carried out in two periods: i) April 2015 corresponding to the maximum water salinity in the estuary and ii) in November of the same year with high standing of fresh water under a El Nino event. Circular and ring patches of Spartina species
lvora Island have been extensively identified and mapped in on Po
lis et al., aerial photos through GIS imaging processing tools (Le 2001; Costa and Marangoni, 2010; Marangoni and Costa, 2012). We studied two S. densiflora patches: circular (D~7 m; hereinafter called “CP”) and ring (D~20 m; “RP”) located approximately 70 meters from each other (Fig. 1). RP and CP were located 45 m and 95 m respectively from the east margin of the island. The RP central gap showed an incipient re-colonization by Scirpus maritimus (<5 tillers m
2). In the last 12 years CP diameter increased 75% over the mud flat, whereas RP diameter keep practically constant surrounded by S. alterniflora patches and its central gap developed after 2008. These two patches were chosen as representative of the observed vegetative propagation pattern of S. densiflora at local

106

N. Mirlean, C.S.B. Costa / Estuarine, Coastal and Shelf Science 189 (2017) 104e114

marshes. The samples of surface sediment were selected with irregular interval steps along the profiles passing through the centers of both structures, the first and last samples of the profiles were selected beyond the outer edge of the structures. When sampling was carried out on RP the thickening of sampling points was done and the samples were taken from the center of its circular vegetated band, as well as from internal and external parts of the band. All sediment samples were collected using PVC tubes (11 cm diameter and 15 cm length) by pressing the tube in sediment surface. A total of 22 and 30 sediment samples were collected on CP and RP, respectively. The tubes were hermetically sealed by PVC caps under pressure and transported in a vertical position to the laboratory, where they were stored at 4C  . Extraction of interstitial water was realized by centrifuging of sediment subsamples from 0 to 8 cm interval at 3000 rpm for 30 min in 50 ml polypropylene tubes which were purged with nitrogen for 1 min and hermetically sealed before centrifugation. Topographic profiles were carried out transversely to the two Spartina patches and height measurements were made with the help of a laser leveler FG-L3 FPM Holding GmbH. Surface water salinity (using an Orion 3-Star conductivity meter, Thermo Scientific™) and water level (to the nearest centimeter on a graduated ruler) were determined daily, between 10:00 and 11:00 h, at a fixed station located 2 km north of the study site, and related to the mean water level (MWL) of the estuary between 2001 and 2015. Simultaneous measurements of the water level at the fixed station and the study site were made, allowing for an estimation relative height difference and of flooding frequency (percentage of time flooded) at S. densiflora patches over the 60 days before each sampling period. In the brackish water period, the number of active N. granulata burrows in five 1.0 m2 quadrats uniformly placed along each transect in the S. densiflora patches was quantified as a proxy of crab density (Alberti et al., 2007). The density of crab burrows ranged between 20 and 31 m
2, and the averages in CP and RP were 22.8 ± 2.3 and 28.5 ± 2.8 m-2 (±standard deviation), respectively. RP had a significantly (t-test ¼ 3.52, p ¼ 0.008, df ¼ 8) higher density of crab burrows than CP. Crab burrows were not quantified at fresh water period sampling but very few active individuals were observed under the high standing fresh water. 2.3. Analytical

Fig. 1. Study area and sampling site. Profiles across the S. densiflora vegetation structures are shown (red lines): RP e ring patch structure (with plant die-back at its central part), CP e circular patch structure (densely vegetated). Photographs source: Google Earth. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

The reduction/oxidation potential (Eh) was measured at room temperature in subsurface of sediment core samples by pushing down to the depth of 3e4 cm the naked Pt-electrode Analion®. The Pt electrode was verified in standard solutions using the Ag/AgCl/ KCl reference electrode. Electrical conductivity and pH were immediately measured in one the subsamples of interstitial water. The pH was measured with a Oakton® apparatus previously calibrated with a 0.1 M KCl solution and the pH was measured with a Oakton® pH-meter, using two buffer solutions of pH 4 and pH 7. The interstitial water from the second subsample was filtered through a 0.22 mm Millipore® membrane and divided in two portions, one of which was acidified. These portions of interstitial water samples were stored in hermetically sealed bottles at 4 C , for posterior analysis of anions and cations using Ion Chromatography Metrohm® apparatus, and Corg by apparatus TOC-L Shimadzu®. The Relative Standard Deviation (RSD%) for triplicate analyses of anions and cations ranged from 0.5 to 7.5% The detection limit for Corg was 4 mg/L. Free sulfide (H2S) analysis in sediments was realized using an Orion™ advanced portable ISE/pH/mV/ORP/temperature meter model 290A with a Model 9616 BNC Ionplus Silver/Sulfide electrode (Wang and Chapman, 1999). The meter has a concentration range of

N. Mirlean, C.S.B. Costa / Estuarine, Coastal and Shelf Science 189 (2017) 104e114

0.0e19.9 M (S2-) and a relative accuracy of ±0.5% of the reading. A basic sulfide antioxidant buffer solution (SAOB) and sulfide standards (Na2S$9H2O) were made up in the morning and at mid-day on each sampling day. The S2- electrode was calibrated before and after each batch of not more than 12 samples. Three S2- standards (10, 100, 1000 and 10 000 M) were used for a four point electrode calibration. The RSD% for triplicate analysis was below 5%. The content of acid volatile sulfides (AVS) in sediments was determined using the standard method (USEPA, 1991). The chromium-reducible sulfides (CRS) were determined in wet sediment samples using the two-step distillation technique described by Fossing and Jørgensen (1989), using zinc acetate to precipitate sulfides. The amounts of AVS and CRS were assessed by iodometric titration procedure (Burton et al., 2008). 3. Results In the present study, we compared the geochemical parameters in RP and CP in two polar environments seasonally defined in the estuary by predominance of river discharges and sea water incursions. Sampling was conducted in the freshwater environment, chronologically later than the sampling within the brackish period in the estuary. However, we prefer to present the obtained data in reverse order that we believe will contribute to a better justification of the proposed scheme - the formation of ring patches of S. densiflora. 3.1. Geochemical parameters of sediments in freshwater period During all 60 days before sampling both S. densiflora patches remained flooded and surface water salinity average was 0.1 mS cm
1 (Fig. 2a). On the sampling day, the electrical conductivity of pore water in the upper horizon of sediment (0e8 cm layer) of CP varied within a small range (1.8e2.2 mS cm
1) which approximately corresponds to the content of soluble salts about 1 g L
1 (Fig. 3a). This value of pore water salinity is 3e5 times higher than the salinity of surface € ller et al., 2009; freshwater in Patos Lagoon and its tributaries (Mo Marangoni and Costa, 2012). That, probably, is the influence of a weak salt incursion one week before the sampling (Fig. 2), which is also marked by high salinity of pore water (about 3 g L
1) in the soil lower layer (8e15 cm; data not presented). The pH values in pore water varied within a narrow range (6.16e6.74; average ¼ 6.61). Sediments in the outer border zone (first and last sample points on the profile) presented lower pH values (Fig. 3a). These values are closer to the average value of pH (6.7e7.0) in the surface waters of the estuary catchment area (Coradi et al., 2009). The Naþ/SO24 ratio in the pore water along the CP profile changed by more than 4-fold (Fig. 3a). The minimal values of Naþ/ SO24 were recorded in sediments at the outer borders of CP and were close to the ratio (20e30) characteristic to the surface water in humid subtropical landscapes which occupy most of the catchment area of Patos Lagoon (Perelman and Kasimov, 1999). At the central part of CP, pore water sulfate concentration was greatly reduced and Naþ/SO24 reached a value close to 100. The decrease of sulfate ions concentration in the pore water from the CP central part, densely vegetated by Spartina, has probably occurred due to its microbiologically induced reduction. The pore waters in CP sediments contained large amounts of dissolved organic carbon (DOC), which reached at the central part as much as 90 mg L
1 and dropped towards the outer borders to values as low as 20 mg L
1; figure similar the average content of DOC in phreatic and surface waters of neighboring coastal landscapes. Along the entire CP profile, sediments of the upper horizon were characterized by negative Redox values varying from
65 mV

107

to
199 mV. Most strongly reduced conditions were met in the central part, while in the outer borders' sediment the Redox potential was notably higher (Fig. 3b). The concentration of free dissolved sulfide in the sediment of CP central point was as much as 27 mM L
1. Outside CP, soluble sulfide concentration in sediments was significantly lower. The other sulfide species, but specially CRS, also showed distribution patterns along the profile, opposite to what was observed for Redox, with AVS (55 mg kg
1) and CRS (up to 500 mg kg
1) values, reaching peaks at the central part of CP and decreasing towards the border. The pH in RP pore water was almost the same along the profile of this S. densiflora structure (Fig. 3c) but RP sediments had higher electrical conductivity (average 3.7 mS cm
1) than CP pore water. Perhaps this is due to the difference in the positions of structures on the island: CP closer to the island margin than RP, where, due to more intense water exchange, the desalination of sediment occurs faster. The variation of sodium-sulfate ions ratio (10e117) and DOC (20e100 mg L
1) in the pore water of the RP profile showed slightly large variation but very different patterns in comparison with CP. Both parameters in RP profile were characterized by two peaks in the bordering vegetated areas (Fig. 3c). Changes in sodium-sulfate ions ratio occurred associated with reduced content of sulfate ion in the vegetated border (7 mg L
1) and a 10-fold higher average value at the unvegetated center of the ring. Likewise DOC was up to 5 times lower in RP center than in its border. In the sediments at the outer borders of RP the ions ratio and DOC reached about the same value as in its center (Fig. 3c). RP sediment showed strong reducing condition at the densely vegetated border and a large amplitude variation of Eh values from
50 mV to
365 mV (Fig. 3d). At its vegetated outer limits RP Redox potential roses 3e7 times in relation to its center and outer borders. Contents of water soluble sulfides including free hydrogen sulfide vary in pore water over the RP profile from 10 to 33 mM L
1. The higher content of free sulfide was observed at the densely vegetated border of the ring, in the same sampling points where minimal values of Redox potential were found (Fig. 3d). Additionally, highest concentrations of AVS and CRS on the RP profile were also found at the vegetated border, respectively, 82 and 811 mg kg
1. The comparison of the geochemical parameters of pore water in CP and RP sediments under high flooding frequency by freshwater revealed the following common characteristics: i) the development of strong reducing conditions in the areas occupied by vegetation; ii) depletion of pore water sulfate and enrichment of soluble organic matter equally occur in areas with the lowest Redox potential; iii) in the rhizosphere zone under the live vegetation different sulfide species accumulate, from hydrogen sulfide to pyrite and, possibly, elemental sulfur. 3.2. Geochemical parameters of sediments in brackish water period The sediments under the structures of CP and RP within brackish period were sampled about two months after the fresh water conditions changed to salted one, when the water level in the estuary reduced down to exposure of the surface of sediment in some parts of the saltmarsh. The averages of surface water salinity and flooding frequency on S. densiflora patches during the 7 days before sampling were 20 mS cm
1 (z13 g NaCl L
1) and 14% (RP)29% (CP) of the time, respectively, whereas during the 60 days before sampling were 10 mS cm
1 and 23% (RP)-48% (CP) of the time (Fig. 2a). On the sampling day of brackish period, the electrical conductivities of the pore water in the CP sediments were significantly higher and ranged along the profile from 33.9 to 47 mS cm
1, and the average value corresponded to salinity of about 20 g L
1. The

108

N. Mirlean, C.S.B. Costa / Estuarine, Coastal and Shelf Science 189 (2017) 104e114

(A) 160 140

25

brackish period

fresh water period

Water level (cm MWL)

100 80

MWL

MWL

Salinity

Salinity

15

60 10

40 20

Salinity (mS/cm)

20

120

5

0 -20

0

-40 60 57 54 51 48 45 42 39 36 33 30 27 24 21 18 15 12 9 Dayes before sampling

6

3

60 57 54 51 48 45 42 39 36 33 30 27 24 21 18 15 12 9 Dayes before sampling

0

6

3

0

(B) 0

RP

CP

Water level (cm MVL)

-5

-10

-15

-20

Average RP height = -9±2 cm MVL

Average CP height = -14±3 cm MVL

-25 0

2

4

6

8

10

12

14

16

18

20

22

24

26

m

0

1

2

3

4

5

6

7

8

9

10

11

m

Fig. 2. (A) Daily variation in water level and salinity of Patos estuary along 60 days before the collection of sediment samples for the periods of freshwater (spring 2015) and brackish water conditions (autumn 2015). (B) Height of the sediment surface across S. densiflora vegetation structures. RP e ring patch structure (with plant die-back at its central part), CP e circular patch structure (densely vegetated). Height expressed as cm MWL (mean water level of the estuary between 2001 and 2015). The S. densiflora vegetated segments are marked as green bar below the x-axis. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

ratio of sodium and sulfate ions in the pore water of the profile fluctuated within a small range from 4.2 to 3.9, with an average value of 4.0. This value corresponded to diluted seawater, which usually occurs at this part of Patos Lagoon estuary (Costa et al., 2003; Marangoni and Costa, 2012). Fluctuations of both parameters of the pore water salinity did not exhibit any consistent patterns (Fig. 4a). The pore water of CP structure in brackish period showed lower pH (from 5.7 to 6.3; mean ¼ 5.9) than that of the normal estuarine water with similar salt content and even less than the pore water of the same structure during the freshwater period. The ionic composition did not show any particular distinctive distribution pattern in this sampling period. The DOC content in the CP pore water under brackish conditions significantly decreased in comparison with the freshwater period. DOC did not exceed 30 mg L
1 (mean ¼ 17 mg L
1). The maximum values of DOC in the profile were found in the peripheral areas of the СР structure and its outer borders. Minimal DOC values were observed within the circle (Fig. 4a). During the less flooded-brackish water condition there was a strong oxidation of the marsh sediments; redox potential drastically increased in CP from negative to þ140 mV. The Eh distribution along the profile showed a maximum value at the center and minimum value (three times smaller) in the periphery of CP (Fig. 4b).

If compared to freshwater conditions, the amount of soluble sulfide in the brackish water environment was strongly reduced and, in 30% of the sediment samples from CP, soluble sulfide was not detected. The average free sulfide content was as low as 2.3 mM L
1. There was no pattern of sulfide concentration along the CP profile (Fig. 4b). AVS and CRS contents have also decreased significantly; approximately 4 and 7 times lower than those in freshwater conditions, respectively. At the bordering vegetated areas of CP, both AVS and CRS concentrations were much lower than those in the center (Fig. 4b). During brackish water condition, the electrical conductivity of pore water in the RP was almost the same as in CP, but DOC average was about 2.5-folds lower than that of the freshwater period. These two parameters did not show any distinctive pattern along the RP profile (Fig. 4c). Pore water pH inside RP exhibited unexpectedly low values. The average pH ¼ 5 was well below local freshwater bodies and much lower than the pH of estuarine water, which at that time was about 8. The pH distribution along the profile demonstrated a sharp decrease under the vegetated bordering area of the S. densiflora ring in comparison with its center and outer borders (Fig. 4c). Five of six sediment samples collected inside RP had pore water pH values well below 5, and one of them even considerably lower than 4 (pH ¼ 3.4). The average ratio of sodium and sulfate ions in the pore water of the RP profile was close to seawater (3.9) and followed a

N. Mirlean, C.S.B. Costa / Estuarine, Coastal and Shelf Science 189 (2017) 104e114

(A)

550

7

100

6 80

5

DOC

60

+

4

pH

3

Na/SO24

40

Cond.

2

20

pH; Cond. (mS cm-1)

Na+/SO42-; DOC (mg l-1)

(B)

8

1

0

Eh (mV); S2- (μM/L); AVS,CRS (mg/kg)

120

1

2

3

4

5

6

7

8

450 350

Eh 2S AVS CRS

250 150 50 -50 -150 -250

0 0

109

0

9

1

2

3

4

(C)

Na+/SO42-; DOC (mg l-1)

9

+

Na /SO

24

8

pH 80

Cond.

7 6

60

5

40

4 20

3 2

0 0

2

4

6

8

10

12

14

16

18

20

22

24

Eh (mV); S2-(uM/l); AVS, CRS (mg/kg)

DOC 100

6

7

8

9

(D)

10

pH; Cond. (mS cm-1)

120

5

m

m

800

Eh 2S AVS CRS

600 400 200 0 -200 -400 0

2

m

4

6

8

10

12

14

16

18

20

22

24

m

Fig. 3. Spatial distribution of average values for several geochemical parameters of the interstitial water from the upper layer (0e8 cm) of sediments of S. densiflora vegetation structures at the freshwater period (spring 2015). Data of circular patch (A and B) and a ring patch (C and D) are shown. The S. densiflora vegetated segments are marked as green bar below the x-axis. Legend: DОС, рН, Naþ/SO42-, Cond. ¼ conductivity; S2- ¼ sum of free sulfide; AVS ¼ acid-volatile sulfides; CRS ¼ chromium-reducible sulfides. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

pH similar distribution pattern (Fig. 4c); the pore water sample with the lowest pH contained a relatively large quantity of sulfateion. During the brackish water period, the pore water Eh in RP was significantly higher than in CP, and ranged from þ184 to þ300 mV. Redox potential increased from the ring center towards the more densely vegetated bordering area of RP (Fig. 4d). RP sediments showed large amounts of soluble sulfide (up to 139 mM L
1). The highest values of SS2- were found in the inner parts of vegetated bordering areas, while in the central part and its outer borders sulfide concentrations were much lower (Fig. 4d). AVS average was significantly lower than in CP and in half of sediment sample values were below the detection limit of the method. In the brackish condition, RP average content of CRS was about 3 times lower than that observed in CP, as well as it was 13 times lower than that of freshwater condition. The CRS distribution along the RP profile demonstrated a sharp concentration decrease (down to 8 mg kg
1) under the vegetated bordering areas, at the same points where the maximum of free water-soluble sulfide, almost absent of AVS were determined (Fig. 4d). The comparison of geochemical parameters inside CP and RP within less flooded-brackish condition led to the following generalizations: i) a change from strong reducing to oxidative ambient occurs in both structures; ii) a strong acidification of pore water happens, which can reach pH < 4 inside RP; iii) a significant change in the sulfide speciation occurs in pore water within the live

rhizosphere of the vegetated part of the ring structure - while the amount of total reduced sulfur decreases a marked increase of the water-soluble sulfide and lowering of pH are registered. 4. Discussion Combining our results and already published information, we are able to construct a conceptual model showing RP formation of S. densiflora in the irregularly flooded marshes of southern Brazil (Fig. 5). Summarizing, an active process of sulfate reduction accompanied by a strong decrease of the Redox potential occurs in marsh sediments during the high water level and low salinity period. However, at that time, both S. densiflora structures (CP and RP) showed low content of toxic free sulfide in their sediments, which was effectively removed from the system by binding to iron in the form of AVS and CRS. Contrariwise, during the brackish water period, oxidized condition prevail due to the low water levels, and very low pH values (<4) were established at inner edges of RP due to higher oxidation of solid sulfide phases, which was not observed in CP located lower in the intertidal. Thus, aeration and acidification of sediments of the uplifted S. densiflora structures during salt water incursion in the estuary may explain the die-back pattern of RP. Although acidification of aerated marsh sediments have been recorded before, as far as we are aware this is the first time that this process has been related to the die-back of central part of expanding patches of saltmarsh plants. More detailed explanations

N. Mirlean, C.S.B. Costa / Estuarine, Coastal and Shelf Science 189 (2017) 104e114

150

DOC Cond. 2+ Na /SO4 pH

125

25

100 7 6

20 15

5

Eh (mV)

30

80

Eh S2AVS CRS

8

pH; Na+/SO42-

35

100 90

9

40

70 60 50

75

40 50

30

10 5

4

25

3

0

20 10

0 0

1

2

3

4

5

6

7

8

9

10

11

(C)

0

8 30

7

DOC Cond. pH + 2Na /SO4

20

6 5

10 4 0

3 8

10

12

14

m

16

18

20

22

24

Eh (mV); CRS (mg/kg); AVS (mg/kg)

40

6

2

3

4

5

6

7

8

9

10

11

m

(D) 9

4

1

300

10

50

2

0

m

pH; Na +/SO42-

DOC (mg l-1); Cond. (mS cm-1)

45

DOC (mg l-1); Cond. (mS cm-1)

(B)

10

S 2-(uM/L); AVS (mg/kg); CRS (mg/kg)

(A)

50

50 45

250

40

Eh 2S AVS CRS

200

35 30

150

25 20

100

S 2-(μM/L)

110

15 10

50

5 0

0 2

4

6

8

10

12

14

16

18

20

22

24

m

Fig. 4. Spatial distribution of average values for several geochemical parameters of the interstitial water from the upper layer (0e8 cm) of sediments of S. densiflora vegetation structures at the brackish water period (autumn 2015). Data of circular patch (A and B) and a ring patch (C and D) are shown. The S. densiflora vegetated segments are marked as green bar below the x-axis. Legend: DОС, рН, Naþ/SO42-, Cond. ¼ conductivity; S2- ¼ sum of free sulfide; AVS ¼ acid-volatile sulfides; CRS ¼ chromium-reducible sulfides. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 5. Conceptual model for the formation of ring patches of S. densiflora. in the irregularly flooded saltmarshes of southern Brazil. The relative magnitude of the geochemical parameters of sediments is indicated by arrows dimensions and directions.

N. Mirlean, C.S.B. Costa / Estuarine, Coastal and Shelf Science 189 (2017) 104e114

on the mechanisms that regulate RP formation in southern Brazil saltmarshes are given in the following paragraphs.

4.1. Free sulfide binding and solid sulfides formation in freshwater flooded marsh In the Patos Lagoon estuary the dissolved H2S was frequently low in pore waters of S. densiflora structures, an expected scenario when there is a reactive-Fe which scavenges the sulfide from solution (Panutrakul et al., 2001). Inside CP and RP, the fraction of pyrite (CRS; conservative fraction FeS2) that depended on the site and salinity composes 50e95% of total reduced sulfur and always prevails over the AVS fraction (FeS, Fe3S4). These results agree with Howarth and Merkel (1984) that demonstrated in saltmarshes even for short term incubations (a few hours) the majority of the reduced 35 2S transited in the pyrite extracted fraction and not to AVS. Similar process of sulfide reaction with Fe2þ to form pyrite preventing the build-up of toxic H2S was described for flooded wetland soils on the coastal plain of North Carolina subjected to saltwater incursion (Hopfensperger et al., 2014). An increase of the content of solid phase fraction of reduced sulfur in sediments occurred during the period of prolonged freshwater flooding of the marsh (Fig. 5). The process of sulfate reduction was accompanied by a strong decrease of Redox potential in sediments, which distribution demonstrates that the most strongly bacterial activity is developed in areas densely occupied by live tillers of S. densiflora. Previously, Costa et al. (2003) showed lowering in the sediment Eh values throughout Polvora Island marsh at high-water (winter-spring) period. Another parameter which confirms a strong sulfate reduction process is the local change in pore water composition. During the freshwater period, in the areas with high content of AVS and CRS factions inside our structures we observed a sharp (up to 10 times or more) increase of Naþ/SO24 , i.e., reduction of the sulfate concentration and confirmation of the active process of sulfate reduction at the depth of Spartina's rhizosphere. The depletion or even disappearance of sulfate from the interstitial waters in sulfide rich sediment is known since the first marine science expeditions (Murray and Irving, 1895). Metabolism byproducts of the sulfate reducing bacterial populations are greatly responsible for the geochemical circumstances in the studied S. densiflora structures. Bacterial growth is fostered by organic matter consumption and DOC concentration in pore water showed maximum values (up to 800 mg L
1) in freshwater environment and minimal (down to 15 mg L
1) in brackish conditions. At Patos Lagoon estuary, the freshwater environment predominates during the rainy winter/spring (Marangoni and Costa, 2012) when S. densiflora shows minimal growth and a greater income of its detritus into the sediment occurs (Costa, 1997; Peixoto and Costa, 2004). Therefore, during this period, the peaks of DOC concentration were observed just in the central portion of CP or under the vegetated border of RP, while the central part of RP free of live S. densiflora plants the DOC concentrations were always small. Besides, secretions of living plants such as sediment cyanobacteria could be a source of DOC in saltmarsh pore water (Lyons et al., 1982), the leaching and microbial decay of S. densiflora remains seem to be responsible for most of the high DOC content observed during the freshwater period. Additionally the high discharge of rivers in southern Brazil during the austral spring (Marangoni and Costa, 2012) contributes to the enrichment of pore water with DOC concentration in freshwater period. In any case, larger DOC and iron availability and reducing condition during freshwater flooding could promote abundant CRS (pyrite) formation inside S. densiflora structures.

111

4.2. Sediment acidification and RP formation induced by aerated brackish condition The changing of fresh water by salty water in pore solutions significantly acidifies the sediments of both S. densiflora structures, but the pH in RP was much lower and at one point in the inner edge of the ring it decreased under 4 (Fig. 5). Contrasting with its negative values of the freshwater period, the Redox potential in brackish condition was positive and its absolute value was 2-folds higher in RP than that of CP. Higher density of N. granulata burrows in RP than in CP might contributed to differences in soil Redox potential (Fanjul et al., 2007; Martinetto et al., 2016), however the observed average burrow densities (28.5 and 22.8 burrows m
2, respectively) were among the mid-bottom values recorded for N. granulata in Spartina dominated marshes of Patos Lagoon estuary and other coastal areas of the southwestern Atlantic (Costa et al., 2003; Alberti et al., 2007; Freitas et al., 2016; Martinetto et al., 2016). Furthers studies are necessary to understand the role of crab burrowing in soil Redox potential of irregularly flooded Spartina marshes. The literature indicates a large variation in pH of pore waters in saltmarshes (Costa et al., 2003; Marques et al., 2011; Negrin et al., 2011; Duarte et al., 2013). However, the most frequently reported range of values in the estuaries with daily tidal changes of water is 6.2e8.2, which corresponds approximately to riverine waters and incursion of marine waters. Acid or weakly acidic porous waters are also reported in the literature (Otero and Macias, 2002). In their study of physical stress gradient along
lvora Island marsh, Costa et al. (2003) found at a low water Po autumn period significantly lower pH values (down to 5.4) in organic-rich sediments of S. densiflora. More decisively Liu et al. (2008) explained the presence of acid solutions in saltmarsh sediments as the result of iron sulfides and amines oxidation and the formation of sulfuric and nitric acids during the dry season, which seems the likely source of the elevated protons concentration in pore solutions. The pH decrease inside S. densiflora structures in aerated brackish condition is clearly caused by oxidation of solid sulfide phases, as it is indicated by a significant decrease of AVS and CRS concentrations in sediments. Herein at 40% of the sampling points inside RP, AVS were oxidized to undetectable amounts, and the curve of CRS distribution roughly followed that of pH. Schoepfer et al. (2014) also observed that both AVS and CRS decreased as salt water incursion events intensified at a coastal wetland, and associate these processes with either sediment oxidation or consumption as electron donors. Homologous acidification processes have been described by other researchers. McKee et al. (2004) showed that marshes in Lousiana (U.S.A.) subject to extensive die-back had their sediments acidified upon oxidation, while healthy marshes did not, and they pointed out a greater pyrite concentrations in the die-back marshes as the main cause of the acidification (pH down to 4). Major soil acidification occurred in east Australia in estuarine sediments of coastal floodplains when, due to drainage and excavation to facilitate a range of human activities, acid sulfate soils are formed when iron pyrite oxidizes to sulfuric acid exposed to atmosphere, lowering soil pH below 4 (as low as pH 2.7) (Sammut et al., 1995). Very high SS2- concentration (up to 135 mM L
1) and very low pH (4.1 and 3.4) in the internal border of RP sediments reflect the reaction of appeared sulfuric acid with AVS in pore waters, which promotes the formation of free hydrogen sulfide (Rickard and Morse, 2005). The maximum sulfide concentration found in RP may not be lethal, since species of Spartina genus is also known to survive in high concentrations of sulfide, up to 8000 mM L
1, but concentrations between 1000 and 3000 mM L
1 may impair their growth (Lamers et al., 2013). However, acidify pore water can be

112

N. Mirlean, C.S.B. Costa / Estuarine, Coastal and Shelf Science 189 (2017) 104e114

directly toxic for plants, and solubility of Al3þ, Fe2þ, Mn2þ and other metals increases several times with decreases in pH, and may reach toxic levels for plants (Sammut et al., 1995; McKee et al., 2004; Lamers et al., 2013; Hopfensperger et al., 2014). Thus, very low pH such as those found inside RP of S. densiflora can directly or indirectly impair and even kill saltmarsh plants. The difference in pore water chemistry between S. densiflora structures can be explained by their distinct position at the intertidal height. The situation with extremely acid pore water was registered in the period of low water levels in the lagoon when RP sediments were exposed and directly contacted with the atmosphere. At that period, CP sediments were flooded 48% of the time with an average water layer cover of 8 cm, which obviously contributed to the lower level of sulfide oxidation and pH did decrease below 5.7. The CP was located in a lower region of the marsh, while RP elevated height was reached only by the microtides 23% of the time. The hypsometric difference between the centers of the structures was only 5 cm, however it was enough for the appearance of very contrasting geochemical situations between them, during a critical lowering of water level in the estuary. Some authors suggested that events of acute marsh die-back of large extensions of Spartina dominated marshes of the Gulf and east coast of the U.S.A. could be caused by acidification and/or toxic compounds accumulation due to marsh surface prolonged exposition, and hydrology or associated factors with ameliorated effects along shorelines where plants were frequently inundated by tides (McKee et al., 2004; Hopfensperger et al., 2014). By contrast, Hughes et al. (2012) did not find Redox and pH changes as the cause of acute die-back in a saltmarsh inundated twice daily at South Carolina. Thus, marsh die-back associated with sediment acidification induced by aerated brackish-salt condition seems to be found in irregularly flooded marshes. The distinctive geochemical conditions of S. densiflora CP and RP patches seems to support this conclusion, and acidification does not seem to be important only to extensive episodic die-back events but it might explain successional die-back and gap formation in tussocks of bioengineer marsh plants, as reported by several researchers (Lewis et al., 1990, 2001; Castellanos et al., 1994; Castillo et al., 2003). Our present findings highlight a formation process different than it is usually used to explain vegetation ring patches or “historical die-back” of saltmarshes, which is caused by excessive water logging of organic rich sediments that limits aeration of plant roots and allows buildup of soil phytotoxins such as sulfide (McKee et al., 2004). It occurs gradually and typically requires years for a marsh to die completely. Similar historical pattern has been described for submerged vegetated patches of the seagrass Zostera marina. The die-off in the central part of Z. marina patches was associated with excessive sand accumulating inside the patches during storm events (Fonseca et al., 1983; Fonseca and Kenworthy, 1987) and toxic sulfide accumulation in the sediment of the inner side of patches (Borum et al., 2014), which, in the ongoing process of vegetation expansion, the colonized circular patch transforms into a ring. This second explanation was supported by the low iron availability in the carbonate-rich environment colonized by Z. marina. Like seagrass rings, in the clonal tussock of S. densiflora with gradual accumulation of organic matter in the sediment, sulfate reduction and sulfide accumulation, unfavorable conditions for plant life, may occur. In contrast to the conditions of carbonaterich environment, where mobility of iron is very reduced due to higher pH, in our marsh system the resulting hydrogen sulfide is effectively removed from the system by binding to iron in the form of AVS and CRS. Different from frequently flooded marshes of the east coast of North America, solid-phase sulfides formed in subaerial conditions due to the iron rich environment are periodically oxidized by atmospheric oxygen, during seasonal low water level in

the Patos Lagoon estuary, producing sulfuric acid and making soluble large quantity of hydrogen sulfide into pore water. Particularly at intertidal heights just below the mean water level, the sediments appear strongly acidified and enriched with hydrogen sulfide and probably other toxic compounds. The die-back of S. densiflora tillers obviously starts at the organically enriched sediment in the center of its circular patches, triggered by periodic increase of sulfides oxidation combined with a sharp decrease of pH in the rhizomes layer during prolonged surface exposition atmosphere oxygen. Healthy tillers in the outer edge of the patch carry on spreading radially into adjacent mud flat with non-toxic sediment while the vegetation in the inner part dies, thus a ring is formed and its diameter increases. The toxic environment for plant roots appears as a result of the Redox potential increases and solid sulfides oxidation. The emergence of the oxidizing conditions is produced by the uplift of sediment by this bioengineer grass and/or periodic lowering of the water surface below a certain critical level. If these factors (sediment uplift and/or water level dropping) will progress we can expect in the future the development of a ring structure from an actual circular structure. RP may even be replaced by another plant community that will be established in the bare mud flat created by this process. The dissolution of pyrite and sediment acidification after intense subsurface oxidation of iron rich and salinized marsh sediment may have much more important and widespread ecological importance than it is recognized nowadays. The displacement of marsh pioneers that promote sediment uplift by other plants has been attributed to higher competitive ability but acidification/free sulfide stress could be an underlying process shifting the capacity of the pioneer withstands competition. For example, the presence and response of arbuscular mycorrhizal fungi (AMF) to toxicity of sulfide and or acid-sulfate sediment could induce pioneer die-off and later replacement by other species. Lamers et al. (2013) pointed out that sulfides are known to seriously decrease the vitality of ectomycorrhizae and sulfide toxicity on AMF may impair plant nutrient uptake. On the other hand, Maki et al. (2008) showed that AMF isolated from the field significantly promoted the growth of pioneer grasses and legume shrubs in acid sulfate soil at pH 3.4, and thus, acid-tolerant AMF may play an important role in the establishment of plants on this kind of soils. Similar ecological processes could be happening in the marshes. Daleo et al. (2008) showed that AMF can affect the competitive ability of saltmarsh plants. Indeed, at the SW Atlantic coast of South America, the root colonization by AMF increased S. densiflora growth at low nutrient levels, and fungicide additions resulted in S. alterniflora migrating to higher marsh elevations, displacing S. densiflora. Thus, S. densiflora AMF sensitivity to acidification/free sulfide stress might be associated with RP formation. Furthermore, the decrease of plant growth inside the central upraised part of colonizing patches of bioengineer plants before replacement by competitor have been reported for saltmarshes in coastal lagoons of SE Spain (S. maritima, Castellanos et al., 1994) and Argentina (S. perennis, Alberti et al., 2008; S. argentinensis, Lewis et al., 2001). Besides the records of high Redox potential of uplifted patches, these studies lack detailed geochemical analysis of sediments, but it is possible to speculate that acidification may have also occurred and contributed to the emergence of toxic conditions for pioneer plants and/or their AMF in these metal rich saline environments. 5. Conclusion The formation of ring structures (RP) of S. densiflora on saltmarshes in Patos Lagoon estuary is the result of periodic changes of physico-chemical parameters and the composition of pore water in the surface sediments of the rhizosphere zone. Under periods of

N. Mirlean, C.S.B. Costa / Estuarine, Coastal and Shelf Science 189 (2017) 104e114

freshwater flooding of the sediments, S. densiflora vegetation patch accumulates significant amounts of AVS and CRS, and is characterized by sharply reducing environment. During this period, the pH of pore water corresponds to that of the surface water and the amount of soluble forms of sulfides is very low, or is not detectable. This geochemical conditions in the sediments, most likely, does not prevent the development of S. densiflora patches. The seasonal water changes from fresh to brackish conditions and simultaneous lowering of the water level down to the exposure of the sediment. Mostly AVS and some portion of CRS oxidize and produce sulfuric acid that causes a dramatic lowering of the pore water pH to below 4. The reaction of acid with sulfides released hydrogen sulfide in the pore solution, which, coupled with a very low pH had a detrimental effect on S. densiflora. Formation of ring structures begins in the uplift center of the circular structure where over time accumulated organic matter produced by the plant and solid sulfides, and therefore toxic components, are produced in larger quantities. The die-back of S. densiflora in the central part and the lateral expansion of the tussock at its perimeter increase over time the inner radius of the ring. In a healthy circular structure, a lower positioning at the intertidal height allows the presence of a layer of water over the sediment surface. Then it is subjected to a moderate oxidation of sulfides and physical-chemical parameters of pore water apparently are non-toxic to plants, which retains their growth vigor. Development of S. densiflora ring structures in the intertidal flats probably indicates a bioengineer lifting of sediment surface and/or change in the hydrological regime of this irregularly flooded estuary that results in lowering of the average water level. Acknowledgements The research was funded by CNPq-Brazilian National Research Council through grant 303848/2013-8. References Alberti, J., Daleo, P., Iribarne, O., Silliman, B.R., Bertness, M., 2007. Local and geographic variation in grazing intensity by herbivorous crabs in SW Atlantic salt marshes. Mar. Ecol. Prog. Ser. 349, 235e243. Alberti, J., Escapa, M., Iribarne, O., Silliman, B., Bertness, M., 2008. Crab herbivory regulates plant facilitative and competitive processes in Argentinean marshes. Ecology 89, 155e164. Borum, J., Raun, A.L., Hasler-Sheetal, H., Pedersen, M.Ø., Pedersen, O., Holmer, M., 2014. Eelgrass fairy rings: sulfide as inhibiting agent. Mar. Biol. 161, 351e358. Burton, E.D., Sullivan, L.A., Bush, R.T., Johnston, S.G., Keene, A.F., 2008. A simple and inexpensive chromium-reducible sulfur method for acid-sulfate soils. Appl. Geochem. 23, 2759e2766. Castellanos, E.M., Figueroa, M.E., Davy, A.J., 1994. Nucleation and facilitation in saltmarsh succession: interactions between Spartina maritima and Arthrocnemum perenne. J. Ecol. 82, 239e248. Castillo, J.M., Figueroa, M.E., Palomo, T.L., Rubio-Casal, A.E., Nieva, F.J.J., 2003. Intratussock tiller distribution and biomass of Spartina densiflora Brongn. Lagascalia 23, 61e73. Costa, C.S.B., 1997. Tidal marsh and wetland plants. In Subtropical convergence environments. In: Seeliger, U., Odebrecht, C., Castello, J.P. (Eds.), The Coast and Sea in the Southwestern Atlantic. Springer, Berlin, pp. 24e26. Costa, C.S.B., Marangoni, J.C., 2010. As comunidades das marismas. In: Seeliger, U.,
culo de transOdebrecht, C. (Eds.), O estu
ario da Lagoa dos Patos: um se ~es. Editora da FURG, Rio Grande, pp. 125e133 (In Portuguese). formaço Costa, C.S.B., Marangoni, J.C., Azevedo, A.M.G., 2003. Plant zonation in irregularly flooded salt marshes: relative importance of stress tolerance and biological interactions. J. Ecol. 91, 951e965. Costa, C.S.B., Seeliger, U., Kinas, P.G., 1988. The effect of wind velocity and direction ^ncia Cult. 40, on the salinity regime in the Lower Patos Lagoon estuary. Cie 909e912. ~o da qualidade da a
gua suCoradi, P.C., Fia, R., Pereira-Ramirez, O., 2009. Avaliaça perficial dos cursos de
agua do município de Pelotas - RS, Brasil. Ambiente

4, 46e56 (in Portuguese). Agua, Taubate Daleo, P., Alberti, J., Canepuccia, A., Escapa, M., Fanjul, E., Silliman, B.R., Bertness, M.D., Iribarne, O., 2008. Mycorrhizal fungi determine salt-marsh plant zonation depending on nutrient supply. J. Ecol. 96, 431e437. Dennis, B., Civille, J.C., Strong, D.R., 2011. Lateral spread of invasive Spartina

113

alterniflora in uncrowded environments. Biol. Invasions 13, 401e411. Duarte, B., Couto, T., Freitas, J., Valentim, J., Silva, H., Marques, J.C., Dias, J.M., Caçador, I., 2013. Abiotic modulation of Spartina maritima photobiology in different latitudinal populations Estuarine. Coast. Shelf Sci. 130, 127e137. Escapa, M., Perillo, G.M.E., Iribarne, O., 2015. Biogeomorphically driven salt pan formation in Sarcocornia-dominated salt-marshes. Geomorphology 228, 147e157. Fanjul, E., Grela, M.A., Iribarne, O., 2007. Effects of the dominant SW Atlantic intertidal burrowing crab Chasmagnathus granulatus on sediment chemistry and nutrient distribution. Mar. Ecol. Prog. Ser. 341, 177e190. Feist, B.E., Simenstad, C.A., 2000. Expansion rates and recruitment frequency of exotic smooth cordgrass, Spartina alterniflora (Loisel), colonizing unvegetated littoral flats in Willapa Bay, Washington. Estuaries 23, 267e274. Fonseca, M.S., Kenworthy, W.J., 1987. Effects of current on photosynthesis and distribution of seagrasses. Aquat. Bot. 27, 59e78. Fonseca, M.S., Zieman, J.C., Thayer, G.W., Fisher, J.S., 1983. The role of current velocity in structuring seagrass meadows. Estuar. Coast. Shelf Sci. 17, 367e380. Fossing, H., Jørgensen, B.B., 1989. Measurement of bacterial sulphate reduction in sediments: evaluation of a single-step chromium reduction method. Biogeochemistry 8, 205e222. Freitas, R.F., Schrack, E.C., He, Q., Silliman, B.R., Furlong, E.B., Telles, A.C., Costa, C.S.B., 2016. Consumer control of the establishment of marsh foundation plants in intertidal mudflats. Mar. Ecol. Prog. Ser. 547, 79e89. Hopfensperger, K.N., Amy, J., Burgin, V., Schoepfer, A., Helton, A.M., 2014. Impacts of saltwater incursion on plant communities, anaerobic microbial metabolism, and resulting relationships in a restored freshwater wetland. Ecosystems 17, 792e807. Howarth, R.W., Merkel, S., 1984. Pyrite formation and the measurement of sulfate reduction in salt marsh sediments. Limnol. Oceanogr. 29, 598e608. Hughes, A.L.H., Wilson, A.M., Morris, J.T., 2012. Hydrologic variability in a salt marsh: assessing the links between drought and acute marsh dieback. Estuar. Coast. Shelf Sci. 111, 95e106. Lamers, L.P.M., Govers, L.L., Janssen, I.C.J.M., Geurts, J.J.M., Van der Welle, M.E.W., VanKatwijk, M.M., Van der Heide, T., Roelofs, J.G.M., Smolders, A.J.P., 2013. Sulfide as a soil phytotoxin: a review. Front. Plant Sci. Plant Physiology 4, 1e14.
lis, R.J.F., Costa, C.S.B., Hartmann, C., 2001. Evoluça ~o ao longo de 50 anos de Le ^nicas de marismas no sul do Brasil: Ilha da Po
lvora (Rio Grande, estruturas bioge RS). In: Annals of VIII Congresso da Associaç~ ao Brasileira de Estudos do Quar
rio (ABEQUA). Associaç~ tena ao Brasileira de Estudos do Quarten
ario-UFRGS, Porto Alegre, pp. 483e484 (in Portuguese). Lewis, J.P., Stofella, S.L., Feldman, S.R., 2001. Monk's tonsure-like gaps in the tussock grass Spartina argentinensis (Gramineae). Rev. Biol. Trop. 49, 313e316. Lewis, J.P., Stofella, S.L., Prado, D.E., Pire, E.F., Francheschi, E.A., Carnevale, N.J., 1990. Dynamics and development of floristic richness in the vegetation of a large depressed area of the Great Chaco. Flora 184, 63e77. Liu, Z.F., Lee, C., Aller, R.C., 2008. Drying effects on decomposition of salt marsh sediment and on lysine sorption. J. Mar. Res. 66, 665e689. Lyons, B.W., Gaudette, H.E., Gustafson, H.C., 1982. Dissolved organic carbon in pore waters from a hypersaline environment. Org. Geochem. 3, 133e135. Maki, T., Nomachi, M., Yoshida, S., Ezawa, T., 2008. Plant symbiotic microorganisms in acid sulfate soil: significance in the growth of pioneer plants. Plant Soil 310, 55e65. Marangoni, J.C., Costa, C.S.B., 2012. Short- and long-term vegetative propagation of two Spartina species on a salt marsh in southern Brazil. Estuaries Coast 35, 763e773. Marques, B., Lillebø, A.I., Pereira, E., Duarte, A.C., 2011. Mercury cycling and sequestration in salt marshes sediments: an ecosystem service provided by Juncus maritimus and Scirpus maritimus. Environ. Pollut. 159, 1869e1876. Martinetto, P., Montemayor, D.I., Alberti, J., Costa, C.S.B., Iribarne, O., 2016. Crab bioturbation and herbivory may account for variability in carbon sequestration and stocks in south west Atlantic salt marshes. Front. Mar. Sci. 3, 122. McKee, K.L., Mendelssohn, I.A., Materne, M.D., 2004. Temporary soil desiccation that may have reduced water availability, increased soil salinity, and/or caused soil acidification (via pyrite oxidation) and increased uptake of toxic metals such as Fe or Al. Glob. Ecol. Biogeogr. 13, 65e73. Minkoff, D.R., Escapa, M., Ferramola, F.E., Maraschín, S.D., Pierini, J.O., Perillo, G.M.E., Delrieux, C., 2006. Effects of crabehalophytic plant interactions on creek growth in a S.W. Atlantic salt marsh: a cellular automata model. Estuar. Coast. Shelf Sci. 69, 403e413. €ller, O.O., Castello, J.P., Vaz, A.C., 2009. The effect of river discharge and winds on Mo the interannual variability of the pink shrimp Farfantepenaeus paulensis production in Patos Lagoon. Estuaries Coasts 32, 787e796. Murray, J., Irving, R., 1895. On chemical changes which take place in the composition of the sea water associated with blue muds on the floor of the ocean. Trans. R. Soc. Edinb. 37, 481e507. Negrin, V.L., Spetter, C.V., Asteasuain, R.O., Perillo, G.M.E., Marcovecchio, J.E., 2011. Influence of flooding and vegetation on carbon, nitrogen, and phosphorus dynamics in the pore water of a Spartina alterniflora salt marsh. J. Environ. Sci. 23, 212e221. Otero, X.L., Macias, F., 2002. Spatial and seasonal variation in heavy metals in interstitial water of salt marsh soils. Environ. Pollut. 120, 183e190. Panutrakul, S., Monteny, F., Baeyens, W., 2001. Seasonal variations in sediment sulfur cycling in the Ballastplaat mudflat. Belg. Estuaries 24, 257e265.
ria líquida ae
rea de Spartina denPeixoto, A.R., Costa, C.S.B., 2004. Produç~ ao prima siflora Brong. (Poacea) no estu
ario da Lagoa dos Patos, Rio Grande do Sul, Brasil.

114

N. Mirlean, C.S.B. Costa / Estuarine, Coastal and Shelf Science 189 (2017) 104e114

Iheríngia, Ser. Bot. 59, 27e34 (in Portuguese). Perelman, A.I., Kasimov, N.S., 1999. Geochemistry of Landscape, third ed. Astreia, Moscow. 768 pp. Perillo, G.M.E., Iribarne, O.O., 2003. Processes of tidal channel development in salt and freshwater marshes. Earth Surf. Process. Landforms 28, 1473e1482. Rickard, D., Morse, J.W., 2005. Acid volatile sulfide (AVS). Mar. Chem. 97, 141e197. Sammut, J.M., Callinan, R., Fraser, G., 1995. Estuarine acidification: impacts on aquatic biota of draining acid sulphate soils. Austalian Geogr. Stud. 33, 89e100. Schoepfer, V.A., Bernhardt, E.S., Burgin, A.J., 2014. Iron clad wetlands: soil iron-

sulfur buffering determines coastal wetland response to salt water incursion. Jornal Geophys. Researsch 119, 2209e2219. USEPA, 1991. Draft Analytical Method for Determination of Acid Volatile Sulfide in Sediment. EPA-821-R-91-100, Washington, DC. €ller, O.O., Almeida, T.L., 2006. An
Vaz, A.C., Mo alise quantitativa da descarga dos rios afluentes da Lagoa dos Patos. Atlantica 28, 13e23 (in Portuguese). Wang, F., Chapman, P.M., 1999. Biological implications of sulfide in sediment e a review focusing on sediment toxicity. Environ. Toxicol. Chem. 18, 2526e2532.