The effect of salinity on the development ofLoxothylacus panopaei larvae (Crustacea: Cirripedia: Rhizocephala)

The rhizocephalan barnacleLoxothylacus panopaei, which parasitizes the mud crabRhithropanopeus harrisii, releases its larvae as nauplii. The nauplii develop through four stages to the cypris stage; both larval forms are lecithotrophs. Parasitized crabs were acclimated in 10‰, 15‰, and 20‰ water during August and the development of releasedL. panopaei larvae was monitored in a range of salinities. After 48–60 h, the nauplii and cyprids in each experimental salinity were counted and the numbers used to calculate a developmental index. The range of salinities in which successful development to cyprids occurred (developmental index >70%) could be shifted as a consequence of the acclimating salinity.L. panopaei larvae can develop successfully over a 10–30‰ salinity range, indicating that settlement onto host crabs and virgin externae is likely to take place in nature within this range.     

Distribution and abundance of grapsid crabs (Grapsidae) in a mangrove estuary: Effects of sediment characteristics,salinity tolerances,and osmoregulatory ability

Crabs (Grapsidae,Sesarma) are the dominant macrofaunal group of mangrove forest soils in northern Australia. Little is known about the ecology of these crabs or the factors that influence their distribution in mangrove forests. Pitfall traps were used to sample grapsid crabs in the Murray River estuary in north Queensland. Sampling was conducted at five sites along a salinity gradient from <1‰ at upstream sites to >35‰ at the river mouth. At each site, trapping was done in both low and high intertidal forests. We characterized the sediments at each site by measuring percent sand, silt, clay and organic matter, Eh, pH, and soil pore-water salinity. Four species of grapsids dominated the crab fauna along the Murray River (Sesarma semperi-longicristatum, S. messa, S. brevicristatum, andS. brevipes). Distinct zonation patterns were found along the salinity gradient and between high and low intertidal forests.S. messa was dominant in high intertidal, downstream forests, high and low intertidal forests in the middle to downstream portion of the river, and in low intertidal forests in the central reach of the river.S. brevipes was dominant in both low and high intertidal zone forests at low salinity upstream sites.S. brevicristatum was most abundant in the central reaches of the river and only in the high intertidal zone.S. semperi-longicristatum was found only in the low intertidal zone, downstream forest. Subsequently, tests of salinity tolerances of these crabs were carried out in the laboratory. These indicated very wide tolerances over salinities from completely fresh to hypersaline (60‰). The osmoregulatory abilities of the crabs were also found to vary. However, neither their salinity tolerance nor osmoregulatory ability adequately explain the zonation patterns were measured in the field. For example,S. brevicristatum had the most restricted distribution, but it had the second broadest salinity tolerance and osmoregulatory ability. Sediment characteristics explained a significant amount of the variation in abundance for two of the crab species. Pore-water salinity provided no explanatory power for any of the species. Individual species abundances are probably influenced by additional factors such as interspecific competition and predation.     

Growth and mortality rates of larval American shad,Alosa sapidissima,at different salinities

The tolerance of post yolk-sac American shad Alosa sapidissima larvae to salinities typically seen in estuaries was assessed experimentally. Sixteen-day-old Hudson River (experiment I) and 35-d-old Delaware River (experiment II) larvae were held for 8 d and 9 d respectively in low (0–1‰), medium (9–11‰), and highly (19–20‰) brackish water, and mortality and growth rates were measured. Growth rates did not vary significantly among salinity treatments. Mortality in experiment I did not vary significantly among salinity treatments however, in experiment II, mortality was zero at 10‰ but higher and statistically indistinguishable between 0‰ and 20‰ In experiment II relative condition increased with salinity. These results imply that estuarine salinities neither depress growth rates nor elevate mortality rates of larval American shad when compared with freshwater conditions. We conclude that ecological factors other than the physiological effects of salinity have played more important roles in the evolution of the upriver spawning and nursery preference shown by this species.     

Salinity preferences in the stage I zoeae of three temperate zone fiddler crabs,genus Uca

Zoeae of three species of temperate zone fiddler crabs, Uca pugnax, U. minax, and U. pugilator, were reared in the laboratory. The zoeae of each species were placed individually in artificial salinity gradients and observed for specific salinity preferences. Each species of zoeae displayed a salinity preference that reflected the salinity patterns of the adult crabs of the same species. Zoeae of U. pugnax and U. pugilator, like the adult crabs, displayed a preference for higher salinities (i.e., 20.6‰±3.5 and 21.5‰±3.0, respectively). Zoeae of U. minax, like the adult crabs, displayed a preference for lower salinities (i.e., 9.8‰±2.9).     

Salinity tolerance of brown shrimpFarfantepenaeus aztecus as it relates to postlarval and juvenile survival, distribution, and growth in estuaries

The brown shrimp,Farfantepenaeus aztecus, is the major component of the Gulf of Mexico shrimp fishery, and it is critical that we understand its environmental requirements. Brown shrimp spend a large portion of their post-larval (PL) and juvenile life within estuaries distributed along salinity gradients and yet our understanding of the salinity tolerance of various age groups is limited. A series of 48-hr bioassays were conducted in which various ages ofF. aztecus (PL-10, PL-13, PL-15, PL-17, PL-20, and PL-23) were acclimated from a salinity of 26‰ to 1‰, 2‰, 4‰, 8‰, 12‰, and 26‰ in order to determine their tolerance to these salinities. Finally, PL-80.F. aztecus were transferred directly from 25‰ to 2‰, 4‰, and 8‰ waters to study the effects of rapid salinity reductions on juvenile survival. Survival of 10-and 13-day-old PLs was significantly, different from the control (26‰) for all salinities tested. Survival of PL-15 shrimp and older was significantly lower than survival of the controls at 1‰ and 2‰ but similar to the control at all other salinities tested. A 4-wk growth trial was conducted with juvenile shrimp at 2‰, 4‰, 8‰, and 12‰. There was no significant difference in survival among treatments, although shrimp maintained at 8‰ and 12‰ grew significantlymore than shrimp maintained at 2‰ and 4‰. There was no growth difference between shrimp at the two low salinities or between shrimp at the two high salinities. Survival of juveniles transferred directly from 25‰ to various salinities were 100% at 25‰, 94.2% at 8‰, 67.3% at 4‰, and 63.5% at 2‰. These results suggest that PL-13 and younger brown shrimp would have a better chance of survival by delaying entry into estuaries susceptible to rapid salinity declines. The brown shrimp juveniles would, be more densely distributed in areas with salinities greater than 4‰ than in salinities less than 4‰. Although food availability, and bottom type also affect shrimp distribution survival and growth, salinity may also greatly affect the shrimp and its fishery.     

Effect of salinity and temperature on survival and development of young zebra (Dreissena polymorpha) and quagga (Dreissena bugensis) mussels

We reared larval zebra mussels,Dreissena polymorpha, and quagga mussels,D. bugensis, through and beyond metamorphosis (settlement) at salinities of 0–8‰. Juvenile zebra mussels gradually acclimated to 8‰ and 10‰ have been reared at these salinities for over 8 mo. Tolerance to both higher temperatures and higher salinities increases with larval age in both species (though zebra mussel embryos and larvae have a greater degree of salinity tolerance than quagga mussel embryos and larvae). Thus, only 6% of 3-day-old zebra mussel veligers survived after exposure to 4‰ for 8 additional days, whereas there was 22% survival of veligers placed in 4‰ at day 13 and grown to settlement 11 d later. Zebra mussel pediveligers, acclimated to increasing salinity in 2‰ increments beginning at day 23, continued to survive and grow in 8‰ after 5-mo exposure, though the growth rates of these juveniles were significantly less than those of juveniles reared in lower salinities. Quagga mussels did not metamorphose and settle as quickly as zebra mussel pediveligers. No quagga mussel pediveligers had settled before exposure to artificial fresh water (AFW), 2‰ 4‰, 6‰, and 8‰ on day 30. Percent settlement of these quagga mussel juveniles (based on 100% survival at the start of experiments on day 30) was 90% in AFW, 67% at 2‰, 69% at 4‰, 46% at 6‰, and 0.1% at 8‰.     

Recruitment of the Crabs Eurypanopeus depressus, Rhithropanopeus harrisii, and Petrolisthes armatus to Oyster Reefs: the Influence of Freshwater Inflow

Oyster reefs provide structural habitat for resident crabs and fishes, most of which have planktonic larvae that are dependent upon transport/retention processes for successful settlement. High rates of freshwater inflow have the potential to disrupt these processes, creating spatial gaps between larval distribution and settlement habitat. To investigate whether inflow can impact subsequent recruitment to oyster reefs, densities of crab larvae and post-settlement juveniles and adults were compared in Estero Bay, Florida, over 22 months (2005–2006). Three species were selected for comparison: Petrolisthes armatus, Eurypanopeus depressus, and Rhithropanopeus harrisii. All are important members of oyster reef communities in Southwest Florida; all exhibit protracted spawning, with larvae present throughout the year; and each is distributed unevenly on reefs in different salinity regimes. Recruitment to oyster reefs was positively correlated with bay-wide larval supply at all five reefs examined. Species-specific larval connectivity to settlement sites was altered by inflow: where connectivity was enhanced by increased inflow, stock–recruitment curves were linear; where connectivity was reduced by high inflows, stock–recruitment curves were asymptotic at higher larval densities. Maximum recruit density varied by an order of magnitude among reefs. Although live oyster density was a good indicator of habitat quality in regard to crab density, it did not account for the high variability in recruit densities. Variation in recruit density at higher levels of larval supply may primarily be caused by inflow-induced variation in larval connectivity, creating an abiotic simulation of what has widely been regarded as density dependence in stock–recruitment curves.     

Using submerged aquatic vegetation to establish minimum and maximum freshwater inflows to the Caloosahatchee estuary, Florida

Species of submerged aquatic vegetation (SAV) are frequently used in the management of estuarine systems to set restoration goals, nutrient load reduction goals, and water quality targets. As human need for water increases, the amount of freshwater required by estuaries has become an increasingly important issue. While the, science of establishing the freshwater needs of estuaries is not well developed, recent attempts have emphasized the freshwater requirements of fisheries. We evaluate the hypothesis that SAV can be used to establish freshwater inflow needs. Salinity tolerance data from laboratory and field studies of SAV in the Caloosahatchee estuary, Florida, are used to estimate a minimum flow required to maintain the salt-tolerant freshwater species,Vallisneria americana, at the head of the estuary and a maximum flow required to prevent mortality, of the marine speciesHalodule wrightii at its mouth. ForV. americana, laboratory experiments showed that little or no growth occurred between 10‰ and 15‰ In the field, lower shoot densities (<400 shoots m^?2) were associated with salinities greater than 10‰. Results forH. wrightii were more variable than forV. americana. Laboratory experiments indicated that mortality could occur at salinities <6‰, with little growth occurring between 6‰ and 12‰. Field data indicated that higher blade densities (>600 blades m^?2) tend to occur at salinities greater than 12‰ Relationships between salinity in the estuary and discharge from the Caloosahatchee River indicated that flows>8.5 m³ s^?1 would produce tolerable salinity (<10‰) forV. americana and flows<89 m³ s^?1 would avoid lethal salinities (<6‰) forH. wrightii.     

Salinity detection and associated behavior in the dungeness crab,Cancer magister

Dungeness crabs, Cancer magister, showed the same antennular flicking response to brief (<2 min) salinity fluctuations as they have previously shown in detecting chemical food cues during other studies. The threshold concentrations at which 50% of the crabs detected the salinity changes were 29.9‰ and 32.7‰ or 96% and 105% of ambient seawater (31.0‰). At the maximum salinity changes used, other behaviors accompanied the flicking response. In a second experiment where salinity rose or fell continuously, two previously undescribed behaviors, pulsing and closure, occurred. In pulsing, crabs showed a rapid coordinated opening and closing of the outer maxillipeds with rapid beating of the maxillipedal flagellae. In closure, crabs stopped all overt activity, retracted their appendages and tightly closed the buccal cavity with the outer maxillipeds. Under increasing salinity crabs exhibited pulsing at 34.9‰ or 113% of ambient seawater and closure at 36.2‰ or 117% of ambient. Under decreasing salinity crabs showed pulsing at 23.2‰ or 75% of ambient seawater and closure at 15.5‰ or 50% of ambient.     

Seasonal Prevalence of Hematodinium sp. Infections of Blue Crabs in Three South Carolina (USA) Rivers

Blue crab, Callinectes sapidus, commercial landings in the USA have been declining at an alarming rate. In South Carolina, these declines are significantly correlated with years of decreased rainfall and elevated salt marsh salinity. Previous studies suggest that higher salinity increases the risk of infection by Hematodinium sp., a dinoflagellate parasite of blue crabs, C. sapidus. A 4-year survey (June 2008 to March 2012) of blue crabs in the ACE Basin National Estuarine Research Reserve documented (1) the temporal and spatial patterns of Hematodinium sp. infection in relation to salinity, (2) some environmental correlates of disease prevalence, and (3) the characteristics of infected blue crabs. Sampling was conducted four times a year in March, June, September, and December in the South Edisto, Ashepoo, and Combahee rivers beginning in June 2008. Crab hemolymph samples were collected and preserved and DNA was successfully amplified for 2,303 individuals. Hematodinium sp. infection was evaluated by PCR amplification of its 18S rRNA gene and adjacent regions. Prevalence was highest in December 2008 in the Combahee River at sites closest to St. Helena Sound. The spatial and temporal pattern of Hematodinium sp. infection was correlated with several environmental parameters. Infected crabs exhibited differences in carapace shape and body condition compared to uninfected crabs. Overall, these results suggest that blue crabs in regions of higher salinity are at greater risk of infection by Hematodinium sp. and infected individuals exhibit sub-lethal effects of the disease.     

The growth of submersed macrophytes under experimental salinity and light conditions

The growth, morphology, and chemical composition ofHydrilla verticillata, Myriophyllum spicatum, Potamogeton perfoliatus, andVallisneria americana were compared among different salinity and light conditions. Plants were grown in microcosms (1.2 m⁵) under ambient photoperiod adjusted to 50% and 8% of solar radiation. The culture solution in five pairs of tanks was gradually adjusted to salinities of 0, 2, 4, 6, and 12‰. With the exception ofH. verticillata, the aquatic macrophytes examined may be considered eurysaline species that are able to adapt to salinities one-third the strength of sea water. With increasing salinity, the inflorescence production decreased inM. spicatum andP. perfoliatus, yet asexual reproduction in the latter species by underground buds remained constant. Stem elongation increased in response to shading inM. spicatum, while shadedP. perfoliatus had higher concentrations of chlorophylla. In association with high epiphytic mass, chlorophylla concentrations in all species were greatest at 12‰. The concentration of sodium increased in all four species of aquatic macrophytes examined here, indicating that these macrophytes did not possess mechanisms to exclude this ion. The nitrogen content (Y) of the aquatic macrophytes tested increased significantly with higher sodium concentration (X), suggesting that nitrogen may be utilized in osmoregulation (Y = X × 0.288 + 6.10, r² = 0.71). The tolerance ofV. americana andP. perfoliatus to salinity was greater in our study compared to other investigations. This may be associated with experimental methodology, whereby macrophytes were subjected to more gradual rather than abrupt changes in salinity. The two macrophytes best adapted to estuarine conditions in this study by exhibiting growth up to 12‰, includingM. spicatum andV. americana, also exhibited a greater degree of response in morphology, tissue chemistry (including chlorophyll content and total nitrogen), and reproductive output in response to varying salinity and light conditions.     

Subtidal distribution of barnacles (Cirripedia: Balanidae) in Chesapeake Bay,Maryland

From 1977 to 1980, samples of barnacles were collected (as opportunities arose) from 61 subtidal locations (mostly oyster beds) around Chesapeake Bay, Maryland. Three species were identified from the area.Balanus improvisus dominated, comprising 83% of the 8,231 barnacles identified, and was collected at all locations but one. It occurred over a collection salinity range of 0.8‰ to 17.9‰.Balanus subalbidus (14% of the barnacles identified) was collected over the same salinity range, but mainly in lower salinity waters.Balanus eburneus was scarce (2% of the barnacles identified) and was collected at higher salinities (8.5‰ to 17.1‰).     

Mechanisms of expansion for an introduced species of cordgrass, Spartina densiflora, in Humboldt Bay, California

The dominant plant in Humboldt Bay salt marshes in Spartina densiflora, a species of cordgrass apparently introduced from South America. At several salt marshes and restoration sites around Humboldt Bay, distribution of this plant has increased significantly. We investigated the relative contributions of vegetative tiller production and seed germination to the establishment and expansion of S. densiflora. Lateral spread of plants surrounded by competitors were compared to areas without competing plant species. Plants growing in areas without competitors had significantly higher rates of vegetative expansion (p<0.0001). Viable seed production, germination rates, seedling survivorship, and growth of adult plants were measured in six salinity treatments. Approximately 1,977±80 viable seeds are produced per plant (0.25–0.5 m²). The number of germinating seeds was inversely related to increases in salinity. Salinity treatments between 19‰ and 35‰ produced significantly lower germination rates than salinities of 0–18‰ (p<0.0001). Seedling survivorship was 50% at ≤4‰ and 8–14% at ≥11‰. Lateral expansion of adult, greenhouse-grown plants occurred in all salinity treatments, with modest decreases in the highest salinity treatments (p<0.05). Our findings indicate that S. densiflora expands primarily by vegetative expansion, and lateral tillers are produced by throughout the year. Spartina densiflora produces prolific amounts of seed, but recruitment in mature salt marshes may be limited by competitors and higher salinities. At restoration sites, planting of native species such as Salicornia virginica, Distichlis spicata, or Jaumea carnosa may prevent monospecific stands of S. densiflora from developing.     

Swimming speeds of oyster larvaeCrassostrea virginica in different salinities and temperatures

Swimming speeds of oyster larvaeCrassostrea virginica were determined in constant and increasing salinities to learn more about the oyster larval “salinity response”. “Normal” non-directed swimming speeds ranged from less than 1 cm/min for early veligers to 5 cm/min for “eyed” veligers with temperature an important variable. When subjected to hourly salinity increases of 0.5‰, most larvae swam upward or downward at approximately 3 times the above speeds. Larvae with values closed in response to traces of formalin sank at speeds of 5 to 50 cm/min depending on larval stage. The results may explain the differential vertical position of larval stages in estuaries and suggest the presence of a taxic component to the salinity response.     

There is no horohalinicum

Species abundance declines to a minimum (the Artenminimum) between 5 and 8‰, not only in estuaries, but in all bodies of brackish water. Khlebovich (1968) examined published hydrochemical data for estuaries and concluded that sharp changes in the ionic composition of seawater diluted with fresh water occur at salinities below 5 to 8‰. He further argued that these ionic changes constitute a physico-chemical barrier between marine and freshwater faunas. Kinne (1971) gave the name “horohalinicum” to the segment of the salinity gradient between 8 and 5‰. We have re-examined the data used by Khlebovich (1968) and found that, in fact, while the ionic composition of diluted seawater changesslightly between 8 and 5‰, the changes in ionic ratios below 2‰ are much larger. Thus, the proposed physico-chemical barrier does not exist between 8 and 5‰; it cannot then explain the Artenminimum; and there is no basis for the horohalinicum concept of Kinne (1971). Two ecological explanations for the occurrence of the Artenminimum—a species-area effect and the stability-time hypothesis—are discussed and found to be inconsistent with published data on species distributions in brackish waters. The low species diversity of brackish water may be explained, in part, by two factors: few animals evolve those physiological mechanisms required for life in the variable habitat; and these species, which are very eurytopic, have low rates of speciation.     

Larval production of the caridean shrimp,Crangon septemspinosa,in waters adjacent to Chesapeake Bay in relation to oceanographic conditions

Plankton samples of the MECCAS- (Microbial Exchanges and Coupling in Coastal Atlantic Systems) Project, taken in February, June, and August 1985 and April 1986, were analyzed to study the spatio-temporal distribution of sand shrimp, Crangon septemspinosa, larvae of Chesapeake Bay. With up to 250.9 larvae m^?3, results confirm C. septemspinosa as a very abundant decapod larval form in early spring in the study area. The overwhelming majority (94.5%) of the larvae occurred in April 1986; a second minor peak of larval production was observed in February 1985. The first two larval stages comprised 81.1% of the collected larvae, and complete series of all developmental stages including juveniles were obtained in June 1985 and April 1986. Newly hatched larvae occurred over a wide range of salinities (22.00–33.60‰), while more advanced forms were found mainly at higher salinities (>30‰). High larval abundances (>50 larvae m^?3) were obtained between 10°C; another considerably smaller peak in abundance occurred at temperatures ranging from 2.7°C to 4.5°C. Compared to other developmental stages, high abundances of the first two larval stages were collected at the highest chlorophyll concentrations. The significance of phytoplankton as a possible energy source for early stages of planktonic larvae and the role of phytoplankton as a possible chemical stimulus for larval release is discussed in terms of stomach and mouthpart structure and larval sensitivity to their chemical environment. *** DIRECT SUPPORT *** A01BY066 00016     

Fouling community of the Loxahatchee River estuary,Florida, 1980–81

Monthly growth of the fouling community at eight test panel sites in the Loxahatchee River Estuary was related to salinity and temperature. Growth was lowest in January 1981 (averaging 23 g per m², dry weight), and increased during spring and early summer with increasing water temperature. Maximum growth occurred during early or midsummer at upstream locations, before river or canal discharge substantially reduced salinity, and in late summer at downstream locations. Growth was greatest at salinities slightly less than that of seawater and decreased at salinities less than about 10‰. Growth was suppressed throughout the estuary in August 1981, probably because of the sudden decrease in temperature and salinity, and perhaps the increase in physical scouring, caused by runoff from Tropical Storm Dennis. Large loads of nutrients transported to the estuary from storm runoff, however, may have subsequently stimulated growth, which increased in September 1981 to the maximum for the year (averaging 683 g per m², dry weight).     

Dispersion and recruitment of crab larvae in the Chesapeake Bay plume: Physical and biological controls

As part of the Microbial Exchanges and Coupling in Coastal Atlantic Systems (MECCAS) Project, crab larvae were collected in the shelf waters off Chesapeake Bay in June and August 1985 and April 1986. We conducted hydrographic (temperature, salinity, nutrients) and biological (chlorophyll, copepods) mapping in conjunction with Eulerian and Lagrangian time studies of the vertical distribution of crab larvae in the Chesapeake Bay plume. These abundance estimates are used with current meter records and drifter trajectories to infer mechanisms of larval crab dispersion to the shelf waters and recruitment back into Chesapeake Bay. The highest numbers of crab larvae were usually associated with the Chesapeake Bay plume, suggesting that it was the dominant source of crab larvae to shelf waters. Patches of crab larvae also were found in the higher salinity shelf waters, and possibly were remnants of previous plume discharge events. The distribution of crab larvae in the shelf waters changed on 1–2 d time scales as a consequence of both variations in the discharge rate of the Chesapeake Bay plume and local wind-driven currents. Downwelling-favorable winds (NW) intensified the coastal jet and confined the plume and crab larvae along the coast. In April during a downwelling event (when northwesterly winds predominated), crab zoeae were transported southward along the coast at speeds that at times exceeded 168 km d⁻¹. During June and August the upwelling-favorable winds (S, SW) opposed the anticyclonic turn of the plume and, via Ekman circulation, forced the plume and crab larvae to spread seaward. Plume velocities during these conditions generally were less than 48 km d⁻¹. The recruitment of crab larvae to Chesapeake Bay is facilitated in late summer by the dominance of southerly winds, which can reverse the southward flow of shelf waters. Periodic downwelling-favorable winds can result in surface waters and crab larvae moving toward the entrance of Chesapeake Bay. Approximately 27% of the larval crabs spend at least part of the day in bottom waters, which have a residual drift toward the bay mouth. There appears to be a variety of physical transport mechanisms that can enhance the recruitment of crab larvae into Chesapeake Bay.     

A comparison of dispersal strategies in two genera of brachyuran crab in a secondary estuary

Seasonal occurrence and vertical distribution of larvae of two genera of brachyuran crab were studied in a secondary estuary flowing into Delaware Bay. Spawning in the xanthid crabRhithropanopeus harrisii occurred earlier with peak abundance of larvae in June and with a distinct decline in abundance in August. In contrast,Uca spp. larvae reached peak abundance in August. All zoeal stages ofR. harrisii were collected in the river suggesting that larvae of this species are retained in secondary estuaries near areas of prime adult habitat. Only zoea stage I larvae and megalopa ofUca spp. were collected in the river indicating that larvae of these speies may be flushed into the Delaware Bay and may not return to secondary estuaries near areas suitable for adult habitat until the megalopa stage is reached. It is not clear if this dispersal pattern is an active or passive process. *** DIRECT SUPPORT *** A01BY019 00004     

Cattail invasion and persistence in a coastal salt marsh: The role of salinity reduction

The hypothesis thatTypha domingensis (cattail) can invade tidal marshes only after soil salinities are substantially reduced was tested experimentally by comparing the salt tolerance of seeds, seedlings, and plants reared from rhizomes. Germination rates for four southern California populations reached 100% in fresh water, decreasing to 2% at 20‰. The salt tolerance of seeds from three coastal populations was lower than that of the Salton Sea population. Salt tolerance of plants grown in the lab did not increase with age for seedlings up to 8 weeks old. Rhizome-bearing plants had greatly decreased growth at 10‰ and no growth at 25‰ However, rhizomes of about 5% of the plants survived 9 months at 45‰. The seeds and seedlings are salt sensitive, which explains why invasion into tidal marshes is restricted to prolonged periods of low soil salinity. The older, rhizome-bearing plants are salt tolerant, which explains how invading plants persist persist under hypersaline conditions.