Climate change impacts, such as accelerated sea-level rise, will affect stress gradients, yet imp... more Climate change impacts, such as accelerated sea-level rise, will affect stress gradients, yet impacts on competition/stress tolerance trade-offs and shifts in distributions are unclear. Ecosystems with strong stress gradients, such as estuaries, allow for space-for-time substitutions of stress factors and can give insight into future climate-related shifts in both resource and nonresource stresses. We tested the stress gradient hypothesis and examined the effect of increased inundation stress and biotic interactions on growth and survival of two congeneric wetland sedges, Schoenoplectus acutus and Schoenoplectus americanus. We simulated sea-level rise across existing marsh elevations and those not currently found to reflect potential future sea-level rise conditions in two tidal wetlands differing in salinity. Plants were grown individually and together at five tidal elevations, the lowest simulating an 80-cm increase in sea level, and harvested to assess differences in biomass after one growing season. Inundation time, salinity, sulfides, and redox potential were measured concurrently. As predicted, increasing inundation reduced biomass of the species commonly found at higher marsh elevations, with little effect on the species found along channel margins. The presence of neighbors reduced total biomass of both species, particularly at the highest elevation; facilitation did not occur at any elevation. Contrary to predictions , we documented the competitive superiority of the stress tolerator under increased inundation, which was not predicted by the stress gradient hypothesis. Multifactor manipulation experiments addressing plant response to accelerated climate change are integral to creating a more realistic, valuable, and needed assessment of potential ecosystem response. Our results point to the important and unpredicted synergies between physical stressors, which are predicted to increase in intensity with climate change, and competitive forces on biomass as stresses increase. K E Y W O R D S
Sediment cores were collected from five coastal wetlands in England, the Netherlands, and Poland,... more Sediment cores were collected from five coastal wetlands in England, the Netherlands, and Poland, and samples analysed for heavy metal concentrations. Depth profiles of metal concentrations were converted to time-based profiles, using vertical accretion rates obtained from 137Cs dating. Chronologies of metal concentrations from the two cores (high and low marsh) at each site showed good agreement, indicating that sediment profiles represent historical inputs of metals. Sediment metal concentrations showed that some of the areas are highly polluted, with peak sediment concentrations up to five times greater than in the oldest sediment samples (1890s to 1900s). Chronologies indicated different pollution histories for the five areas. Local hydrology appeared to be important in affecting metal accumulation. Metal concentrations have recently decreased in the cores from St. Annaland Marsh (the Netherlands) and Stiffkey Marsh (UK). Levels of metals from the Oder River remained high throughout the upper part of the cores.
This study represents the first report on sediment accretion rates using 137 Cs dating for a sout... more This study represents the first report on sediment accretion rates using 137 Cs dating for a southern California salt marsh. Vertical accretion rates ranged from 0.7 to 1.2 cm yr Ϫ1 , which is at the high end of sediment accretion values for coastal wetlands. This has lead to increases in elevation within the estuary from 18 to 35 cm over the last 35 years. Depth profiles of metal concentrations were converted to time-based profiles using vertical accretion rates. Chronologies for most cores indicate a consistent peak in sediment lead (Pb) concentrations in the early to mid 1980s, corresponding to the historic decline in Pb use, which was completed in the U.S. by the early 1980s, but not begun in Mexico until 1991. Sediment Pb levels ranged from about 6-56 g g Ϫ1 . Other metals did not show any consistent trends in sediment chronology, except for a single core from a mid-marsh site (east-mid 2), which showed a 2-3-fold increase in levels of Cu, Ni, and Zn during the past two decades. Sediment levels of copper (Cu), nickel (Ni), and zinc (Zn) ranged from 6-34 g g Ϫ1 , 11-27 g g Ϫ1 , and 42-122 g g Ϫ1 , respectively. Despite rapid industrial development of the watershed, a comparison of the sediment metal concentrations in the Tijuana Estuary to other anthropogenically-impacted estuaries in the United States and Europe, shows that metal levels in sediments of the north arm of the estuary are relatively low.
We report the first data on belowground tissue mass and nitrogen (N) concentration forSpartina fo... more We report the first data on belowground tissue mass and nitrogen (N) concentration forSpartina foliosa in southern California, assessing one natural and two constructed marshes on San Diego Bay. Biomass at the natural marsh was low compared to that of otherSpartina spp., but higher than values reported forS. foliosa in northern California. In sandy constructed marshes planted 5 and 10 years before this study,S. foliosa had lower belowground tissue N, lower N crop (%N×biomass), and shallower roots than in the adjacent natural marsh. We took advantage of a 2-yr, large-scale fertilization project being performed in the older constructed marsh and examined biomass and N storage after N additions. Although there was a trend toward N accumulation with fertilization, N crop remained at approximately 50% of natural marsh levels, unlike the large aboveground responses to N addition in our previous studies. Lower belowground reserves help to explain poor aerial growth in the created marshes and suggest the need for finer sediments (with greater potential for holding and supplying nutrients) to sustain (S. foliosa. While fine sediments are beginning to accumulate on the surface of the created marshes, vertical accretion is more likely to shift the plant community toward other species than to enhanceS. foliosa growth. We suggest salvaging and importing fine, organic marsh sediments or providing organic amendments to establish proper substrate conditions. Overexcavating and allowing fine sediments to accumulate remains an option, although the time scale is unpredictable due to the stochasticity of accretion events.
New terminology is needed for the assessment of wetland restoration projects, focusing on progres... more New terminology is needed for the assessment of wetland restoration projects, focusing on progress towards ecological goals rather than the yes/no alternative of success versus failure. Mitigation projects should be evaluated based on 'compliance' of specific mitigation criteria. Peer-reviewed assessments of wetland restoration projects have focused on appropriate parameters (topographic, hydrologic, soil, vegetation, and animal components); however, for the most part, they have evaluated small, newly restored wetlands over the short-term with little repeated sampling. Future assessments should focus on understanding problems at a particular site by identifying cause -effect mechanisms. We review a case study from San Diego Bay where detailed assessments led to the identification of problems and subsequent management alternatives. We also recommend the use of two novel approaches: experimentation and the evaluation of ecosystem resiliency to unplanned disturbances. These assessments will provide a better understanding of ecosystem functioning and enhance current comparisons of natural and restored systems. A stronger science base for wetland restoration and adaptive management do not guarantee that restoration targets will be reached; however, these approaches will identify the causes of problems and allow us to predict whether or not restoration targets can be reached in a timely manner.
Although tidal wetland vegetation patterns are typically related to elevation, we hypothesized th... more Although tidal wetland vegetation patterns are typically related to elevation, we hypothesized that the vertical range of a species may shift where the topography is more heterogeneous. We examined plant species occurrences in relation to elevation, proximity to the bay, and proximity to tidal creeks at a near-pristine wetland in San Quintín Bay, Baja California, Mexico. At the whole-wetland scale, most species occurred primarily within a 30-cm elevation band (the marsh plain). However, Spartina foliosa occurred only at the bayward margin, even though “suitable” elevations were present further inland. A similar pattern was found in San Diego Bay. At the microtopographic scale, three species on the marsh plain were strongly influenced by elevation, whereas four species responded to both elevation and proximity to tidal creeks. The latter species tended to “avoid” the lower 10 cm of the marsh plain except near a tidal creek. Species richness was thus greater (by 0.6 species at the lowest 10-cm class) at the tidal creek margin. Better drainage near creeks is the hypothesized cause. Our results help explain why species that are transplanted to constructed wetlands do not always grow at the full range of elevations they occupy in natural wetlands. We recommend that species be introduced to their modal elevation (determined from nearby reference marshes) and that salt-marsh construction designs include topographic heterogeneity (complex tidal creek networks). The analysis of broad-scale and fine-scale patterns of occurrence also suggests new habitat nomenclature. Elevation-based terms (“low,”“middle,” and “high” marsh) should be replaced by a system that considers elevation, landscape position, and conspicuous species. We suggest three habitat designations: (a) the high marsh—a 30- to 70-cm elevation range with Salicornia subterminalis; (b) the marsh plain—a 30-cm elevation range with heterogeneous topography and up to nine common species; and (c) cordgrass habitat—the bayward portion of the marsh plain and lower elevations, all occupied by Spartina foliosa. Although these habitats do not have discrete boundaries, separate terms are needed for wetland restoration plans and these designations will improve recognition that vegetation patterns respond to horizontal, as well as vertical, position.
Our test of the hypothesis that biomass and nitrogen would increase with more species-rich planti... more Our test of the hypothesis that biomass and nitrogen would increase with more species-rich plantings simultaneously vegetated a salt marsh restoration site and demonstrated that on average, randomly chosen, 6-species plantings accumulated more biomass and nitrogen than the mean for 0-and 1-species assemblages, with the mean for 3-species assemblages being intermediate. In addition, we found that individual species (from the pool of eight native halophytes) differed in their functional capacity, with Salicornia virginica (Sv) and Jaumea carnosa contributing the greatest biomass when planted alone, while Triglochin concinna had the highest tissue N concentrations. When planted alone, Sv accumulated comparable amounts of biomass and nitrogen as in the multispecies plots, indicating that individual species can have a large effect on particular functions. Soil TKN in the surface 0-5 cm was greater in 6-species plots than unplanted plots in 1999, while both 3-and 6-species plots were greater than unplanted plots in 2000; however, there were no differences at 5-20 cm depth and no species-specific effects. Root and shoot biomass both increased with species richness, with total biomass of 6-species plots averaging 995.6 Ϯ 120.5 g/m 2 in 2000, compared to the mean for 1-species plots (572.1 Ϯ 90.3 g/m 2 ) and unplanted plots (164.5 Ϯ 24.7 g/m 2 ). Still, at the age of three years, root biomass was only about one-third that of the species-rich reference site, and shoot biomass was one-half to one-fifth the maxima reported for reference salt marshes. Species-specific effects were found for Sv, which had high biomass of both roots and shoots in the multispecies plots (55% of aboveground biomass in 3-species plots and 41% in 6-species plots) and the highest pool of N (52% of the N pool in 3-species plots and 42% in 6-species plots), even though only one-eighth of the initial plantings were Sv. However, when plots with this species were excluded from the analysis, the species-richness effect persisted. Thus, ecosystem function, as measured by biomass and N accumulation, increased with species richness regardless of dominance by the highly productive Sv. We conclude that manipulating the richness and composition of plantings offers ecosystem restorationists an effective tool for accelerating the rate of functional development.
Hypothetical models in the scientific literature suggest that ecosystem restoration and creation ... more Hypothetical models in the scientific literature suggest that ecosystem restoration and creation sites follow a smooth path of development (called a trajectory), rapidly matching natural reference sites (the target). Multi-million-dollar mitigation agreements have been based on the expectation that damages to habitat will be compensated within 5-10 years, and monitoring periods have been set accordingly. Our San Diego Bay study site, the Sweetwater Marsh National Wildlife Refuge, has one of the longest and most detailed records of habitat development at a mitigation site: data on soil organic matter, soil nitrogen, plant growth, and plant canopies for up to 10 years from a 12-year-old site. High interannual variation and lack of directional changes indicate little chance that targets will be reached in the near future. Other papers perpetuate the trajectory model, despite data that corroborate our findings. After reviewing "trajectory models" and presenting our comprehensive data for the first time, we suggest alternative management and mitigation policies.
Following heavy winter storms and sedimentation in 1993,Spartina foliosa (Pacific cordgrass) clon... more Following heavy winter storms and sedimentation in 1993,Spartina foliosa (Pacific cordgrass) clones established on a 6.5-ha mudflat in Tijuana Estuary, with over 80 new clones counted by 1997. El Niño Southern Oscillation (ENSO) storms in 1993 apparently facilitated the habitat conversion through river flooding, which caused a temporary reduction in soil salinity and delivered large volumes of sediment. Extreme sedimentation likely raised mudflat elevations enough to allowSpartina establishment. We hypothesized that clones, once established, increased sedimentation in a positive feedback loop leading to accelerated habitat conversion. We collected data on elevation,Spartina expansion, and sediment accretion in two consecutive years (1998–1999). The elevation range of the mudflats in 1998 (0.5–0.83 m NGVD) was within the elevation range ofSpartina at this site (0.39–0.83 m NGVD), indicating that remaining mudflats are at elevations suitable for further expansion. Sediment accretion ranged from 4.0–12.7 cm between 1997–1998 (ENSO conditions), but was close to long-term averages (˜ cm yr−1) in 1998–1999 (nonflood year) indicating how susceptible Tijuana Estuary is to sedimentation from episodic storms. Although accretion rates were similar withinSpartina clones and on bare mudflats over the ENSO winter, clones were typically dome-shaped suggesting higher sediment retention rates within clones. The radial expansion rates of clones (1.31±0.25 m in 1998; 1.12±0.07 m in 1999) approximated the maxima reported for this species and were not related to clone size or vigor. Conditions on the mudflat appear ideal forSpartina growth, masking differences that might otherwise be observed. Given the likelihood of sedimentation-driven habitat conversion in southern California and other Mediterranean-type estuaries, management efforts are needed to address sedimentation issues on a watershed scale. We recommend that wetland restoration projects in southern California include large areas of intertidal mudflat, both to maintain habitat for shorebird feeding and to allow colonization by salt marsh vegetation.
Tidal wetland mesocosms at Tijuana River National Estuarine Research Reserve failed to elucidate ... more Tidal wetland mesocosms at Tijuana River National Estuarine Research Reserve failed to elucidate effects of hydrologic treatments (excluded, impounded, and fully tidal systems) for most parameters measuring Salicornia virginica (pickleweed). Although soil salinity increased where tidal flushing was excluded for 10 months (salinities rose ف 20 to 50%), pickleweed cover and algal chlorophyll did not differ among treatments. Effects were seen only in pickleweed growth rates ( ف 30% decrease where tides were excluded) and normalized difference vegetation index (NDVI) measurements. We failed to show any differences between impounded and fully tidal conditions, because the mesocosms had coarse sediments, and impounded water drained easily via subsurface flow. However, the problems that we encountered with the mesocosms led to the following advice for future wetland restoration projects: (1) Mesocosms are useful for testing restoration techniques before an actual restoration project takes place. (2) Mesocosms should be used to test factors that may lead to more successful restoration in the future, including planting techniques, substrate conditions, and hydrology. (3) Mesocosms should be used to develop new assessment methods for monitoring wetland ecosystems. Because of the ability to control some environmental parameters while maintaining seminatural conditions, mesocosms offer great potential for the future evaluation of experimental restoration techniques. I Plant Cover and Growth. Pickleweed responses were measured by five methods to document treatment effects. We compared a less labor-intensive procedure (spectral reflectance) with traditional methods of canopy structure (per-
Extensive restoration efforts in southern California coastal wetlands highlight several challenge... more Extensive restoration efforts in southern California coastal wetlands highlight several challenges for urban salt marsh restoration, including: habitat isolation and fragmentation, impacts from exotic species, the loss of transitional upland habitats, and other alterations to hydrologic and sediment dynamics. Habitat isolation impairs colonization by dispersal-limited plants, so planting becomes essential to achieve diverse salt marshes. Low species richness slows the development of salt marsh functions (e.g., biomass and nitrogen accumulation) in southern California. A variety of exotic species have invaded the upper reaches of salt marshes in southern California, most commonly in marshes with hydrological modifications. The replacement of gradual slopes between wetlands and uplands by sharp transitions abutting urban development limits our ability to restore rare plant and animal populations. Where hydrologic connections are impaired by roads and other structures, the natural migration of channels is constrained, and sediment dynamics often lead to lagoon mouth closure.
In a large (8 ha) salt marsh restoration site, we tested the effects of excavating tidal creeks p... more In a large (8 ha) salt marsh restoration site, we tested the effects of excavating tidal creeks patterned after reference systems. Our purposes were to enhance understanding of tidal creek networks and to test the need to excavate creeks during salt marsh restoration. We compared geomorphic changes in areas with and without creek networks (n = 3; each area 1.3 ha) and monitored creek cross-sectional areas, creek lengths, vertical accretion, and marsh surface elevations for 5 yr that included multiple sedimentation events. We hypothesized that cells with creeks would develop different marsh surface and creek network characteristics (i.e., surface elevation change, sedimentation rate, creek cross-sectional area, length, and drainage density). Marsh surface vertical accretion averaged 1.3 cm yr−1 with large storm inputs, providing the opportunity to assess the response of the drainage network to extreme sedimentation rates. The constructed creeks initially filled due to high accretion rates but stabilized at cross-sectional areas matching, or on a trajectory toward, equilibrium values predicted by regional regression equations. Sedimentation on the marsh surface was greatest in low elevation areas and was not directly influenced by creeks. Time required for cross-sectional area stabilization ranged from 0 to > 5 yr, depending on creek order. First-order constructed creeks lengthened rapidly (mean rate of 1.3 m yr−1) in areas of low elevation and low vegetation cover. New (volunteer) creeks formed rapidly in cells without creeks in areas with low elevation, low vegetation cover, and high elevation gradient (mean rate of 6.2 m yr−1). After 5 yr, volunteer creeks were, at most, one-fourth the area of constructed creeks and had not yet reached the upper marsh plain. In just 4 yr, the site’s drainage density expanded from 0.018 to reference levels of 0.022 m m−2. Pools also formed on the marsh plain due to sediment resuspension associated with wind-driven waves. We conclude that excavated creeks jump-started the development of drainage density and creek and channel dimensions, and that the tidal prism became similar to those of the reference site in 4–5 yr.
Knowing that diverse plantings enhanced biomass and nitrogen (N) accumulation in a restored Calif... more Knowing that diverse plantings enhanced biomass and nitrogen (N) accumulation in a restored California salt marsh, we asked if the ''biodiversity effect'' was due to species selection or complementarity. In a two-year greenhouse experiment, we found positive biodiversity effects on total, root, and shoot biomass, total and root N crop, and on biomass and N allocation; negative effects on root and shoot N concentration; and no effect on shoot N crop. Overyielding among trios and sextets was supported by significant deviations in observed yield from that expected relative to solo yields (D T ). However, both trios and sextets strongly underyielded relative to the highest yielding solo in the assemblage (D max ) in all attributes, and to the dominant species in the assemblage (D dom ) in most attributes. When we decomposed biodiversity effects on shoot characteristics, selection effects primarily drove over-and underyielding. The only complementarity effect was underyielding of sextet shoot biomass. These analyses were possible because we replicated assemblages and evaluated 11 response variables. One species (Salicornia virginica) dominated functioning when present; when absent, another dominated (e.g., Frankenia salina). Effects varied with the response tested, however. For both shoot biomass and N crop, S. virginica was the dominant overyielding species (based on D ij and comparisons of trios 6 target species). For shoot N concentration, however, the dominant was Triglochin concinna, a species that had low biomass but was capable of reducing assemblage performance, presumably by concentrating N.
Climate change impacts, such as accelerated sea-level rise, will affect stress gradients, yet imp... more Climate change impacts, such as accelerated sea-level rise, will affect stress gradients, yet impacts on competition/stress tolerance trade-offs and shifts in distributions are unclear. Ecosystems with strong stress gradients, such as estuaries, allow for space-for-time substitutions of stress factors and can give insight into future climate-related shifts in both resource and nonresource stresses. We tested the stress gradient hypothesis and examined the effect of increased inundation stress and biotic interactions on growth and survival of two congeneric wetland sedges, Schoenoplectus acutus and Schoenoplectus americanus. We simulated sea-level rise across existing marsh elevations and those not currently found to reflect potential future sea-level rise conditions in two tidal wetlands differing in salinity. Plants were grown individually and together at five tidal elevations, the lowest simulating an 80-cm increase in sea level, and harvested to assess differences in biomass after one growing season. Inundation time, salinity, sulfides, and redox potential were measured concurrently. As predicted, increasing inundation reduced biomass of the species commonly found at higher marsh elevations, with little effect on the species found along channel margins. The presence of neighbors reduced total biomass of both species, particularly at the highest elevation; facilitation did not occur at any elevation. Contrary to predictions , we documented the competitive superiority of the stress tolerator under increased inundation, which was not predicted by the stress gradient hypothesis. Multifactor manipulation experiments addressing plant response to accelerated climate change are integral to creating a more realistic, valuable, and needed assessment of potential ecosystem response. Our results point to the important and unpredicted synergies between physical stressors, which are predicted to increase in intensity with climate change, and competitive forces on biomass as stresses increase. K E Y W O R D S
Sediment cores were collected from five coastal wetlands in England, the Netherlands, and Poland,... more Sediment cores were collected from five coastal wetlands in England, the Netherlands, and Poland, and samples analysed for heavy metal concentrations. Depth profiles of metal concentrations were converted to time-based profiles, using vertical accretion rates obtained from 137Cs dating. Chronologies of metal concentrations from the two cores (high and low marsh) at each site showed good agreement, indicating that sediment profiles represent historical inputs of metals. Sediment metal concentrations showed that some of the areas are highly polluted, with peak sediment concentrations up to five times greater than in the oldest sediment samples (1890s to 1900s). Chronologies indicated different pollution histories for the five areas. Local hydrology appeared to be important in affecting metal accumulation. Metal concentrations have recently decreased in the cores from St. Annaland Marsh (the Netherlands) and Stiffkey Marsh (UK). Levels of metals from the Oder River remained high throughout the upper part of the cores.
This study represents the first report on sediment accretion rates using 137 Cs dating for a sout... more This study represents the first report on sediment accretion rates using 137 Cs dating for a southern California salt marsh. Vertical accretion rates ranged from 0.7 to 1.2 cm yr Ϫ1 , which is at the high end of sediment accretion values for coastal wetlands. This has lead to increases in elevation within the estuary from 18 to 35 cm over the last 35 years. Depth profiles of metal concentrations were converted to time-based profiles using vertical accretion rates. Chronologies for most cores indicate a consistent peak in sediment lead (Pb) concentrations in the early to mid 1980s, corresponding to the historic decline in Pb use, which was completed in the U.S. by the early 1980s, but not begun in Mexico until 1991. Sediment Pb levels ranged from about 6-56 g g Ϫ1 . Other metals did not show any consistent trends in sediment chronology, except for a single core from a mid-marsh site (east-mid 2), which showed a 2-3-fold increase in levels of Cu, Ni, and Zn during the past two decades. Sediment levels of copper (Cu), nickel (Ni), and zinc (Zn) ranged from 6-34 g g Ϫ1 , 11-27 g g Ϫ1 , and 42-122 g g Ϫ1 , respectively. Despite rapid industrial development of the watershed, a comparison of the sediment metal concentrations in the Tijuana Estuary to other anthropogenically-impacted estuaries in the United States and Europe, shows that metal levels in sediments of the north arm of the estuary are relatively low.
We report the first data on belowground tissue mass and nitrogen (N) concentration forSpartina fo... more We report the first data on belowground tissue mass and nitrogen (N) concentration forSpartina foliosa in southern California, assessing one natural and two constructed marshes on San Diego Bay. Biomass at the natural marsh was low compared to that of otherSpartina spp., but higher than values reported forS. foliosa in northern California. In sandy constructed marshes planted 5 and 10 years before this study,S. foliosa had lower belowground tissue N, lower N crop (%N×biomass), and shallower roots than in the adjacent natural marsh. We took advantage of a 2-yr, large-scale fertilization project being performed in the older constructed marsh and examined biomass and N storage after N additions. Although there was a trend toward N accumulation with fertilization, N crop remained at approximately 50% of natural marsh levels, unlike the large aboveground responses to N addition in our previous studies. Lower belowground reserves help to explain poor aerial growth in the created marshes and suggest the need for finer sediments (with greater potential for holding and supplying nutrients) to sustain (S. foliosa. While fine sediments are beginning to accumulate on the surface of the created marshes, vertical accretion is more likely to shift the plant community toward other species than to enhanceS. foliosa growth. We suggest salvaging and importing fine, organic marsh sediments or providing organic amendments to establish proper substrate conditions. Overexcavating and allowing fine sediments to accumulate remains an option, although the time scale is unpredictable due to the stochasticity of accretion events.
New terminology is needed for the assessment of wetland restoration projects, focusing on progres... more New terminology is needed for the assessment of wetland restoration projects, focusing on progress towards ecological goals rather than the yes/no alternative of success versus failure. Mitigation projects should be evaluated based on 'compliance' of specific mitigation criteria. Peer-reviewed assessments of wetland restoration projects have focused on appropriate parameters (topographic, hydrologic, soil, vegetation, and animal components); however, for the most part, they have evaluated small, newly restored wetlands over the short-term with little repeated sampling. Future assessments should focus on understanding problems at a particular site by identifying cause -effect mechanisms. We review a case study from San Diego Bay where detailed assessments led to the identification of problems and subsequent management alternatives. We also recommend the use of two novel approaches: experimentation and the evaluation of ecosystem resiliency to unplanned disturbances. These assessments will provide a better understanding of ecosystem functioning and enhance current comparisons of natural and restored systems. A stronger science base for wetland restoration and adaptive management do not guarantee that restoration targets will be reached; however, these approaches will identify the causes of problems and allow us to predict whether or not restoration targets can be reached in a timely manner.
Although tidal wetland vegetation patterns are typically related to elevation, we hypothesized th... more Although tidal wetland vegetation patterns are typically related to elevation, we hypothesized that the vertical range of a species may shift where the topography is more heterogeneous. We examined plant species occurrences in relation to elevation, proximity to the bay, and proximity to tidal creeks at a near-pristine wetland in San Quintín Bay, Baja California, Mexico. At the whole-wetland scale, most species occurred primarily within a 30-cm elevation band (the marsh plain). However, Spartina foliosa occurred only at the bayward margin, even though “suitable” elevations were present further inland. A similar pattern was found in San Diego Bay. At the microtopographic scale, three species on the marsh plain were strongly influenced by elevation, whereas four species responded to both elevation and proximity to tidal creeks. The latter species tended to “avoid” the lower 10 cm of the marsh plain except near a tidal creek. Species richness was thus greater (by 0.6 species at the lowest 10-cm class) at the tidal creek margin. Better drainage near creeks is the hypothesized cause. Our results help explain why species that are transplanted to constructed wetlands do not always grow at the full range of elevations they occupy in natural wetlands. We recommend that species be introduced to their modal elevation (determined from nearby reference marshes) and that salt-marsh construction designs include topographic heterogeneity (complex tidal creek networks). The analysis of broad-scale and fine-scale patterns of occurrence also suggests new habitat nomenclature. Elevation-based terms (“low,”“middle,” and “high” marsh) should be replaced by a system that considers elevation, landscape position, and conspicuous species. We suggest three habitat designations: (a) the high marsh—a 30- to 70-cm elevation range with Salicornia subterminalis; (b) the marsh plain—a 30-cm elevation range with heterogeneous topography and up to nine common species; and (c) cordgrass habitat—the bayward portion of the marsh plain and lower elevations, all occupied by Spartina foliosa. Although these habitats do not have discrete boundaries, separate terms are needed for wetland restoration plans and these designations will improve recognition that vegetation patterns respond to horizontal, as well as vertical, position.
Our test of the hypothesis that biomass and nitrogen would increase with more species-rich planti... more Our test of the hypothesis that biomass and nitrogen would increase with more species-rich plantings simultaneously vegetated a salt marsh restoration site and demonstrated that on average, randomly chosen, 6-species plantings accumulated more biomass and nitrogen than the mean for 0-and 1-species assemblages, with the mean for 3-species assemblages being intermediate. In addition, we found that individual species (from the pool of eight native halophytes) differed in their functional capacity, with Salicornia virginica (Sv) and Jaumea carnosa contributing the greatest biomass when planted alone, while Triglochin concinna had the highest tissue N concentrations. When planted alone, Sv accumulated comparable amounts of biomass and nitrogen as in the multispecies plots, indicating that individual species can have a large effect on particular functions. Soil TKN in the surface 0-5 cm was greater in 6-species plots than unplanted plots in 1999, while both 3-and 6-species plots were greater than unplanted plots in 2000; however, there were no differences at 5-20 cm depth and no species-specific effects. Root and shoot biomass both increased with species richness, with total biomass of 6-species plots averaging 995.6 Ϯ 120.5 g/m 2 in 2000, compared to the mean for 1-species plots (572.1 Ϯ 90.3 g/m 2 ) and unplanted plots (164.5 Ϯ 24.7 g/m 2 ). Still, at the age of three years, root biomass was only about one-third that of the species-rich reference site, and shoot biomass was one-half to one-fifth the maxima reported for reference salt marshes. Species-specific effects were found for Sv, which had high biomass of both roots and shoots in the multispecies plots (55% of aboveground biomass in 3-species plots and 41% in 6-species plots) and the highest pool of N (52% of the N pool in 3-species plots and 42% in 6-species plots), even though only one-eighth of the initial plantings were Sv. However, when plots with this species were excluded from the analysis, the species-richness effect persisted. Thus, ecosystem function, as measured by biomass and N accumulation, increased with species richness regardless of dominance by the highly productive Sv. We conclude that manipulating the richness and composition of plantings offers ecosystem restorationists an effective tool for accelerating the rate of functional development.
Hypothetical models in the scientific literature suggest that ecosystem restoration and creation ... more Hypothetical models in the scientific literature suggest that ecosystem restoration and creation sites follow a smooth path of development (called a trajectory), rapidly matching natural reference sites (the target). Multi-million-dollar mitigation agreements have been based on the expectation that damages to habitat will be compensated within 5-10 years, and monitoring periods have been set accordingly. Our San Diego Bay study site, the Sweetwater Marsh National Wildlife Refuge, has one of the longest and most detailed records of habitat development at a mitigation site: data on soil organic matter, soil nitrogen, plant growth, and plant canopies for up to 10 years from a 12-year-old site. High interannual variation and lack of directional changes indicate little chance that targets will be reached in the near future. Other papers perpetuate the trajectory model, despite data that corroborate our findings. After reviewing "trajectory models" and presenting our comprehensive data for the first time, we suggest alternative management and mitigation policies.
Following heavy winter storms and sedimentation in 1993,Spartina foliosa (Pacific cordgrass) clon... more Following heavy winter storms and sedimentation in 1993,Spartina foliosa (Pacific cordgrass) clones established on a 6.5-ha mudflat in Tijuana Estuary, with over 80 new clones counted by 1997. El Niño Southern Oscillation (ENSO) storms in 1993 apparently facilitated the habitat conversion through river flooding, which caused a temporary reduction in soil salinity and delivered large volumes of sediment. Extreme sedimentation likely raised mudflat elevations enough to allowSpartina establishment. We hypothesized that clones, once established, increased sedimentation in a positive feedback loop leading to accelerated habitat conversion. We collected data on elevation,Spartina expansion, and sediment accretion in two consecutive years (1998–1999). The elevation range of the mudflats in 1998 (0.5–0.83 m NGVD) was within the elevation range ofSpartina at this site (0.39–0.83 m NGVD), indicating that remaining mudflats are at elevations suitable for further expansion. Sediment accretion ranged from 4.0–12.7 cm between 1997–1998 (ENSO conditions), but was close to long-term averages (˜ cm yr−1) in 1998–1999 (nonflood year) indicating how susceptible Tijuana Estuary is to sedimentation from episodic storms. Although accretion rates were similar withinSpartina clones and on bare mudflats over the ENSO winter, clones were typically dome-shaped suggesting higher sediment retention rates within clones. The radial expansion rates of clones (1.31±0.25 m in 1998; 1.12±0.07 m in 1999) approximated the maxima reported for this species and were not related to clone size or vigor. Conditions on the mudflat appear ideal forSpartina growth, masking differences that might otherwise be observed. Given the likelihood of sedimentation-driven habitat conversion in southern California and other Mediterranean-type estuaries, management efforts are needed to address sedimentation issues on a watershed scale. We recommend that wetland restoration projects in southern California include large areas of intertidal mudflat, both to maintain habitat for shorebird feeding and to allow colonization by salt marsh vegetation.
Tidal wetland mesocosms at Tijuana River National Estuarine Research Reserve failed to elucidate ... more Tidal wetland mesocosms at Tijuana River National Estuarine Research Reserve failed to elucidate effects of hydrologic treatments (excluded, impounded, and fully tidal systems) for most parameters measuring Salicornia virginica (pickleweed). Although soil salinity increased where tidal flushing was excluded for 10 months (salinities rose ف 20 to 50%), pickleweed cover and algal chlorophyll did not differ among treatments. Effects were seen only in pickleweed growth rates ( ف 30% decrease where tides were excluded) and normalized difference vegetation index (NDVI) measurements. We failed to show any differences between impounded and fully tidal conditions, because the mesocosms had coarse sediments, and impounded water drained easily via subsurface flow. However, the problems that we encountered with the mesocosms led to the following advice for future wetland restoration projects: (1) Mesocosms are useful for testing restoration techniques before an actual restoration project takes place. (2) Mesocosms should be used to test factors that may lead to more successful restoration in the future, including planting techniques, substrate conditions, and hydrology. (3) Mesocosms should be used to develop new assessment methods for monitoring wetland ecosystems. Because of the ability to control some environmental parameters while maintaining seminatural conditions, mesocosms offer great potential for the future evaluation of experimental restoration techniques. I Plant Cover and Growth. Pickleweed responses were measured by five methods to document treatment effects. We compared a less labor-intensive procedure (spectral reflectance) with traditional methods of canopy structure (per-
Extensive restoration efforts in southern California coastal wetlands highlight several challenge... more Extensive restoration efforts in southern California coastal wetlands highlight several challenges for urban salt marsh restoration, including: habitat isolation and fragmentation, impacts from exotic species, the loss of transitional upland habitats, and other alterations to hydrologic and sediment dynamics. Habitat isolation impairs colonization by dispersal-limited plants, so planting becomes essential to achieve diverse salt marshes. Low species richness slows the development of salt marsh functions (e.g., biomass and nitrogen accumulation) in southern California. A variety of exotic species have invaded the upper reaches of salt marshes in southern California, most commonly in marshes with hydrological modifications. The replacement of gradual slopes between wetlands and uplands by sharp transitions abutting urban development limits our ability to restore rare plant and animal populations. Where hydrologic connections are impaired by roads and other structures, the natural migration of channels is constrained, and sediment dynamics often lead to lagoon mouth closure.
In a large (8 ha) salt marsh restoration site, we tested the effects of excavating tidal creeks p... more In a large (8 ha) salt marsh restoration site, we tested the effects of excavating tidal creeks patterned after reference systems. Our purposes were to enhance understanding of tidal creek networks and to test the need to excavate creeks during salt marsh restoration. We compared geomorphic changes in areas with and without creek networks (n = 3; each area 1.3 ha) and monitored creek cross-sectional areas, creek lengths, vertical accretion, and marsh surface elevations for 5 yr that included multiple sedimentation events. We hypothesized that cells with creeks would develop different marsh surface and creek network characteristics (i.e., surface elevation change, sedimentation rate, creek cross-sectional area, length, and drainage density). Marsh surface vertical accretion averaged 1.3 cm yr−1 with large storm inputs, providing the opportunity to assess the response of the drainage network to extreme sedimentation rates. The constructed creeks initially filled due to high accretion rates but stabilized at cross-sectional areas matching, or on a trajectory toward, equilibrium values predicted by regional regression equations. Sedimentation on the marsh surface was greatest in low elevation areas and was not directly influenced by creeks. Time required for cross-sectional area stabilization ranged from 0 to > 5 yr, depending on creek order. First-order constructed creeks lengthened rapidly (mean rate of 1.3 m yr−1) in areas of low elevation and low vegetation cover. New (volunteer) creeks formed rapidly in cells without creeks in areas with low elevation, low vegetation cover, and high elevation gradient (mean rate of 6.2 m yr−1). After 5 yr, volunteer creeks were, at most, one-fourth the area of constructed creeks and had not yet reached the upper marsh plain. In just 4 yr, the site’s drainage density expanded from 0.018 to reference levels of 0.022 m m−2. Pools also formed on the marsh plain due to sediment resuspension associated with wind-driven waves. We conclude that excavated creeks jump-started the development of drainage density and creek and channel dimensions, and that the tidal prism became similar to those of the reference site in 4–5 yr.
Knowing that diverse plantings enhanced biomass and nitrogen (N) accumulation in a restored Calif... more Knowing that diverse plantings enhanced biomass and nitrogen (N) accumulation in a restored California salt marsh, we asked if the ''biodiversity effect'' was due to species selection or complementarity. In a two-year greenhouse experiment, we found positive biodiversity effects on total, root, and shoot biomass, total and root N crop, and on biomass and N allocation; negative effects on root and shoot N concentration; and no effect on shoot N crop. Overyielding among trios and sextets was supported by significant deviations in observed yield from that expected relative to solo yields (D T ). However, both trios and sextets strongly underyielded relative to the highest yielding solo in the assemblage (D max ) in all attributes, and to the dominant species in the assemblage (D dom ) in most attributes. When we decomposed biodiversity effects on shoot characteristics, selection effects primarily drove over-and underyielding. The only complementarity effect was underyielding of sextet shoot biomass. These analyses were possible because we replicated assemblages and evaluated 11 response variables. One species (Salicornia virginica) dominated functioning when present; when absent, another dominated (e.g., Frankenia salina). Effects varied with the response tested, however. For both shoot biomass and N crop, S. virginica was the dominant overyielding species (based on D ij and comparisons of trios 6 target species). For shoot N concentration, however, the dominant was Triglochin concinna, a species that had low biomass but was capable of reducing assemblage performance, presumably by concentrating N.
Uploads
Papers by John Callaway