Temporal trends in tidal marsh restoration at Hester Marsh, Elkhorn Slough, California
Data files
Jan 07, 2025 version files 6.61 MB
-
Hester_Compare_Phase_Salicornia.csv
857 B
-
README.md
16.51 KB
-
Wasson_Compare_Phase_Multispecies.csv
9.03 KB
-
Wasson_Compare_Years_Distichlis_Size.csv
1.26 KB
-
Wasson_Compare_Years_Distichlis_Survival.csv
928 B
-
Wasson_Compare_Years_Salicornia.csv
840 B
-
Wasson_Facilitation.csv
1.06 KB
-
Wasson_Inundation.csv
1.20 KB
-
Wasson_Seedlings.csv
9.30 KB
-
Wasson_Soil_Trajectory.csv
6.09 KB
-
Wasson_Temperature_Precipitation.csv
5.69 MB
-
Wasson_UAV_Cover.csv
876.91 KB
Abstract
Understanding restoration trajectories and their sensitivity to climate is critical for designing effective adaptation strategies for restoration projects. Tidal marsh restoration often involves initial bare earth conditions that may be stressful to colonizing plants, especially on high elevation marsh platforms built to be sea-level resilient. Under these circumstances, stressors such as soil salinity may increase over time, but can be mitigated by strong rainfall. At Hester Marsh, a large tidal marsh restoration site in Elkhorn Slough, California, we evaluated passive restoration success, tracking colonization by plants whose seeds arrived naturally on tides, and active restoration success, monitoring greenhouse-grown transplants. Our investigation revealed non-linear restoration trajectories with high climate sensitivity, at the scale of the entire landscape and of individual plants. We found strong effects of drought on marsh restoration success indicators. Restoration success decreased dramatically in the first area to be completed, which experienced drought conditions following construction. In contrast, restoration success declined more slowly in the second area, which had consecutive rainy years following construction. Both passive and active restoration showed strong differences across these areas and across dry and rainy years. Facilitation can sometimes improve conditions for later arriving plants, but we found higher mortality of seedlings under existing vegetation than in bare areas. Thus, plant colonization may slow over time both due to increasing abiotic stress and through competition by early colonizers. Our findings lead to concrete recommendations for climate adaptation strategies for tidal marsh restoration. Since we found that the first year following construction appeared to have the least stressful conditions, we recommend managers invest especially heavily in supporting plant colonization through passive and active restoration during this early window of opportunity. We also found plant size and species affected drought tolerance and recommend larger plant sizes and hardy species be incorporated into active tidal marsh restoration. Furthermore, we recommend planning for phased completion of restoration projects to generate a mosaic of areas with different trajectories and increase the probability that some areas will be completed during optimal climate conditions. We thus illustrate how an understanding of climate sensitivity of restoration trajectories can enhance restoration success.
README: Temporal trends in tidal marsh restoration at Hester Marsh, Elkhorn Slough, California
https://doi.org/10.5061/dryad.djh9w0w91
Description of the data and file structure
Below is a summary table of which of the 11 datafiles correspond to which figures and tables in the paper.
Filename | Used for Article |
---|---|
Wasson_Precipitation_Temperature | Fig 3 |
Wasson_Inundation | Fig 3 |
Wasson_Soil_Trajectory | Fig 4 |
Wasson_UAV_Cover | Fig 5 |
Wasson_Seedlings | Fig 6, Table S5 |
Wasson_Facilitation | Fig 7 |
Wasson_Compare_Phase_Multispecies | Table 2, Table S6 |
Wasson_Compare_Phase_Salicornia | FigS1, Table S7 |
Wasson_Compare_Years_Distichlis_Survival | Fig 8 |
Wasson_Compare_Years_Distichlis_Size | FigS2, Table S8 |
Wasson_Compare_Years_Salicornia | FigS3, Table S9 |
Abiotic conditions relevant to plant restoration success
Weather and tides
Filename “Wasson_Precipitation_Temperature.csv”
We measured precipitation and air temperature at 15-minute intervals using the ESNERR weather station. Data, metadata and standard operating procedures can be accessed at www.nerrsdata.org.
Filename “Wasson_Inundation.csv”
We monitored tidal surface water levels using a water level recorder (ESNERR station, Vierra Marsh, data also accessed at www.nerrsdata.org ) installed nearby. We georeferenced both the marsh surface and the water level recorder and calculated the percentage of time the marsh surfaces in Phase I and Phase II were tidally inundated.
Soil properties by phase and year
Filename “Wasson_Soil_Trajectory.csv”
We collected shallow soil cores (0-5 cm depth) within close proximity (~10 cm) of selected recruitment plots. Samples were taken at 6-12 plots (varying by sampling date) in Phase I and 10 plots in Phase II. We collected the cores within the two phases across the three study years, with sample dates on 22 Feb 2022, 23 April 2023, 12 Sept 2023, 8 April 2024, and 25 Sept 2024. We analyzed all samples for salinity and water content immediately after collection. We also analyzed samples from 12 Sept 2023 for bulk density, organic matter and soil texture and used a static cone penetrometer (Humboldt Mfg. Co., USA) in the field to measure surface compaction.
*Passive restoration success: vegetation cover *
Filename “Wasson_UAV_Cover.csv”
We collected UAV imagery to track passive restoration success - colonization of the marsh platform from seeds washed into the restoration area on tides. We quantified vegetation cover change over time in Phase I and II. To compare similar soil conditions across phases, we limited the analysis to an area of interest where sediment removed from adjacent hillsides was placed on former degraded marsh, with no subsequent ripping to decompact it, and no planting. We divided each area of interest into replicate quadrats (5 x 5 m) to calculate differences in colonization rate over time. We identified each pixel of the UAV imagery as either vegetated or unvegetated through supervised pixel-based classification in Google Earth Engine, using a random forest model. We assessed the accuracy of these classifications by comparing the predicted class of 500 randomly selected pixels to the true class, as determined through photo-interpretation. We used the measured error of omission and admission to adjust the percent cover estimates to more accurately track the trend of revegetation over time. Though we did not classify the species identity of pixels, we determined through field reconnaissance that the majority of cover consisted of perennial Salicornia, though the annual Spergularia marina episodically accounted for a portion of the cover (~ 5-10% total cover). Technical details of the processing and classification of UAV imagery are described in Appendix S1 of the article.
*Passive restoration success: wild seedlings *
Filename “Wasson_Seedlings.csv”
To compare passive restoration success across phases and years in the field, we quantified seedling density in permanent quadrats (0.5 x 0.5 m) in 2022-2024. The quadrats were located in portions of Phase I (24 quadrats) and II (20 quadrats) with similar soil histories (where sediment removed from adjacent hillsides was placed on former degraded marsh, with no subsequent ripping to decompact it), and were part of larger fixed plots where we planted greenhouse-grown plants (see “Active restoration” section below).
We counted live seedlings at each sampling date; seedlings were defined as plants <10 cm in height. Virtually all seedlings encountered were either Salicornia pacifica *or *Spergularia marina *so we only report on these species. Since *Spergularia *is an annual, all seedlings germinated in the same year as they were counted. For *Salicornia, a perennial, we could not determine whether seedlings had germinated in the present or a previous year, though generally seedlings grow more than 10 cm in the first year, so most of our counts likely represent seedlings that germinated in the year they were counted.
*Effect of previous year’s vegetation cover on new wild seedlings *
Filename “Wasson_Facilitation.csv”
To determine whether there is facilitation or inhibition of Salicornia *seedlings by established *Salicornia, we conducted a mensurative experiment to track wild seedlings in areas with and without presence of Salicornia *cover from plants established in the previous year. In late May 2023, we established pairs of small adjacent plots with and without existing *Salicornia *cover in Phase II. We chose an area that had been ripped to decompact it, because *Salicornia *seedling densities were notably more abundant in resultant furrows than in nearby compacted, unripped areas, and we sought areas where there would be sufficient sample sizes to determine effects. Paired bare and vegetated plots (0.15 x 0.15 m) were located near (0.4-0.9 m) each other within the same furrow, in visually similar soil conditions. The established *Salicornia *in the vegetated plots *was approximately one year old, from wild recruitment in the first growing season that Phase II was open to tides. We attempted to set up the plot pairs in areas with approximately similar seedling densities, so we could track mortality from a similar starting point. We set up 10 pairs, which averaged 6 m from each other (range 1.5-15 m apart). We counted *Salicornia *seedling density and measured the height of the largest seedling in each plot on 25 May and 14 Dec 2023.
*Active restoration experiments: transplant survival and growth *
Comparing restoration success between phases
Filename “Wasson_Compare_Phase_Multispecies.csv”
2022 experiment with small and medium-sized plants: On 26 Jan 2022, we planted small-sized (grown in 107 ml “conetainer” pots) and medium-sized (grown in 216 ml “Deepot16” pots) Distichlis, *Frankenia *and *Jaumea *into compacted sediment addition areas of Phase I (n=24) and Phase II (n=20), adjacent to the recruitment quadrats described above for quantifying seedling densities. All plants were watered twice in the first two weeks after planting. We monitored survival on 23 Jun 2022; by this date almost all the plants in Phase I were dead, so that is the last date of relevance for this comparison of phases.
Filename “Wasson_Compare_Phase_Salicornia.csv”
2023 experiment with large plants: Since almost all small and medium-sized greenhouse-grown plants had died in Phase I in 2022, and we had no large greenhouse-grown plants available, we tried a new transplant technique for increasing vegetation cover in 2023, using large direct *Salicornia *transplants from nearby creeks in 2023. On 10 Feb 2023, we transplanted large *Salicornia *(approx 3 l root balls) from a creek in Phase I to areas of Phase I (n=12) and II (n=12) near where the experiment with small and medium-sized transplants had been conducted the previous year. They were hand-watered twice in the first two weeks after planting, then not irrigated again, however there were rains at least weekly for the first months after planting. We monitored survival and size on 12 Sept 2023; for size, we measured maximum length, width and height of each plant, and multiplied those together to estimate volume.
Comparing restoration success among years
Filename “Wasson_Compare_Years_Distichlis_Size.csv”
Filename “Wasson_Compare_Years_Distichlis_Survival.csv”
Distichlis *transplants: We planted medium-sized (216 ml) greenhouse-grown *Distichlis into the same 20 plots in a compacted, sediment addition area of Phase II in winter of three consecutive years (26 Jan 2022, 14 Feb 2023, 26 Jan 2024). We measured transplant size in the first summer after planting.We monitored survival of 2022 transplants for 593 days, survival of 2023 transplants for 545 days, and survival of 2024 transplants for 257 days.
Filename “Wasson_Compare_Years_Salicornia.csv”
*Salicornia *transplants: We planted large creek-transplant *Salicornia *into nearby compacted sediment addition areas of Phase II in two years (16 April 2022, n=10, 14 Feb 2023, n=12) using similar methods and initial hand watering for the first weeks. Size and survival were monitored in late summer of each year (8 Aug 2022 and 12 Sept 2023, respectively). We used a gamma GLM model with Year as a fixed effect to look at differences in growth between years.
Files and variables
File: Wasson_Inundation.csv
Description: Data on inundation frequency of the two restoration areas being studied
Variables
- Month-Year: the month and year
- PI_Percent_Time_Inundated: percent of total time that Phase I was inundated during this month
- PII_Percent_Time_Inundated: percent of total time that Phase II was inundated during this month
File: Wasson_Seedlings.csv
Description: Data on naturally recruiting seedlings in permanent plots in the two restoration areas
Variables
- Date: date
- PlotID: identifier of the permanent monitoring plot
- Phase: restoration phase, I or II
- Pickle_Recruit_Density: density per m2 of pickleweed (Salicornia pacifica) seedlings in the plot
- Sperg_Recruit_Density: density per m2 of sand spurry (Spergularia marina) seedlings in the plot
File: Wasson_Compare_Phase_Multispecies.csv
Description: data on transplant success in the two phases in 2022
Variables
- PlotID: identifier of the permanent monitoring plot
- Date: date
- Phase: restoration phase, I or II
- Genus: genus of transplant (Distichlis spicata, Frankenia salina, Jaumea carnosa)
- Size: volume in cm3, generated from measuring and multiplying length x width x height in cm
- Alive: whether the plant was alive at the monitoring date
File: Wasson_Facilitation.csv
Description: data on paired vegetated vs. bare plots designed to determine whether there was facilitation of new seedlings by existing vegetation
Cells with NA had no sizes provided because there were zero seedlings in one of the paired plots, and so no paired comparison of size could be made.
Variables
- Pair: number to identify each pair of plots
- Watered: whether plot was watered or not
- Vegetated_Count_May: count of pickleweed (Salicornia pacifica) seedlings per quadrat (15 x 15 cm) in vegetated plot in May 2023
- Bare_Count_May: count of seedlings (Salicornia pacifica) per quadrat (15 x 15 cm) in bare plot in May 2023
- Vegetated_Count_December: count of pickleweed (Salicornia pacifica) seedlings per quadrat (15 x 15 cm) in vegetated plot in December 2023
- Bare_Count_December: count of seedlings (Salicornia pacifica) per quadrat (15 x 15 cm) in bare plot in December 2023
- Bare_Largest_P_Dec: height of largest pickleweed (Salicornia pacifica) in bare plot in December 2023
- Vegetated_Largest_P_Dec: height of largest pickleweed (Salicornia pacifica) in vegetated plot in December 2023
- Bare_Survival_May_Dec: proportion of pickleweed (Salicornia pacifica) seedlings that survived from May to December 2023 in bare plot
- Vegetated_Survival_May_Dec: proportion of pickleweed (Salicornia pacifica) seedlings that survived from May to December 2023 in vegetated plot
File: Hester_Compare_Phase_Salicornia.csv
Description: data on large transplanted Salicornia pacifica size differences after one growing season in 2022 in two restoration phases
Variables
- Phase: restoration phase, I or II
- PlotID: identifier of the permanent monitoring plot
- P_height: height (cm) of pickleweed (Salicornia pacifica)
- P_length: length (cm) of pickleweed (Salicornia pacifica)
- P_width: width (cm) of pickleweed (Salicornia pacifica)
- Pickleweed_Volume: volume (cm3) of pickleweed (Salicornia pacifica)
File: Wasson_Compare_Years_Distichlis_Size.csv
Description: data comparing size of Distichlis spicata planted into the same plots in three different years (2022-2024)
Cells with NA are plants that were dead at that date, and therefore had no size.
Variables
- Planting_Year: Year in which the Distichlis was planted
- PlotID: identifier of the permanent monitoring plot
- H: height (cm) of Distichlis spicata
- L: length (cm) of Distichlis spicata
- W: width (cm) of Distichlis spicata
- Volume: volume (cm3) of Distichlis spicata
File: Wasson_Compare_Years_Distichlis_Survival.csv
Description: data comparing survival of Distichlis spicata planted into the same plots in three different years (2022-2024)
Variables
- Planting_Year: Year in which the Distichlis was planted
- PlotID: identifier of the permanent monitoring plot
- Censor: 0 if the plant was still alive at least monitoring, 1 if it was dead
- Days_Survived: number of days the plant survived, or if it was still alive, number of days to last monitoring
File: Wasson_Compare_Years_Salicornia.csv
Description: data comparing sizes of Salicornia pacifica transplanted in different years
Variables
- Planting_Year: year in which the transplant occurred
- PlotID: identifier of the permanent monitoring plot
- P_height: height (cm) of pickleweed (Salicornia pacifica)
- P_length: length (cm) of pickleweed (Salicornia pacifica)
- P_width: width (cm) of pickleweed (Salicornia pacifica)
- Pickleweed_Volume: volume (cm3) of pickleweed (Salicornia pacifica)
File: Wasson_Temperature_Precipitation.csv
Description: weather data during the study period, from the ESNERR weather station
Variables
- DateTimeStamp: date and time
- Air_Temperature: temperature in C
- Precipitation: rainfall in mm
File: Wasson_Soil_Trajectory.csv
Description: data on soil conditions over the study period, determined from soil cores analyzed in the laboratory, and field penetrometer readings
Cells with NA indicate data that were not collected at a particular location and date.
Variables
- Plot ID: identifier of the permanent monitoring plot
- Date: date
- Phase: restoration phase, I or II
- Moisture_Percent: moisture as volumetric water content
- Salinity: salinity in ppt
- Bulk_Density: bulk density in grams per cm3
- Organic_Matter_Percent: percent organic matter
- Sand_Percent: percent sand
- Silt_Percent: percent silt
- Clay_Percent: percent clay
- Depth to refusal_cm: depth to refusal as measured in the field with penetrometer
File: Wasson_UAV_Cover.csv
Description:
Variables
- Date: date
- Phase: restoration phase, I or II
- ID: identifier of the 5 x 5 m grid cell into which the landscape was broken for UAV analysis
- Percent_Cover: percent cover of vegetation in the grid cell
Code/software
Files are CSVs and don't require any particular program to view
Access information
Other publicly accessible locations of the data:
- NA
Data was derived from the following sources:
- NA