USA: Texas: West Galveston Bay Seagrass Restoration

Snake Island restoration; Credit: Galveston Bay Foundation


In spring 1994, 2.5 acres (1 hectare) of Halodule wrightii (shoal grass) were planted at two sites in West Galveston Bay using donor plants from eastern and western Matagorda Bay. This project was intended as a pilot project to guide the eventual restoration of over 567 hectares of submerged aquatic vegetation in Galveston Bay under the Galveston Bay Comprehensive Conservation and Management Plan. As such, it examined the effects of planting depth, planting density, and fertilizer application on the survival of transplanted Halodule. While one of the planting sites–Snake Island Cove–experienced a complete die-off within one year, the other–Redfish Cove–exhibited high rates of success (especially in shallower areas of the site), and has since expanded to cover four additional acres through natural propagation. In fact, this site was used in 1999 as a donor site for another planting project in West Galveston Bay.

Quick Facts

Project Location:
West Galvestone Bay, 29.2506859, -94.9610649

Geographic Region:
North America

Country or Territory:
United States of America


Coral Reef, Seagrass & Shellfish Beds

Area being restored:
2.5 acres

Organization Type:
Governmental Body


Project Stage:

Start Date:

End Date:

Primary Causes of Degradation

Urbanization, Transportation & Industry

Degradation Description

In the western portion of the Galveston Bay estuary, seagrass acreage declined from 890 hectares (2200 acres) in 1956 to zero by 1989. Most of these seagrass meadows (primarily shoal grass, Halodule wrightii) grew along the barrier island edges of western West Bay. The only remaining seagrass beds (about 36 hectares or 89 acres) still in existence are found in Christmas Bay, a semi-isolated embayment adjoining West Bay. Seagrass loss has been attributed primarily to direct and indirect effects of dredging activities, wastewater discharges, increased turbidity, increased wave action after bulkheading, and unidentified point-source pollution.

Reference Ecosystem Description

The West Galveston Bay segment is a moderate-salinity (mesohaline) area, and true seagrasses–consisting of predominately shoal grass (Halodule wrightii)–formerly occurred in this part of the bay system, both along the back side of Galveston Island and on the northern shoreline along the mainland. Based on historical photoanalysis, about 458 ha (1,132 acres) of shoal grass were estimated along the back side of Galveston Island during 1956, and 125 ha (308 acres) estimated during 1965 (Pulich and White, 1991). In 1975 photographs, only 37 ha (91 acres) of shoal grass beds were visible in the 3.6-km (2-mi) shoreline section just north of San Luis Pass. Wigeon grass had always been found mixed in with these shoal grass beds and continued to be found sporadically in some shoreline areas of Galveston Island after 1979.

Christmas and Drum Bays contain well-developed, polyhaline seagrass communities, including predominately shoal grass and star grass, and some turtle grass. This area represents the northernmost location for turtle grass on the Texas coast and is somewhat of a distribution anomaly because the closest known population is far to the south in Aransas Bay near Rockport.

Since the late 1950s, seagrass vegetation in upper Galveston and Trinity Bays (excluding the Trinity River delta) has consisted entirely of wigeon grass beds. There are anecdotal reports that shoal grass occurred there in the first half of the 1900s, but these reports are unconfirmed. In West Galveston Bay, shoal grass has historically been the predominant seagrass. In certain seasons, however, particularly in spring, localized grassbeds will contain some wigeon grass. As mentioned previously, a few small patches of turtle grass also occurred at the lower end of Galveston Island near San Luis Pass up until the late 1970s. These patches were interspersed in the large beds dominated by shoal grass.

Project Goals

The goal of this project was to determine whether small-scale restorations in West Bay resulted in viable Halodule habitat, and to thereby determine the feasibility of the proposed large-scale restoration of Halodule beds in Galveston Bay under the Galveston Bay Comprehensive Conservation and Management Plan.


The project does not have a monitoring plan.


This was a pilot project intended to assess the feasibility of large-scale seagrass restoration in Galveston Bay, as proposed by the Galveston Bay Estuary Program. As such, its primary stakeholders are the Estuary Program’s partners and member organizations, which include local, state, and federal governments, business, industry, universities, conservation organizations, bay user groups, and private citizens.

Description of Project Activities:
Restoration was attempted at two sites on western Galveston Island. One bed was planted at Redfish Cove (29º06'15.9"N, 95º06'30.9"W), and two beds (East and West) were planted at Snake Island Cove (29º09'21.4"N, 95º01'58.5"W). Halodule transplanting units (TPUs) were collected from donor beds in Matagorda Bay near the towns of Palacios and Port O'Connor, Texas, approximately 15 km apart and 150 km southwest of West Bay. A total of 3500 TPUs was removed from the Palacios bed and transported to Redfish Cove during 26-28 April 1994. Another 7100 TPUs were removed from the Port O'Connor bed and transported to Snake Island Cove during 3-5 and 10-12 May 1994. TPUs were prepared by the peat pot method developed by Fonseca (1994). A 7.5cm-diameter circular sod plugger was used to extract Halodule from donor beds. Each plug was then inserted directly into a 7.5cm-diameter peat pot. Time-release fertilizer (Osmocote: 14% N, 14% P, 14% K; 5.25g +/- 0.14 SE, n = 10) was added to all but 1200 TPUs from Port O'Connor prior to insertion of Halodule plugs. TPUs were placed in seawater-filled holding tanks and covered with wet burlap to prevent desiccation during transport. Once at the transplant site, the sod plugger was used to create holes into which TPUs were placed, after the sides of the peat pots were ripped to allow rhizome propagation (spread from the original TPU). Individual transplant beds ecompassed 62m x 42m and were arrayed perpendicular to the shoreline at the low-water mark observed after passage of a strong cold front in December 1993. The perimeter of each bed was marked by galvanized fence posts every 2m and was surrounded by 1.2m- high black plastic screen with 10mm x 16mm mesh during the first week of April 1994. This mesh size was expected to exclude all but the smallest of the local fishes and decapods, several species of which could affect transplant success through disturbance or herbivory (Fonseca 1994; Hammerstrom et al. 1998). After fencing, each bed was swept from the seaward edge to the landward access gate with a 75m-wide, 10mm mesh seine to force large organisms out; then the gate was closed. Snake Island East and West were constructed 20m apart and 9km northeast of Redfish Cove. Each bed received TPUs arranged in three 8m x 54m strips. Within each strip, we delineated areas representing combinations of three planting densities (high = 0.25m centers; medium = 0.5m centers; low = 1.0 m centers) and three relative water depth ranges (shallow, medium, and deep). Within each strip, TPUs were planted along marked guidelines as follows: 40 rows at 9 TPUs per row on 1.0 m centers (360 TPUs); 24 rows at 17 TPUs per row on 0.5 m centers (408 TPUs); and 12 rows at 33 TPUs per row on 0.25m centers (396 TPUs). Thus, 3492 TPUs were placed into each bed and 10,476 TPUs were placed overall. A fourth 8m x 54m strip within each bed was left bare as a control to monitor natural recruitment. Transplanted and natural grassbeds were examined monthly during June-October 1994 and in April, June, and September 1995. Halodule survival and propagation were determined by snorkeling over randomly chosen rows of TPUs at each site. TPUs were scored as live (presence of green leaves) or dead (no leaves apparent), and live TPUs were scored as propagating (rhizomes extending beyond the original TPU) or dormant (no rhizomes outside of the original TPU). In addition, we swam transects across control sections to determine whether these areas remained nonvegetated.

Ecological Outcomes Achieved

Eliminate existing threats to the ecosystem:
Coverage in restored beds was significantly lower than that of Christmas Bay--the natural control site--except in September 1994 when the coverage at Redfish Cove equaled that of the natural bed. Moreover, coverage was significantly higher at Redfish Cove than at either Snake Island bed in all months except June 1994. No recruitment of Halodule into the control areas at either site was observed at any time during the initial monitoring period. However, More than 75% of all TPUs showed signs of rhizome propagation outside the peat pots by the end of the first growing season. Production of new shoots (i.e. those located outside the original TPUs) was significantly related to site, planting density, and water depth. New shoot densities (number per unit area) at the transplant sites were significantly lower than those in natural seagrass beds, except in September 1994 when Redfish Cove densities equaled those at Christmas Bay East. New-shoot densities at Redfish Cove were also significantly higher than those at Snake Island East and West in September 1994 and in June and September 1995. High-density plantings produced significantly higher numbers of new shoots than did medium- and low-density plantings in all but the first and last months. Shallow plantings produced higher new-shoot densities than did medium and deep plantings, significantly so in September and October 1994 and June 1995. Survival was significantly higher at Redfish Cove in late 1994 and in 1995 than at either Snake Island bed. Survival at Redfish Cove was 70% by September 1995, while almost all transplants at Snake Island Cove had died by then. The deterioration at Snake Island was a two-stage process that began at the end of the growing season in 1994, when average survival declined from 84% in August to 56% in September and October. Survival stabilized at that level through June 1995, but then the decline increased through September 1995, resulting in a complete loss at Snake Island East and only 6% survival at Snake Island West. We estimate that the total area covered by Halodule at Redfish Cove, including surviving TPUs, was 1014 m² as of June 1996. Halodule coverage ranged from solid to patchy, with isolated patches as small as 2.3 m². Site maps indicate a loss of deeper water areas but stability and expansion of the shallowest region. In fact, the Redfish Cove site has now become permanently established, and a 0.4-ha (1-acre) initial bed has expanded to cover an area of about 2 ha (5 acres). This bed was used in 1999 as a donor site and seemed to recover most of its shallow vegetated area by the 2000 survey. Seagrass shoot densities remain below those of Christmas Bay, but root and rhizome biomass has been similar to that of Christmas Bay in all years. During the 1994-1997 monitoring period, densities of fishes, shrimps, crabs, and benthos (e.g. worms, clams, and small crustaceans) in the planted bed were below densities found in natural seagrass of Christmas Bay but above those observed in adjacent unplanted sands. Since 1998, volunteer beds have begun growing west of the experimental site and may be a result of propagules breaking off of the planted beds.

Factors limiting recovery of the ecosystem:
Halodule survival, coverage, and new shoot densities were affected by site, planting density, and water depth. Planted beds at Redfish Cove exhibited favorable rates of success, while those at Snake Island Cove failed completely. Analysis of various factors influencing the dynamics of this die-off at Snake Island suggests that water quality and sedimentation were the primary factors, and that water column light conditions, salinity, and eutrophic conditions were less influential (Sheridan and others, 1998). Planting density also proved of considerable importance to the ultimate success of the planted beds. In September and October 1994 and in April 1995, survival of high-density transplants was significantly greater than that of medium- and low-density plantings; and in July and September 1995, survival of both high- and medium-density plantings significantly exceeded that of low-density beds. Indeed, success rates were significantly higher when planted on 0.25-m or 0.5-m centers rather than on 1.0-m centers, and propagation (spreading from transplant units) was also greater from 0.25-m or 0.5-m center plantings. Finally, water depth was of considerable import in the long-term viability of newly planted beds. Survival in shallow sections often significantly exceeded that in middle depths, and was always significantly greater than in deep sections.

Socio-Economic & Community Outcomes Achieved

Economic vitality and local livelihoods:
Successful restoration of seagrass beds will increase habitat for species of commercial and recreational importance such as penaeid shrimp, blue crab, and spotted sea trout, as well as their prey, and it will also stabilize shorelines and slow erosion.

Key Lessons Learned

There are several recommendations to be made concerning transplanting seagrasses into the Galveston Bay Estuary. Given that we do not know much about the physical, chemical, and biological status of specific restoration sites, it would benefit resource agencies to conduct various screening tests (including the planting of small experimental beds) prior to full-scale restoration at any given site in West Bay. These tests should help avoid the catastrophic failure we found in Snake Island Cove. Reconnaissance of potential chemical contaminants such as biocides or hydrocarbons in sediments and water along the remaining undeveloped shoreline needs to be conducted, as do assessments of sediment nutrient chemistry with reference to possible nutrient limitation. Current velocities should be examined during times of maximum tidal ranges, particularly if widely spaced plantings (e.g. 1.0m centers) are anticipated. An array of light meters should be deployed at each site to estimate maximum depth limits for transplants. Once a potential planting site passes these screening criteria, test plantings on the order of 100 m² beds should be installed and monitored for at least two growing seasons.

Our experiments and those of Fonseca et al. (1994, 1996a) and Hammerstrom et al. (1998) indicate that, once final sites are delimited, transplanting should be conducted during spring months on 0.25m or 0.5 m centers, in relatively shallow water, by the peat pot method, and including time-release fertilizer. These techniques have proven to be cost-effective and should foster the rapid growth and coalescence of seagrasses and thus long-term resistance to physical and biotic disturbances in West Bay.

Long-Term Management

The results of this project will be used by the Texas Natural Resources Conservation Commission’s Galveston Bay Program to plan and conduct further seagrass restoration actions, with a goal of restoring 567 hectares (1400 acres) of submerged aquatic vegetation in 10 years.

Sources and Amounts of Funding

$75,000 USD The Environmental Protection Agency’s Near Coastal Waters Program provided funding for this project. Other participating agencies and institutions donated a total of 200 volunteer hours to assist with project activities, a contribution representing approximately $25,000 in in-kind support.

Other Resources

Galveston Bay Estuary Program

Galveston Bay Foundation

Primary Contact

Organizational Contact