2004, Engeman et al. in press). Once these injuries occur, further erosion of the carbonate sediment can occur if the topographic damage is great enough to prevent natural recruitment of the seagrass. Seagrass rhizomes grow from the apical meristems, which are not able to grow when there is a topographic difference of 20cm or greater (Kenworthy 2002, Engeman et al. in press); therefore, erosion over time will increase the amount of damage, leaving a much larger injury footprint. Despite the fact that shallow seagrass beds in LKSLMA are delineated by “No Combustion Zone” signs which mark areas of the seagrass flats that are off limits to motorized vessels due to depth, grounding events still occur. Restoration of these seagrass habitats is necessary to prevent further deterioration and to restore critical habitat for floral and faunal species. Herein, we quantify and discuss inprogress restoration techniques applied for damaged seagrass beds at LKSLMA, and our Year 1 preliminary findings.

METHODS 
This restoration project was funded through a grant from the Keys Environmental Restoration Fund and through the Lignumvitae Key Botanical State Park “Help Our State Parks” (HOSP) fund. Protocols for seagrass restoration are well documented and summarized by Fonseca et al (1998) and were applied in LKSLMA. Restoration was conducted at six sites within the park and included topographic restoration, “bird stake” installation, and shoal grass (Halodule wrightii) planting unit installation. Because of variation in the physical damage, prop scar topographic features, and restoration requirements for each site, methods utilized also were site specific.

Phase one consisted of topographic restoration at three of the six restoration sites: “Princess Jullin,” “Curved Scar,” and “Power Cat” sites. Native fill material consisted of 0.635cm pea rock with a top layer of limerock screening. The fill was transported by barge to the area, where it was then loaded into a mobile box that is situated on a smaller vessel capable of accessing the shallow site. Once over the injury site, the box is gradually lifted up so that the fill slides into the injury feature (see Figure 1). Once sufficient fill was placed in the feature, the fill was graded by hand using garden rakes so that it was level with or just slightly above the level of the surrounding seagrass flat. PVC pipes were installed and marked at the level of the fill to measure erosion over time. Fill material was carefully kept away from the adjacent and undamaged seagrass beds.

A biodegradable cotton cloth was secured on top of the limerock screening in two small areas at the “Princess Jullin” and “Curved Scar” sites. The purpose of this was to determine whether a physical barrier placed over the fill

material would provide any protection from erosion since the restoration sites selected experience high velocity water movement across the seagrass flat. Within five months the cloth had degraded at both sites and no erosion was observed either where the cloth had been installed or in the uncovered areas.

The next phase of the project was to install the “bird stakes.” Halodule wrightii, which is the pioneer seagrass species and the first to colonize a disturbed site, responds to a limited increase in nutrient levels (Kenworthy et al. 2000, Hall et al. 2006). Bird stakes encourage roosting by many species of marine birds, particularly double-crested cormorants (Phalacrocorax auritus), terns (Sterna spp.), and gulls (Larus spp.) and the nutrient-rich fecal waste that is deposited by the birds directly over the restoration site enhances the growth of Halodule. Bird stakes were constructed of a 3.03m length PVC pipe with a 5.08cm x 10.16cm wood block attached on its edge to the top of the pipe. This allows for maximum deposition of waste in the water. Stakes were installed so that the wood blocks are 25.4cm above the mean high tide mark for maximum use at all tidal phases.

Bird stakes were installed at all six restoration sites. In prop scars and filled blowholes, they were installed on 2m centers 0.5m from the edge of the undamaged seagrass flat. Stakes were installed the length of single prop scars, but alternated in twin prop scars. At the “Curved Scar” erosion had widened the original injury feature so the bird stakes were installed across the scar on 2m centers staying 0.5m from the edge. “Robbie’s Flat,” the “Stake Array,” and the “T-Array,” sites consisted of numerous prop scars over a large area of the seagrass flat. At these sites, bird stakes were installed as stake arrays on 3m centers 2m from the edge of the unscarred seagrass flat (see Figure 2).  

Phase three of the project consisted of the installation of Halodule planting units at the “Curved Scar” and the “Princess Jullin” sites. Halodule was harvested in two trips from one donor site within LVKSLMA approximately ten minutes from the restoration sites. Halodule was harvested by hand, placed in buckets of water to minimize stress to the grass, and transported to the restoration sites. Individual  

Continued on page 4