Seagrass meadows form key ecosystems that can be found growing in relatively sheltered coastal environments throughout temperate and tropical waters. In recent decades seagrass loss has become an increasing problem worldwide, owed in large to the desire for humans to live near the coast where we create extensive infrastructure and modifications to the marine environment, including discharging pollution directly into marine waters (Shepherd et al. 1989; Walker & McComb 1992; Short & Wyllie-Echeverria 1996). One of the unfortunate consequences is the loss of seagrass meadows and associated communities of plants and animals. Seagrasses, like terrestrial grasses, are well known for their ability to consolidate and bind sediments (Fonseca et al. 1985; Walker & Woelkerling 1988; Fonseca 1996; Koch 2001), consequently their loss destabilizes the seafloor leading to sediment erosion and usually high turbidity associated with the resuspension of fine sediments, contributing to overall poor water quality perpetuating further seagrass loss (Larkum & West 1983; Seddon et al. 2003). In cases where the sources of seagrass loss have been removed or ameliorated, regeneration is usually notoriously slow (e.g. Kirkman 1989; Meehan & West 2000). Sometimes the loss of seagrass can lead to irreversible changes in the nature of the environment and habitat, rendering the site no longer suitable for seagrass survival (e.g. West et al. 1990). One of the most significant meadow-forming seagrasses in Australia is Posidonia (commonly known as ribbon, strap or tape weed). There is considerable evidence that Posidonia meadows, and similar slow-growing seagrasses elsewhere in the world, can take decades to recover after a major disturbance (Kirkman 1997; Meehan & West 2000). In contrast, seagrasses with faster-growing rhizomes, such as Heterozostera (Table 1), have been known to colonize bare areas within 6 months (Kirkman 1989). Even so, if the disturbance is large enough, even recovery by fast-growing species may take several decades (Olesen & Sand-Jensen 1994; Orth et al. 1994). The increasing loss of seagrass habitat and its slow regeneration time are issues that have concerned people for a long time and has provided the impetus for marine researchers to direct their attention and efforts toward developing methodologies for seagrass restoration and rehabilitation, or in some cases, meadow creation (Fig. 1, Box 1). The restoration and rehabilitation of seagrass meadows not only involves the development of transplanting techniques but also requires diving skills. Divers here are coring seagrass from a Posidonia meadow at 6 m depth. At this relatively shallow depth a tank of air will last around 75 min, but each diver is limited to a maximum of 4 dives per day. Photo by Sonja Venema (SARDI Aquatic Sciences). Restoration underwater, and of seagrasses in particular, is a relatively recent endeavour when viewed against developments in habitat restoration in terrestrial environments (Box 2). Some early pioneering attempts at seagrass restoration in the United States were made by Addy (1947; cited Fonseca et al. 1998) and Philips (1960; cited Fonseca et al. 1998) who reported varying degrees of success. Over the last 20–30 years, however, there has been intense activity and developments in this area, but unfortunately a large proportion of these seagrass restoration attempts have only resulted in limited transplant survival and coverage. For instance, of the 53 restoration projects Fonseca et al. (1998) reviewed, only 5% of the plantings of sprigs or cores had 100% survival, whereas, the mean planting unit (PU) survival for these studies was 42%, so that many of these projects would require replanting to attain the desired coverage. In Australia, comparatively less effort has been directed toward studies on seagrass restoration, despite the significant declines in seagrass coverage that have occurred in Australian waters over recent decades (Walker & McComb 1992). The majority of the research has occurred over the last 10–15 years, with most of this activity centred in Western Australia – the location of highest seagrass diversity in the world (Walker 2003) and so a hotspot for seagrass research in general, although some early work on restoration methodologies was initiated by Larkum (1976) in Botany Bay (NSW). A comprehensive review by Lord et al. (1999) found that there had only been a total of 21 restoration projects in Australia in the 20 years prior to 2000, with over 60% based in Western Australia. As found in the United States and Europe, success has been limited, with by far the majority of failures in Australia attributed to high wave action, particularly as a result of storm activity (e.g. West et al. 1990; Paling 1995; Paling et al. 2001a). Many different transplanting and planting methodologies have been developed and tested over the years (Box 3 and 1, 2), with varying degrees of success depending on factors such as species, sediment type and current speeds. Success with traditional techniques is highly variable, as discussed earlier, but even when high PU survival and good coverage are reported, there are limits to the area that can be realistically restored or rehabilitated using these techniques. Coring seagrass is a common technique used to collect donor material for transplanting to the rehabilitation site. This diver has just extracted a plug of seagrass (Amphibolis antarctica) using a 10-cm PVC core, from a donor meadow off Henley Beach in Adelaide, South Australia. Photo by David Miller (SARDI Aquatic Sciences). Getting enough areal coverage is a major problem. While acknowledging that failures are common, Lewis (1987) concluded that there have been successful restoration and meadow creation projects of up to 6 ha in size. This is a huge area of seagrass for divers to transplant. However, even this falls well short of the area required to rehabilitate many of the sites of significant seagrass loss in Australia and elsewhere. Recently a team in Western Australia lead by Eric Paling developed mechanical techniques to enable large sods of seagrass to be extracted and transplanted. These sods have a much greater mass than cores and, therefore, better anchorage in high wave energy environments compared with cores and sprigs. In 1996 Ocean Industries Pty., on behalf of Cockburn Cement Ltd, constructed ECOSUB1 – a purpose built, fully submergible machine capable of transplanting large sods measuring approximately a quarter of a m2, designed as part of an operation to salvage seagrass removed as a result of a sand mining operation (Paling et al. 2001a). Later in 2000, ECOSUB2 (Fig. 3 and Box 3) went into operation with a significantly improved design enabling sods of twice the area (and considerably greater mass) of ECOSUB1 to be extracted, plus the addition of an extra storage unit (shuttle), greatly increased the number of planting units that could be transplanted per day from about nine sods, to an estimated 75 sods. Average long-term survival of the mechanically transplanted sods, around 70% over 3 years for ECOSUB1, has proven to be much higher using this technique than other transplant methods (cores and sprigs) used in similar locations (Lord et al. 1999; Paling et al. 2001b). The need to salvage large areas of seagrass as a result of sand mining operations in Western Australia led to the development of submergible machines (such as ECOSUB1 and 2) capable of transplanting large sods of seagrass. The main advantage of large sods is that they have a greater mass and a better rate of survival in high wave energy environments compared with those transplanted by hand. Clockwise from top left: (a) ECOSUB2 and accompanying sod shuttle prior to launching (photo reproduced with permission of Cockburn Cement Limited, Western Australia); (b) a diver storing a seagrass sod in the ECOSUB2 sod shuttle at Success Bank in Western Australia; and (c) commercial divers operating ECOSUB2 underwater (photos courtesy of Brian Richards, Murdoch University). Even still, mechanical transplantation by ECOSUB2 at the maximum predicted rate of 75 sods per day – approximately 40 m2 of 100% seagrass cover (Paling et al. 2001b), which could potentially cover 227 m2 per day at the planting site (assuming 1 m between sods), would still take 44 days to plant just 1 ha. It is clear from these studies and many more documented in the grey literature, that it is possible to successfully regenerate small-scale seagrass loss (several hectares at most). However, many of the problem locations in Australia, and indeed worldwide, have lost hundreds and even thousands of hectares of seagrass habitat (Walker & McComb 1992; Short & Wyllie-Echeverria 1996). What new techniques might be developed to address these situations? To add a further complication, there has been considerable concern about the damage caused by traditional techniques that depend upon donor meadows (e.g. cores and sprigs) and for this reason, many believe that these vegetative techniques should not be used for large-scale projects (especially for slow-growing seagrasses such as Posidonia or Thalassia). As Lewis (1987) points out, 'the future of seagrass restoration and creation efforts will depend increasingly on the availability of plants obtained without damage to existing healthy seagrass meadows . . . or through salvage efforts'. (p. 154) Evidently there is a need for inexpensive methodologies, in terms of time, labour and materials, in order to replant large areas of seagrass loss. One relatively new area of focus in Australia involves the use of seagrass seeds and seedlings to rehabilitate an area, sometimes referred to as donor meadow independent methods (Box 3 and Fig. 4), because they do not require mature seagrass to be extracted from healthy meadows – a major consideration in a region where seagrass is scarce or significant areas have been lost. Developing techniques for the successful growth of seagrass seedlings on a large-scale is a relatively new and ongoing endeavour. This year-old Posidonia seedling is part of a trail where seedlings were grown in biodegradable Jiffy pots at the South Australian Aquatic Sciences Centre (from Seddon et al. 2004b). Might it be possible to achieve restoration of larger areas, and reduce impacts, by broadcasting seed or planting seedlings, thereby relying more heavily on facilitating natural regeneration? Like terrestrial grasses, seagrasses have two forms of propagation; sexual (i.e. seeds) and vegetative (i.e. rhizome extension). Many seagrasses are prolific annual fruiters, but there is evidence that some species are biannual or perhaps even sporadic fruiters depending on the location, and possibly also environment factors that we are yet to determine (Kirkman 1998; Seddon et al. 2004b). Regardless of mode of sexual reproduction (e.g. release of fruits, viviparity), most species spread as a result of vegetative methods rather than by recruitment of seedlings (Kuo & Kirkman 1996). For seagrasses that do produce seeds, there are two significant processes that influence whether an area of seagrass loss is able to regenerate naturally and where recruitment facilitation could play an important role; fruiting and recruitment. For instance, circumstances can occur where there is an overall lack of recruitment success owing to extremely low (or absent) fruiting for a given year. This could happen when there is very little of the original meadow left, or perhaps existing meadows are under metabolic stress and unable to sustain the energy reserves required for fruiting. Second, circumstances can occur where fruiting occurs in high numbers but there is a lack of recruitment success. This could happen if seedlings are unable to attach or remain attached as a result of storms and strong currents, or if the seafloor is unconsolidated, or in some cases if the seeds/seedlings just don't make it to the location of seagrass loss. Tidal and wind driven currents may represent a barrier to recruitment if the prevailing currents at the time of fruiting simply do not carry the propagules to areas of seagrass loss (e.g. Box 4). Furthermore, even if there has been good recruitment, seedlings may not survive due to factors such as poor water quality and associated factors such as smothering and shading due to excessive algal growth. Assuming that poor water quality is not an issue, an important consideration for all restoration programs (e.g. Fonseca et al. 1998; Meehan & West 2002), one of the simplest ways to maximize the success of any given recruitment event may be to collect the fruits and disperse the seed or seedlings at desired locations to enhance recruitment. For example in Chesapeake Bay (USA) good results have been achieved by broadcasting eelgrass (Zostera marina) seeds over large areas. In this case around 15% of viable seeds became established, improving to 41–56% when protective burlap bags were used (Orth et al. 1994; Harwell & Orth 1999). There are a couple of prerequisites that make the seed broadcast method possible with regard to large-scale seagrass rehabilitation in the US. First, Z. marina is a prolific annual fruiter and, therefore, a reliable source of abundant propagules (Granger et al. 2000), and second, the prevailing hydrodynamic conditions in Chesapeake Bay are relatively calm, that is, low swell. Although Orth et al. (1994) state that the current speed is 'quite high', up to 1.2 cm s−1 (0.5 m s−1 is considered too strong for most restoration attempts; Fonseca et al. 1998), they postulate that this is not a problem because the small-scale topography shields the tiny seeds from strong currents. In Australia, there are several relatively small, fast-growing endemic species (e.g. Z. mucronata, Z. muelleri and Heterozostera tasmanica) that are likely to be amenable to reseeding via the broadcast method. Unfortunately the second prerequisite rules out most sites in temperate Australia as candidates for the seed broadcast method, as most sites are subject to high current speeds and more importantly, strong swell. Recruitment facilitation is a method researchers are starting to seriously consider with a view to large-scale seagrass rehabilitation in Metropolitan Adelaide (South Australia), where over 5200 ha of seagrass has been lost over the last 50 years (Seddon 2002). In particular, those of us involved with seagrass rehabilitation in South Australia believe that we should be working with natural life history processes, such as reproduction and recruitment, by enhancing the success of recruitment events (through the capture of seedlings) and facilitating seedling survival and subsequent spread via lateral expansion. To this end, we recently started investigating methods of trapping seedlings using biodegradable hessian matting (a material previously used in dune revegetation projects and a good substitute for remnant seagrass fibre mat) and encouraging the seedlings to attach and grow (Fig. 5). Preliminary results show that Amphibolis seedlings do effectively attach to the hessian strips, with initial numbers up to 18 seedlings per linear meter (Seddon et al. 2004b). A diver adding a final layer of sediment along a biodegradable strip of hessian matting which is partially buried and secured into the seafloor with steel pegs, leaving an edge free and exposed to initially trap seedlings (particularly Amphibolis) drifting past with the current. Once the seedlings are captured, the matting provides a point of anchorage for the seedlings to attach their roots to in what would otherwise be a bare substrate (from Seddon et al. 2004b). Recolonization and succession are both key processes driving seagrass dynamics, and in this respect, manipulations of these processes could be utilized to augment or accelerate recovery. Like any habitat dominated by a long-lived plant community, seagrasses also tend to form vast meadows of slow growing, high biomass species; climax species. In the situation where a major disturbance has removed the climax species, the resulting bare substrate is usually colonized by fast growing, low biomass, colonizing species (e.g. Box 4). These species quickly cover space, stabilize sediments, and create a microclimate more conducive for the survival of the seedlings of climax species. Much of the recent thinking in seagrass rehabilitation relates to methodologies to 'fast-track' the natural process of succession – also referred to as 'compressed succession' (Fonseca et al. 2000), by encouraging the growth of colonizing species, then in-planting with climax species either collected as seedlings or cultured seedlings. Certainly results by van Keulen et al. (2003) indicate that cores of Posidonia fair much better when planted into a bed of Heterozostera compared with bare sand or even bare sand with mesh protection. Nevertheless, using colonizing species to cover ground, then in-planting with seedlings, still remains largely untested in the context restoration or rehabilitation methodologies. In South Australia we have started a program of collecting the fruits of the climax species Posidonia spp., germinating the seeds and culturing the subsequent seedlings for future planing into areas of seagrass loss (Fig. 5). The advantages of this technique are that seedlings can be held in culture until the conditions are optimal for planting (usually spring, when the photoperiod is increasing and the winter storms are over), thereby maximizing chances of success. Previous results with aquarium grown Posidonia oceanica seedlings by Balestri et al. (1998) in the Mediterranean are very encouraging as approximately 70% of the cultivated seedlings survived for 3 years when planted-out, compared with 66% survival over the same period for naturally recruited seedlings, whereas survival of surf grass (Phyllospadix torreyi) seedlings planted-out in California (USA) was much lower, but still approximated naturally recruited seedlings at 30% (Holbrook et al. 2002). It is by exploring and testing these relatively new methods of recruitment facilitation and fast-tracking succession that we are hoping to develop cost-effective tools to restore and rehabilitate large-scale seagrass loss that has otherwise previously been intractable both logistically and financially. Its much easier to find a way to go with the flow than fighting against it! The seagrass rehabilitation research in South Australia discussed in this paper is a joint collaborative project between SARDI Aquatic Sciences and the Coastal Protection Branch, Department for Environment and Heritage. Many thanks to the SA rehabilitation team, especially David Miller and Sonja Venema for all their efforts, and to Doug Fothringham for his enthusiasm and support of all things seagrass. I would also like express my appreciation to my colleagues Mike van Keulen, Eric Paling and Marion Cambridge from Western Australia and Bob Orth and Mark Fonseca from the United States, whose expertise, advice and encouragement contributed to the formulation the ideas expressed in this paper. Finally, thankyou for comments and editing suggestions from two anonymous reviewers.
No takes yet. Share an insight, caveat, or question.
Stephanie Seddon (2004) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: