The mosaic of patterns observed in seagrass landscapes is often a result of natural perturbations such as erosion and burial by sand waves (Harlin and Thorne-Miller, 1982; Fonseca et al., 1983; Marba et al., 1994; Marba and Duarte, 1995; Fonseca and Bell, 1998; Bell et al., 1999; Fig. 12) and/or disturbances caused by fauna (Orth, 1975; Ogden, 1980; Preen, 1995), storms (Preen etal., 1995; Fonseca and Bell, 1998) and/or disease (den Hartog, 1987). Anthropogenic causes (eutrophication, boat and mooring scars, fishing gear scars) can also contribute to seagrass patchiness (Cambridge, 1975; Walker etal., 1989; Creed and Filho, 1999; Orth etal., 2002). Here we will focus on flow-related generation of patchi-ness.
The disturbances that lead to patchiness in seagrass landscapes range in scale from the complete but local destruction of seagrass ecosystems (Dan et al., 1998) to the creation of smaller (10's of meters) unvegetated depressions in continuous meadows, termed "blowouts" (Patriquin, 1975; see review in Short and Wyllie-Echeverria, 1996; Fonseca and Bell, 1998). The magnitude and frequency of disturbances is what determines the degree of patchiness of a seagrass meadow (Fonseca et al., 1983; Fonseca and Bell, 1998; Hemminga and Duarte, 2000). It can therefore be assumed that more extreme events would result in erosion and complete or partial loss of seagrasses (with little recovery), whereas periods of reduced disturbance may result in coalescence of the patches and the formation of more continuous meadows (Fonseca et al., 1983; Kirkman and Kirkman, 2000). This concurs with numerous reports of widespread loss of meadows as a result of hurricanes and cyclones (Birch and Birch, 1984; Williams, 1988; Rodriguez et al., 1994; Preen et al., 1995; Moncreiff et al., 1999; Whitfield et al., 2002), and the extensive colonization of seagrasses (pioneering or climax species) during periods of reduced disturbance (Kirkman, 1985; Kendrick et al., 2000). Note that some studies report no damage to seagrasses after the passage of hurricanes (Thomas etal., 1961; Tilmantetal., 1994; Dawes etal., 1995).
In areas with continuous high wave energy, sea-grass ecosystems can be: (i) non-existent (Dan et al.,
1998); (ii) depth restricted (when sufficient light is available, seagrasses colonize areas below the maximum wave penetration depth; Krause-Jensen et al., 2003; Middelboe et al., 2003); (iii) dominated by more robust species (e.g. Amphibolis griffithii and Posidonia coriacea); and (iv) patchier as the disturbance of high waves may hinder the lateral expansion of some seagrass beds (Kendrick et al., 2000; Frederiksen et al., 2004). In contrast, in sheltered waters, seagrass meadows tend to be more continuous and are colonized by relatively more fragile species (e.g. some Posidonia spp) (Kirkman and Kuo, 1990). Under calm conditions, creation of openings in meadows appears not to lead to further wide-scale loss, although regrowth into the damaged areas can be slow (Walker et al., 1989; Meehan and West, 2000; Orth et al., 2002). The degree of wave exposure can be quantified by applying the relative wave exposure index (REI) first developed by Keddy. This index takes into account the wind direction and intensity and fetch and has been successfully linked to landscape features in seagrass habitats (Fonseca and Bell, 1998; Fonseca et al., 2000, 2002; Hovel et al., 2002; Krause-Jensen et al., 2003; Frederiksen et al., 2004). When wave-dissipating structures (e.g. sand bars, sills, coral or oyster reefs) occur in the seagrass system, the bathymetry may also have to be taken into account in order to properly estimate the wave exposure using the REI.
A mixture of unvegetated and densely vegetated areas in close proximity may characterize intermediate disturbance regimes. For example, Posidonia sinuosa meadows in habitats of moderate water flow are characterized by dense rows of plants interchanged with strips of bare sand (Bridgwood, 2002). A similar gradient of patterns has been reported for meadows of Zostera marina subject to gradients in velocity and wave energy (Fig. 12), with the proposal that the structural integrity of the habitat would deteriorate with increasing habitat fragmentation (den Har-tog, 1971; Fonseca et al., 1983; Fonseca and Bell, 1998).
Another common cause for complex seagrass mosaic patterns is the hydrodynamically-mediated movement of sand waves and sand dunes through seagrass meadows (Marba et al., 1994; Walker et al., 1996; Bridgwood, 2002; Paling et al., 2003; van Keulen and Borowitzka, 2003; Frederiksen et al., 2004). Changes in sediment height, a result ofwater flow, can be significant andrapid (10's of cm over periods of hours) (Paling et al., 2003), and larger sand dunes can travel through meadows on a time scale of months (Walker et al., 1996; Bridgwood, 2002; van Keulen and Borowitzka, 2003). The degree to which these large amounts of sediment negatively affect the seagrasses creating unvegetated patches depends on their tolerance for sedimentation, the amount of sediment deposited, and the period the plants remain buried. Sedimentation rates of 2 to 13 cm year-1 can be coped with by large (e.g. En-halus acoroides) as well as by fast growing (e.g. Halophila) seagrasses as well as by plants with vertical stem elongation (e.g. Cymodocea nodosa, C. serrulata and Syringodium isoetifolium) (Vermaat et al., 1996). Subaqueous dune migration appears to maintain Cymodocea meadows in a continuous state of colonization which is ultimately responsible for the characteristic patchy landscape (Marba and Duarte, 1995). Even when sediments completely cover the leaves of small, slow-growing plants, some seagrasses are able to survive as long as the sediment is removed by currents or waves in a matter of weeks (Halodule wrightii survived after being buried for 2 months; Phillips, 1980). Other seagrasses such as Posidonia oceanica are able to increase vertical growth from 5 to 7 mm year-1 to 52 mm year-1 when (partially) covered by sand waves (Boudouresque and de Grissac, 1983). In contrast, Zostera marina seems to have little or no tolerance to sedimentation regardless of the sediment type (Mills and Fonseca, 2003).
Was this article helpful?