4. Discussion
4.1. Impact of SLR on reef top wave dynamics
The large magnitude of the hydrodynamic changes observed from the present modeling are likely to have considerable impacts on biogeochemical fluxes in a wide range of processes acting upon corals and algae. These changes to the reef wave dynamics can clearly impact on coral and algal health through changes in production (Jokiel, 1978 and Tribble et al., 1994), (Carpenter and Williams, 1993 and Carpenter and Williams, 2007), through increased mixing (Skirving et al., 2006 and Wangpraseurt et al., 2012) and by imposing damaging wave forces (Massel and Done, 1993 and Madin and Connolly, 2006). Similarly, changes in nearshore wave height in the lagoon due to SLR may strongly influence near shore processes and wave energy reaching nearshore corals, seagrass, mangroves and beaches (Sheppard et al., 2005). The bathymetry of an individual reef has a very strong influence on the wave height over the reef flat, at the leeward reef edge and in the lagoon for present sea level (see Fig. 3). However, this influence weakens for average wave conditions with increasing SLR, i.e. the energy dissipation induced by wide barrier reefs at present SLR is reduced, consistent with Sheppard et al. (2005). Similar effects occur on fringing reefs (e.g. Grady et al., 2013). The greatest influence of SLR occurs on shallow reefs, and these systems may already be under stress because of increases in water temperature and reductions in water quality. SLR induced changes in wave height are very significant. For example, for a reef depth of 1 m, the change in wave height for SLR = 1 m is equivalent to a reef reducing in width from 1200 m to less than 200 m (Fig. 3). For wider reefs, even a modest increase of a SLR = 0.25 m is equivalent to a change in reef width of nearly 1000 m. Conversely, in the reef lagoon, changes in wave height near the shoreline due to SLR are equivalent to the lagoon width increasing from 50 m to 1800 m at the current baseline sea level (Fig. 5). Likewise, for narrow lagoons, a SLR of 25 cm has an impact equivalent to doubling the lagoon width from 200 m to 400 m. Thus, the impact of SLR is equivalent to much greater fetch lengths for the local wind waves within the lagoon.
While changes in bathymetry lead to monotonic variations in wave height, the same changes lead to more complex variations in near bed velocity over the reef flat, with a maximum in the near bed wave induced orbital velocity often occurring for particular reef bathymetry. For example, for many bathymetries, maximum velocities occur for reef flats of order 1–1.5 m deep (Fig. 4). Therefore, for narrow and shallow reefs, velocities tend to decrease with SLR; conversely, for wide and deep reefs velocities increase with SLR (Fig. 7). Taking the results for the complete range of bathymetry and wave conditions, the change of response to SLR is a function of the reef zone, both the reef flat width and the depth, and the roughness, illustrated by the contour lines for Urms/Urms0 = 1 in Fig. 7. This change over in response occurs at reef widths of order 200–300 m on shallow reefs but at reef widths of order 1 km on deep reefs with the present model and assumptions. On the fore-reef, the response is similar for all bathymetries, with a reduction in wave induced velocity with increasing SLR. Conversely, under cyclonic conditions, the velocity increases for all bathymetries and in all reef zones, with no change over in response. Nevertheless, the magnitude of the change in the hydrodynamic conditions continues to vary reef by reef.
The changes in wave forces due to SLR are equally complex and specific with regard to location, reef bathymetry and coral species (Fig. 8). While the greatest changes again occur on narrow shallow reefs, the flip over in response now occurs because of different coral species (here represented by a representative diameter), and between different zones on the reef, rather than for different reef bathymetry. The interrelationship between bathymetry, wave height, water depth and wave induced velocity, in conjunction with the dynamics of the wave forcing, results in the complete contrast in the influence of SLR on intermediate and massive corals compared to those on branching corals, particularly on shallower reefs where the impacts of SLR are greatest. Thus, SLR only is detrimental in terms of increased wave forces for branching corals and then primarily only on the reef flat. SLR is generally beneficial, i.e. decreased cyclonic wave forces, for massive corals at all locations. Therefore, different species will be impacted in different ways, and some species may in fact benefit from SLR in terms of reduced risk of cyclone damage.
Climate impacts and anthropogenic activities other than SLR may degrade the reef habitat such that a healthy rough reef loses species diversity and becomes more uniform and hence smoother (Kennedy et al., 2013). It should be noted that loss of diversity may result in dominance of a massive coral (Fabricius et al., 2011) and although this might preserve the prior macro-scale roughness, the loss of structural complexity and fine-scale roughness may result in less overall wave damping by friction and drag. Further work is required to resolve this issue. While the impacts of SLR are generally similar on rough reefs and smooth reefs for all parameters discussed above, subtle variations do occur, notably for wave-orbital velocity. Further, the interplay of reef depth, width and roughness in controlling the wave height means that different bathymetries show variations in sensitivity to changes in roughness at different sea levels. For example, for wide reefs, wave height is insensitive to roughness at present sea level, but differences of order 10% are predicted for different roughness at SLR = 1 m (Fig. 3). Similarly, roughness has little impact on wave heights on shallow reefs at present sea level (since energy dissipation is dominated by breaking), but again differences of order 15% emerge at higher sea levels. Therefore, in combination, SLR and changes in reef composition can lead to gradual but significant changes in both reef top and nearshore wave dynamics (Fig. 4, Fig. 5, Fig. 6, Fig. 7 and Fig. 8). Likewise, in the absence of SLR, changes in water depth over the reef flat may still occur from loss of coral from breakage or erosion which directly reduces reef flat elevations (Sheppard et al., 2005 and Grady et al., 2013). Similarly, reef accretion rates may or may not keep up with SLR (Buddemeier and Smith, 1988). Such variations in the reef elevation relative to sea level can be immediately assessed from the present model results by considering them as a pseudo-sea level rise.
4.2. Implications for ecosystem health
Paradoxically, the changes in reef wave dynamics from SLR may be both beneficial and detrimental. For example, it is well recognized that greater damage to coral is likely from higher wave induced loads (see Section 3.3). However, coral growth rates increase with increasing flow (Jokiel, 1978) and increasing wave induced velocity enhances particle capture (Sebens et al., 1998), such that larger waves may benefit coral growth, albeit up to a limit, and consequently, SLR (and pseudo sea level rise) may be beneficial for coral production on certain reefs and in certain zones of a range of reefs, even without considering increased accommodation space. This modeling commences the process of identifying such reefs and such zones, and conversely, those ecosystems potentially at greatest risk. Further, the nature of the ecological processes will determine if positive or negative feedback occurs between the ecological and hydrodynamic processes as a result of SLR. For example, taking roughness as a proxy for coral health, a smoother reef results in greater wave induced velocities. Therefore, if increased average wave-orbital velocity is beneficial (up to a limit) to coral health (Fig. 4 and Fig. 7), SLR will be beneficial, and there is negative feedback and stability. If however, increased wave-orbital velocity is detrimental to a coral population as a whole, SLR will be detrimental and positive feedback occurs as the roughness continually reduces, leading to greater velocities and further reduction in coral health. Consequently, in this scenario, a tipping point could exist, beyond which SLR induced velocities become very detrimental. These impacts will also influence the distribution of coral species on the reef. Likewise, the varied response of different species in terms of likely changes in breakage rates under cyclonic conditions will influence future species distribution. Consequently, there is a pressing need to investigate geographic trends in colony strength further in order to get realistic models for the future of colony fragmentation and loss during cyclonic conditions with added SLR. Similarly, not all species necessarily benefit from increased wave-orbital velocity, e.g. seagrasses, and so the response to SLR would be reversed in those cases. Overall, the modeling clearly demonstrates that predicting SLR impacts arising from changing wave dynamics on coral reefs requires careful consideration of the reef bathymetry, the reef zone and the mix of coral species.