Discussion

survival and recovery in D. antillarum populations (Lacey et al. 2013; Bodmer et al.

2015; Chapter 5), which will likely be exacerbated by continued reef flattening.

The ex situ laboratory trials show that structure is not only important for providing shelter from predation, but also potentially has a protective effect against the negative impacts of ocean warming on D. antillarum fitness. In the absence of reef material, D. antillarum PAB decreased in response to elevated water temperature, which agree with previous findings (Bodmer et al. 2017; Chapter 3). This relationship has the potential to further hinder D. antillarum population recovery due to the simultaneous occurrence of the phenomena of ocean warming and reef flattening (Alvarez-Filip et al. 2009); as reef complexity decreases, urchins become more reliant on PAB for survival, but as SST increases, PAB is reduced. However, the negative relationship between SST and PAB is shown to be attenuated by the provision of habitat complexity.

Under control temperature treatments, there is a stepwise reduction in PAB associated with increasing levels of structural complexity. Addition of habitat structure enables individuals to gain fitness benefits associated with reduced energy expenditure without incurring the fitness costs associated with increased predation pressure (Millott and Yoshida 1960). In the absence of reef material, PAB in the medium temperature treatment was lower than that observed at the control temperature, but this difference disappeared when urchins were provisioned with low complexity habitat. In the high-temperature-low-complexity treatment, PAB was significantly lower than under the low-complexity treatments of either of the other two temperature scenarios, which indicates that the mitigating effect of low complexity reef architecture may not occur if the most severe climate change predictions come to fruition. However, regardless of the temperature scenario, D. antillarum PAB was

similar across all high complexity treatments, which implies that complex reef structures may help facilitate D. antillarum survival in a warming world by providing a secondary line of predatory defence and reducing reliance on innate predator avoidance behaviours that will be negatively impacted by rising SST.

The in situ habitat complexity preference results, coupled with the ex situ laboratory trials, support the assertion that deployment of ARs is a suitable strategy for restoration of D. antillarum populations and their ecological benefits to ecosystem health and resilience. Enhancement of habitat complexity at the 5-15 cm spatial scale will not only provide D. antillarum with suitable predation refugia, but will also buffer them against the negative effects of future climatic changes. For artificial reefs to be successful in achieving their conservation goals, materials must be cheap and readily available (Fitzhardinger and Bailey-Brock 1989; Seaman 2000; Baine 2001). Concrete breezeblocks were an obvious candidate, not only because of their ubiquity throughout the Caribbean, but also because their openings have standard dimensions of ca. 10 x 10 cm. The complexity signature of a breezeblock is similar to D. antillarum’s in situ habitat preference, and therefore provides structure at the spatial scale required for size-specific protection against predation at all life-history stages.

The threefold increase in D. antillarum population size and significant benthic recovery observed around 29 ARs on a highly degraded macroalgal dominated reef system over a 24-month period provides strong evidence that deployment of ARs is a viable restoration strategy at the ecosystem level. Given D. antillarum’s important role as a macroalgal grazer (Carpenter 1984; Haley and Solandt 2001; Mumby et al. 2006;

Chiappone et al. 2013), it is unsurprising that urchin population growth was associated with decreases in macroalgae cover and increases in hard coral cover.

It must be noted that the observed increase in the means of both D. antillarum population size and percent cover of scleractinian hard coral between 2015 and 2017 is accompanied by an almost threefold increase in standard deviation. This may indicate that, while there have been significant improvements in urchin population status and metrics of reef health, D. antillarum recovery, and therefore reinstatement of its associated ecosystem functions, has been patchy.

Many studies show a positive correlation between D. antillarum population density and abundance of juvenile coral recruits (e.g. Edmunds and Carpenter 2001;

Idjadi et al. 2010). Coral recruit abundances decreased by 74 % between 2015 and 2017 on the control reefs at La Ensenada, possibly because high levels of sedimentation and macroalgae in the area have created suboptimal settlement conditions (Rogers 1990). However, the stability of the juvenile coral recruit community around the ARs suggests that reinstatement of D. antillarum grazing has acted to protect the reef from environmental/ecological factors driving declines observed elsewhere.

The benefits associated with deployment of artificial reefs appear to be highly localised, as control reefs in La Ensenada located just 100 m away from the structures saw no change in urchin population size or metrics of reef health throughout the study period. Conservationists attempting region-wide restoration will need to account for this by designing strategies that facilitate recovery in a stepwise manner. It is hypothesised that population recovery is density-dependent (Rogers and Lorenzen 2016), and will initially occur in high quality (i.e. complex) habitats. Once populations in optimal habitats reach carrying capacity, intraspecific competition will force individuals into adjacent suboptimal reef areas. These lower complexity reefs will benefit from D. antillarum grazing as macroalgae cover decreases and scleractinian

coral cover increases to enhance the structure of the environment, and promote further augmentation of urchin populations. Provision of strategically placed artificial reef structures over larger geographical areas may stimulate this cycle of localised recovery and spill-over, and ultimately enable D. antillarum to create new complex habitat that will reinforce its own recovery. Numerous studies addressing the SLOSS (single large or several small) question in marine environments conclude that conservation of small areas can have far-reaching benefits, both spatially and temporally, as a result of spill-over effects (Jones et al. 2007). The small-scale intensive management intervention proposed here is therefore based on a tried and tested approach to marine conservation.

The manpower required to stimulate the recovery of D. antillarum will be huge, and conservationists will need to look beyond the scientific community for help.

A major advantage of artificial reefs is that the theory behind their use is easy to understand, and the simple and inexpensive nature of this design means they could be deployed by interested non-experts (Seaman 2000) such as the recreational dive industry. In theory, artificial structures only ever need to be deployed once on any given local reef system because they will catalyse the occurrence of a positive feedback loop; restoration of D. antillarum ecosystem services will lead to increased hard coral cover and structural complexity, which will stimulate further increases in urchin population size. Using a combination of reef surveys, ex-situ behavioural trials and field-based experimental manipulations this study has demonstrated that deployment of artificial reefs, constructed from cheap and readily available materials, is a viable conservation strategy for the restoration of D. antillarum in the Caribbean, and may even serve to give some protection to populations against the negative impacts of climate change.