Harnessing natural recovery processes to improve restoration outcomes: an experimental assessment of sponge-mediated coral reef restoration

Abstract

Background: Restoration is increasingly implemented to reestablish habitat structure and function following physical anthropogenic disturbance, but scientific knowledge of effectiveness of methods lags behind demand for guidelines. On coral reefs, recovery is largely dependent on coral reestablishment, and substratum stability is critical to the survival of coral fragments and recruits. Concrete is often used to immobilize rubble, but its ecological performance has not been rigorously evaluated, and restoration has generally fallen short of returning degraded habitat to pre-disturbance conditions. Fragments of erect branching sponges mediate reef recovery by facilitating rubble consolidation, yet such natural processes have been largely overlooked in restoring reefs.

Methods: On two reefs in Curacao, four treatments - coral rubble alone, rubble seeded with sponge fragments, rubble bound by concrete, and concrete "rubble" bound by concrete - were monitored over four years to investigate rubble consolidation with and without sponges and the ecological performance of treatments in terms of the number and diversity of coral recruits. Species specific rates of sponge fragment attachment to rubble, donor sponge growth and tissue replacement, and fragment survival inside rubble piles were also investigated to evaluate sponge species performance and determine rates for sustainably harvesting tissue. FINDINGS/SIGNIFICANCE: Rubble piles seeded with sponges retained height and shape to a significantly greater degree, lost fewer replicates to water motion, and were significantly more likely to be consolidated over time than rubble alone. Significantly more corals recruited to sponge-seeded rubble than to all other treatments. Coral diversity was also greatest for rubble with sponges and it was the only treatment to which framework building corals recruited. Differences in overall sponge species performance suggest species selection is important to consider. Employing organisms that jump start successional pathways and facilitate recovery can significantly improve restoration outcomes; however, best practices require techniques be tailored to each system.

Figure 1. Location of study sites.

A. Map of the Caribbean with the location of Curaçao indicated. B. Map of Curaçao, Netherlands Antilles. Filled triangle and circle indicate location of study sites Sea Aquarium reef and Barracuda Point reef, respectively.

Figure 2. Erect branching sponge species and experimental substratum treatments.

A. Aplysina cauliformis. B. Aplysina sp. C. Niphates erecta. D. Coral rubble alone. E. Coral rubble with sponge fragments inserted into pile. F. Concrete bound coral rubble. G. Concrete bound concrete “rubble”.

Figure 3. Influence of treatment and water depth on rubble pile height over time.

Conditional boxplots of rubble pile height for Sea Aquarium and Barracuda point. A and C. Rubble pile height conditional on water depth at Sea Aquarium and Barracuda Point, respectively. B and D. Rubble pile height conditional on treatment at Sea Aquarium and Barracuda Point, respectively. Thick solid and broken bars inside boxes indicate median and mean height, respectively. Letters above boxes indicate significant differences (P<0.05) between factor levels based on pairwise comparison of means with Bonferroni correction.

Figure 4. Sponge fragments stabilizing rubble in piles three months post deployment at Sea Aquarium.

Sponge fragments inserted into piles grew and adhered to adjacent pieces of rubble in less than three months. A and B. Aplysina cauliformis stabilizing sections of rubble piles. C. Coral rubble stabilization by Aplysina sp.

Figure 5. Consolidation of rubble piles with and without sponges by carbonate secreting organisms.

A. Sea Aquarium. B. Barracuda point. Number of unconsolidated piles is represented by light grey bars and consolidated piles by black bars. Asterisks above bars indicate significant differences (* p<0.05, ** p<0.01) in the proportion of consolidated vs. unconsolidated piles between treatments within the same site, at the same time period, by the G-test of independence. Statistical results from each comparison are provided in Table S3.

Figure 6. Number of coral recruits at each site, during each survey, by treatment.

A. Sea Aquarium. B. Barracuda point. For Sea Aquarium (A): open, light gray, dark gray and black bars indicate the number of coral recruits at months 12, 24, 36, and 48, respectively. For Barracuda Point (B): open, light gray, dark gray and black bars indicate the number of coral recruits at months 12, 21, 33, and 45, respectively. Letters above black bars (months 48 and 45 for Sea Aquarium and Barracuda Point, respectively), indicate significant differences (P<0.05) among treatments within each site in the number of coral recruits based on pairwise comparison with Bonferroni correction. Within each site, bars that share the same letter are not significantly different (P>0.05) from one another.

Figure 7. Cumulative replacement of excised tissue (PVR) by sponges.

A. Cumulative mean percent of tissue excised that was replaced at each 3 month period, in terms of volume, by all sponges surviving for 15 months: Aplysina cauliformis (N = 29), Aplysina sp. (N = 34), Niphates erecta (N = 39). Standard error bars are shown. Aplysina cauliformis and Aplysina sp. replaced tissue significantly (P<0.05) more rapidly than N. erecta (determined by pairwise comparison of means with Bonferroni correction) (See Table S4).