restoration, reef restoration, artificial reef, rigorous evaluation, recruitment, Bull Mar Sci, coral reef restoration, coral colonies, ecological success, experimental approach, coral reef ecosystems, restoration projects, artificial reefs, ecological restoration, Southeast Fisheries Science Center, Kenworthy J, pp, R. Bruckner, Lenihan, Lenihan HS, Kenworthy JD, restoration plans, restoration approaches, Coral reef, North Carolina Department of Environmental Health and Natural Resources, reef restoration success, Southeastern United States, conservation and restoration, global context, coral reefs, North Carolina Blue Ribbon Advisory Council on Oysters, W. Precht, M. Fonseca, quantitative evaluation, H. Lenihan, Kenworthy WJ, Ecological Systems Enhancement Technology, Japan International Marine Science and Technology Federation, Raleigh, North Carolina, Center for Coastal Fisheries, NOAA Coast Ocean Office, A. Bruckner, Bull Mar Sci Precht, Ecol
Proceedings 9th International Coral Reef
Symposium, Bali, Indonesia 23-27 October 2000, Vol. 2
The importance of evaluation, experimentation, and ecological process in advancing reef restoration success
M. W. Miller1 ABSTRACT The practice of coral reef restoration in the past has been largely confined to replacement of habitat with artificial structures and the transplant of coral colonies or fragments, either borrowed from nearby populations or rescued from the disturbance. Despite lip-service paid to "monitoring"in most restoration plans, the rigorous evaluation of these efforts has been, often, lacking completely and the declaration of success is often based on the most rudimentary criteria (stability of artificial structures or survival of x% of coral transplants). In very few cases, restoration projects have been designed to test the effectiveness of different restoration approaches (e.g. structure designs) in enhancing the biological performance (e.g. recruitment, growth, disease susceptibility) of key organisms and, in even fewer cases, in enhancing community function. However, these studies demonstrate the power of an adaptive management approach to restoration; that is, rigorously evaluating the ecological performance of alternative restoration approaches in order to do better in the future. Specific evaluative examples from other shallow coastal systems suggest that ecological success can be enhanced by simple and inexpensive restoration approaches. If coral reef restoration is to advance beyond its current "build it and they will come" paradigm, an experimental approach and the evaluation and implementation of ecological restoration
measures built upon our understanding of reef community process and function must be pursued.
Keywords Restoration, Monitoring, Evaluation, Experimentation Introduction Coral reef "restoration" in the UNITED STATES OF AMERICA
is largely confined to acute impacts such as ship groundings. Restoration approaches in this context primarily consist of engineering activities, specifically, the constructional repair of damaged reef framework and the transplantation of adult coral colonies, either from undamaged reefs or "rescued" from the incident itself. Enhancing recovery of ecological function is one of the design criteria for structural restoration projects, but factors such as cost and "aesthetics" are also considered. In the past five years, the US National Oceanic and Atmospheric Administration has undertaken several reef restoration projects in the Florida Keys National Marine Sanctuary. The engineering success of these structural restoration projects has been quite high. That is, restoration structures are mostly stable through hurricanes, and transplanted corals usually stay attached, survive and grow. However, the net benefit of coral transplantation is unclear when transplants are taken from already-depauperate donor populations (as is the case in the Florida Keys) and the success of rescued colony transplants may be compromised by the stress endured in the original disturbance. The evaluation of ecological success and of cost/benefit ratio for these projects is rarely undertaken. Monitoring versus Evaluation Essentially all large scale reef restoration projects at least pay lip-service to including "monitoring" project outcome as an explicit component of the restoration plan. However, such monitoring plans are usually inadequate
to accomplish the rigorous evaluation of ecological restoration success because of poor design or poor (or even complete lack of) execution. For example, monitoring plans may not take into account all of the proper controls (i.e. undamaged controls and damaged but unrestored controls) required to evaluate cost benefit. Often specified monitoring protocols (e.g. photos or video footage) are not carefully executed to yield quantitative data
. In the worst cases, monitoring plans are not executed at all, sometimes due to budget short-falls and/or logistical constraints. In the best cases, quantitative monitoring data is obtained, but it is rarely published or presented in a manner that facilitates the enhancement of future restoration practice. Table 1 Definitions of alternative assessment approaches to restoration projects (Grove 1961) Monitor - Watch, observe, or check for a specific purpose
- Keep track, regulate, control Evaluate - Express numerically or in monetary units - Examination and judgement concern- ing the worth, quality, significance, degree, amount , or condition It is suggested, however, that even the best-designed, best-executed and best-disseminated monitoring plan is inadequate to make the greatest progress in the field of coral reef restoration. Table 1 recounts relevant aspects of dictionary definitions for "monitoring" and "evaluation". If we are to make the greatest progress in advancing any aspect of reef restoration success, we must undertake to obtain quantitative, numerical bases on which to make rigorous judgements of worth and signi-
1 NOAA-Fisheries, Southeast Fisheries Science Center
, 75 Virginia Beach Dr. Miami FL 33149 USA, 305-361-4561, 305-361-4562 (FAX), [email protected]
ficance. That is, monitoring (i.e. watching, observing, and checking) is not enough; we must undertake rigorous evaluation of restoration actions. In a similar vein, Precht et al. (2001) call for an approach where management restoration decisions are considered as hypotheses of ecosystem response and "scientific monitoring" is designed as an experiment to test these predictions. Such programmatic and methodological emphasis on evaluation of artificial reef projects has also recently been advocated by Seaman (2000). Miller and Barimo (2001) used a comparative approach to evaluate ecological success (as defined by in situ recruitment of SCLERACTINIAN CORALS
) of two different reef restoration structures in the Florida Keys National Marine Sanctuary. Density and species richness
of juvenile coral recruits were substantially higher on lime-rock boulder or cobble surfaces than on the flat concrete structure. An experimental component of this study (using standard settlement plates at both sites with isolated design characteristics) is underway to determine what aspects of structure/surface design are responsible for these differences in coral recruitment. Such an experimental approach to evaluation may be required to determine why certain outcomes are observed and is detailed below. Rigorous quantitative evaluation can also provide benefits in the more mundane aspects of ecosystem restoration such as the processes of litigation and assignment of damages to parties causing resource injury. Evaluating natural ecosystem recovery rates is an important pre-requisite to determine the benefit of active restoration. Such quantification and modeling of natural recovery applied to the scaling of restoration projects, known as Habitat Equivalency Analysis, has been advan-
ced in natural resource litigation relating to seagrass ecosystems and has so far been upheld in court (Fonseca et al. 2000). This sort of recovery evaluation and modeling approach may be more difficult to apply in more biocomplex coral reef ecosystems, but warrants undertaking, especially in a global context where coral reefs presumed resilience may be severely compromised by multiple anthropogenic stresses of long duration (Nystrцm et al. 2000, Jackson et al. 2001). Ironically, the legal strictures on settlement funds are interpreted to forbid recovery monitoring as a superfluous "research" activity despite the potential benefits to future litigation (both trustees and responsible parties) of providing a credible way to scale compensatory restoration projects (and therefore their cost) to equal the scale of the natural resource injury over time. Experimentation as a tool in reef restoration As students of science, we learn to contrast the purpose and the mechanics of observational versus experimental studies (for example, see Hairston 1989). Most restoration monitoring studies are observational in nature. They can reveal patterns of abundance or distribution. They cannot discern the cause or causes underlying such patterns. Therefore, if the pattern revealed in a restoration monitoring study is not the pattern expected (e.g. perhaps some target organism fails to recruit), there is no basis for determining why (was there some flaw in the project design, or was there simply a natural dip in recruitment levels in the whole geographical area). Similarly, and perhaps more importantly, a descriptive monitoring study will never be able to tell us how to conduct restoration better.
Fig. 1 Schematic depiction of the continua in types of science, types of restoration studies, related general and specific questions and/or outcomes from each as discussed in this paper
Along with evaluation and discernment of why projects may fail, experimentation is the only method which will allow us to develop and test novel techniques with the goal of improving the effectiveness of the restoration of reef ecological function. The outcomes of such an approach are not always technically complex and more expensive, though sometimes they are. It is imperative for the field of coral reef restoration to move along the continuum from observation/monitoring toward
evaluation and experimentation if it is to advance beyond simpleminded engineering success criteria to accomplishing true restoration of ecological function (Fig 1). Below are some positive examples from other coastal ecosystems of using an experimental approach, guided by a basic understanding of ecological process and function in the given ecosystem, and the benefits to functional restoration success which resulted.
San Onofre Kelp Restoration As mitigation for the destruction of an extensive kelp bed by thermal pollution from a Nuclear Power
station, Southern California
Edison was legally mandated in the mid 1980's to restore over 60 ha of kelp forest. This mandate was followed by a spate of observational ecological studies of existing natural and artificial reefs in southern California that aimed to discern what ecological factors and artificial reef designs would foster kelp recruitment, growth, and persistence. These observational studies clearly identified several design and ecological features of a restorative art
ificial reef that would do just that. For example, in contrast to coral reef ecosystems, greater reef height and topographic complexity were determined to be detrimental to the overall restoration goals of establishing persistent kelp beds as the enhanced fish habitat fostered destructive grazing levels (Patton et al. 1994). Observations suggested that kelp often recruited but failed to persist and were replaced by sessile invertebrates. Experimental studies on several artificial reefs suggested that moderate sized reef elements would provide substrate disturbance (e.g. via overturn of boulders) at an interval (once every 7-8 years) that would help alleviate this competitive replacement (Patton et al. 1995). In other cases, the benefit of certain factors or manipulations in the restoration context were equivocal. Kelp transplantation provided benefits at some artificial reefs, but not at others (e.g. Carter et al. 1994). Also, observations suggested that the spacing of artificial reef boulders and the areal extent of interstitial sand surface
might be important, but it was not clear what spacing would be optimal. Hence, the restoration project was planned in a two-phase approach. The initial phase consists of, essentially, a large scale 3-way factorial experiment to test for the relative benefits of three levels of boulder spacing, two types of boulder material (concrete vs. quarried rock), and active transplantation of juvenile kelp. The so-called "experimental reef" was constructed in 1999, is 9 ha in size, will undergo a five-year evaluation period, and has a total estimated cost
of US$46 million, only half of which is the cost of construction with the remainder going to monitoring and evaluation
activities ( http://www.sce.com/songs_env/reef.html ). The results from this experimental reef will be incorporated into the second phase, a full-scale mitigation reef which will be 61 ha in size. Whereas this planning and pre-evaluation process has been long, cumbersome, and expensive, it has prevented a very large expenditure of mitigation effort and dollars on a restoration project that would have been unsuccessful (i.e. kelp would have been overgrazed on a short term or out-competed by sessile invertebrates on the medium term). Oyster Reef Restoration Oyster populations, landings, and oyster reef habitats in the estuaries of the eastern United States have been in dramatic decline due to destructive harvesting techniques, water quality
degradation, disease, and feedback inter-
actions exacerbating these factors (Lenihan and Peterson 1998, Lenihan et al. 1999). The "just do it" approach had been undertaken by the state of North Carolina
, USA in their restoration program which used dead oyster shell to form mounds about 1 m tall (Lenihan 1999). These restored oyster reefs had minimal success (discerned with minimal evaluation effort via from normal harvest attempt by fishers) as most oysters that settled on the reefs died before they reached harvestable size
(Frankenberg 1995 cited in Lenihan 1999). Thus a series of experimental evaluation studies were undertaken to understand why the restored reefs failed to provide the intended benefits and what alterations in restoration design and execution could enhance oyster success (Lenihan 1999, Lenihan et al. 1999). The primary question was what habitat characteristics foster oyster performance and how? Basic ecology suggests many processes that may be important including physio-chemical tolerance, available trophic resources, recruitment, and predation and/or disease. Lenihan (1999) quantified all these aspects of oyster performance and hydrodynamic and hydrographic conditions on restored oyster reefs of different heights, depths, and on different positions of these reefs. The results were extremely clear. All aspects of oyster performance were enhanced at positions higher up in the water column
(i.e. crests of tall reefs or short reefs in shallow water
) as seasonal anoxia or hypoxia were the major source of mortality of recruited oysters. Also, high water flow at the crests of reefs, especially tall reefs, enhanced oyster growth, condition, survival, and decreased parasitic infections (Lenihan 1999, Lenihan et al. 1999). Hence, in this case, experimental evaluation yields extremely simple, cheap, and effective advise for enhancing oyster reef restoration success: Build TALL reefs! Seagrass Restoration Seagrass meadows are extremely susceptible to anthropogenic physical disturbance in shallow coastal areas favored by recreational boaters. The removal of sediments in such incidents often delays or prevents natural recovery over a decadal time scale and often results in expansion of the unvegetated scar by storms or other natural disturbance (Precht and Gelber, in press). Transplanting seagrasses has been a long-used technique for accelerating recovery of boat scars and numerous techniques have been used across a range of environments with little rigorous evaluation of the relative cost:benefit of the different methods. Fonseca et al. (1994) provide an experimental evaluation of a standard restoration technique (seagrass planting) undertaken in different locations and under a range of protocols (planting methods, enhancements such as fertilizer addition or cages to protect the transplants from destructive bioturbation). They found some geographical variation in the best techniques, but overall, peat-pots were the cheapest planting method. Such experimental evaluations for seagrass restoration techniques have made possible the development of a guidance handbook for restoration managers in the southeastern USA (Fonseca et al. 1998).
Ecological studies of natural seagrass succession suggest that nutrient accumulation in the sediments can be limiting to seagrass developmen (Williams 1990) and the severe sediment excavation of boat groundings suggests that nutrient biogeochemical cycling may be severely disturbed over a long time frame. Nutrient addition was one of the factors that the experimental seagrass transplant evaluation by Fonseca et al. (1994) found to be equivocal, partly due to inconsistent nutrient release in the experiment. A subsequent study by focused on experimental evaluation of two novel nutrient enhancement techniques: mechanical injection of a nutrient solution
from a boat directly into the sediments using a large roller apparatus, and the placement of pvc stakes that serve as bird roosting stakes within the scars (Kenworthy et al. 2000). They found a clear enhancement of seagrass cover (50% within 1.5 yrs) in the bird stake treatments at very low cost (both in terms of materials and effort) and no discernable enhancement from the intensive boat/injection treatments (Kenworthy et al. 2000). This restored seagrass cover was in fact an earlier successional species known to (Halodule wrightii) than the one originally present in the scars (Thalassia testudinum) that is known to be outcompeted by H. wrightii in high nutrient conditions (Fourqurean et al. 1995). However, the nutrient-enhanced colonization of seagrass cover, even if strictly "out-of-kind", decreases the likelihood of scar expansion and will allow the resumption of natural biogeochemical cycling in the sediments. Kenworthy et al. (2000) recommend removal of the bird stakes in one-two years which should result in resumption of T. testudinum dominance (Zieman 1982, Williams 1990), a strategy for restoration termed "compressed succession". Once again, a very simple and inexpensive technique utilizing natural ecological processses and interactors can be extremely effective in coastal restoration. Conclusion Based on these examples, it is clear that an experimental approach can uncover simple and cost-effective techniques for enhancing restoration function. Because such an approach has not really been pursued in coral reef systems, it may be difficult to predict what novel, costeffective ecological restoration techniques (i.e. the bird stakes) for coral reef systems might be. I suspect they may involve broad categories such as manipulation or enhancement of herbivory, enhancing juvenile coral recruitment and growth (e.g. larval seeding, enhancing growth/production in coral nurseries), and possibly the use of sponge transplants for purposes of small rubble stabilization. These cost-effective techniques for coral reefs may not be as simple or cheap as some of the examples given above, but it is quite plausible that some "low-tech" ecological enhancements may be extremely effective and feasible for application even in lessdeveloped countries
. We must move from an observational monitoring mode along the continuum toward rigorous evaluation and experimentation with enhancements of ecological
processes (Fig. 1) if we are truly to restore coral reef ecosystems. This endeavor gains all the more prominence in the global context of reef decline that has been demonstrated at the 9th ICRS, and in a context of suspicion that coral reef systems may have lost much of their ability to recover on their own (Nystrцm et al. 2000). Acknowledgements Support for the production of this manuscript was provided by NOAA-Fisheries (Southeast Fisheries Science Center). Discussion and data were provided by A. Bruckner, R. Bruckner, M. Fonseca, H. Lenihan, and W. Precht and all are greatly appreciated. However, the views expressed and any errors they may contain are strictly my own. References Carter JW, Jessee WN, Foster MS, Carpenter AL (1994) Management of artificial reefs designed to support natural communities. Bull Mar Sci 37:114-128. Fonseca MS, Julius BE, Kenworthy WJ (2000) Integrating biology and economics in seagrass restoration: How much is enough and why? Ecol Engineering 15:227-237. Fonseca MS, Kenworthy WJ, Courtney FX, Hall MO (1994) Seagrass planting in the Southeastern United States: methods for accelerating habitat development. Restoration Ecology 2:198-212. Fonseca MS, Kenworthy WJ, Thayer GW (1998) Guidelines for the conservation and restoration of seagrasses in the Unites States and adjacent waters. NOAA Coastal Ocean Program Decision Analysis Series No. 12. NOAA Coast Ocean Office, Silver Spring, MD: 222 pp. Fourqurean JW, Pwell GVN, Kenworthy J, Zieman JC (1995) The effects of long-term enrichment of nutrient supply on competition between the seagrasses Thalassia testudinum and Halodule wrightii in Florida Bay. Oikos 72:349-358. Frankenberg D (1995) Report of North Carolina Blue Ribbon Advisory Council on Oysters. North Carolina Department of environmental health
and natural resources
, Raleigh, North Carolina, USA. Grove PB (ed) (1961) Webster's 3rd New International Dictionary. G. and C. Merriam, Springfield, MA Hairston NG Sr 1989 Ecological experiments: purpose, design, and execution. Cambridge Univ. Press, Cambridge: 370 pp. Lenihan, HS (1999) Physical-biological coupling on oyster reefs: How habitat structure influences individual performance. Ecol Monogr 69:251-275. Lenihan HS, Micheli F, Shelton SW, Peterson CH (1999) The influence of multiple environmental stress
ors on susceptibility to parasites: An experimental determination with oysters. Limnol Oceanogr 44: 910-924. Lenihan HS, Peterson CH (1998) How habitat degradataion through fishery disturbance enhances impacts of hypoxia on oyster reefs. Ecol Appl 8:128-140. Jackson JBC and 18 others (2001) Historical overfishing and the recent collapse of coastal ecosystems. Science 293: 629-638.
Kenworthy JD, Fonseca MS, Whitfield PE, Hammerstrom K, Schwartzschild AC (2000) A comparison of two methods for enhancing recovery of seagrasses into propeller scars: mechanical injection of nutrient and growth hormone solution vs. defecation by roosting seabirds. Final Report: Center for Coastal Fisheries and Habitat Research NCCOS/ NOS/NOAA. Beaufort, NC: 34 pp. Miller MW, Barimo J (in press) Assessment of juvenile coral populations at two reef restoration sites in the Florida Keys National Marine Sanctuary: Indicators of success? Bull Mar Sci Nystrцm M, Folke C, Moberg F (2000) Coral reef disturbance and resilience in a human-dominated environment. Trends Ecol Evol 15:413-417. Patton ML, Grove RS, Honma LO (1995) Substrate disturbance, competition from sea fans
(Muricea sp.) and the design of an artificial reef for giant kelp (Macrocystis sp.). In: Proceedings, International Conference
on Ecological Systems Enhancement Technology for Aquatic Environments. Japan International
Marine science and technology
Federation, Tokyo. Patton ML, Valle CF, Grove RS (1994) Effects of bottom relief and fish grazing on the density of the giant kelp, Macrocystis. Bull Mar Sci 55:631-644. Precht WF, Aronson RB, Swanson DW (in press) Improving scientific decision-making in the restoration of shop grounding sites on coral reefs. Bull Mar Sci Precht WF, Gelber A (in press) Persistence and enlargement of boat-grounding scars on Thalassia dominated seagrass beds: implications for restoration. Bull Mar Sci Seaman W, Ed. (2000) Artificial reef evaluation with application to natural marine habitats. CRC Press, Boca Raton, FL
: 246 pp. Williams SL (1990) Experimental studies of Caribbean seagrass bed development. Ecol Monogr 60:449-469. Zieman JC (1982) The ecology of seagrasses of South Florida
: a community profile. US Fish and Wildlife Service, Office of Biological Services, Washington DC. FWS/OBS-82/85: 15 pp.