Hard-bottom habitats support commercially important fish species: a systematic review for the North Atlantic Ocean and Baltic Sea

Search protocol

We followed the protocol for a systematic literature review proposed by Pullin & Stewart (2006), which set forth guidelines for planning the review and analysis of the data. We systematically reviewed the literature and evaluated articles returned by three different databases (DTU-Findit, Web of Science, and Scopus) for relevance based on our inclusion criteria (Table 1). Screening was conducted at the level of title, abstract, and full text (Pullin & Stewart, 2006) to each of several ecological study variables (see “search string” below), using the PRISMA protocol (Moher et al., 2009) (Fig. 1). The first author of this publication screened each article by title to eliminate non-relevant articles, and a subset of articles was sent to all co-authors to compare the group’s decisions and ensure we came to consensus on what should be discarded. Then, the same procedure was followed for examining the abstracts of the remaining articles. Finally, we divided the articles for full-text examination among all authors and each person independently extracted data, with a group discussion of which articles were relevant and why, and resolving any discrepancies on what should be included. The search was run on the three different databases on April 7th, 2017. The articles returned by the three different databases were evaluated for relevance based on the inclusion criteria (Table 1), applied at three successive levels: (1) title, (2) abstract and (3) full-text. It is important to note that there are some limitations of our approach, and that we may have missed some research that did not fall into our selected categories based on the search terms we used, and results are only applicable to the species we incorporated into our search (namely ICES-managed species). Our review focused on the species for which ICES gives advice, which resulted in examining the species of importance for ICES Member Countries (i.e., Belgium, Canada, Denmark, Estonia, Finland, France, Germany, Iceland, Ireland, Latvia, Lithuania, the Netherlands, Norway, Poland, Portugal, Russia, Spain, Sweden, United Kingdom, and United States of America; US and Canadian fish stocks are not included in the advice, though these are ICES Member Countries). Therefore, interpretations of these data apply only to the species and habitats we incorporate. Additionally, we examined the current literature (2017–2022) for more recent articles on the same ICES-managed species using hard bottom habitats (database from B Ciotti pers. comm., 2022), which allowed us to identify a couple of case studies to serve as a check on validity of conclusions from our systematic search.

Data extraction

Our primary question was: “What is the habitat importance of hard-bottom natural and artificial structures for temperate fish species managed by ICES?” To answer this question, for each article we determined the studied habitat, the ecosystem type, and species targeted, as detailed below. We broke down habitat importance into the following four main components: (1) distribution-related variables, (2) community-related variables, (3) fitness-related variables, and (4) reproduction-related variables (see details below).

Review selection process

The search process returned an initial 2,651 articles, paring down to 2,333 articles after exclusion of duplicate references. Of those, 45 articles were returned after the subsequent screening criteria and were examined for full-text data extraction (as detailed below and in Fig. 1).

Full text analysis

From the 161 full texts analysed, 116 (72%) did not comply with the exclusion criteria. Most of the articles were excluded based on location (49%), corresponding mostly to studies targeting rockfish in the Pacific Ocean. A total of 21% exclusions represented articles related to non-target species, while 13% studied non-target ecosystems and 7% did not relate to the target topics. A total of 11 articles were excluded for “other” reasons: four were not peer-reviewed, four were reviews, and for three it was not possible to obtain a full text version of the document. Therefore, the review focused on 45 studies distributed along the Baltic and North Seas and northern Atlantic Ocean (Fig. 2). We quantified the numbers of articles or percentage of articles reviewed as a function of the various response variables emphasized.

Articles reviewed as a function of response variables emphasized

Distribution-related variables (i.e., diel cycle, site fidelity, migration) were studied in six articles (13.4%). These targeted primarily site fidelity and migration patterns of Atlantic cod (Gadus morhua; e.g., Reubens et al., 2014; Reubens, Degraer & Vincx, 2014).

Community-related variables (i.e., density, biodiversity, biomass) were the most frequently studied amongst the articles reviewed, being analysed in two thirds of the studies (Fig. 3). Most research targeted a few species (i.e., Atlantic cod, herring (Clupea harengus), and saithe (Pollachius virens)) and a limited set of variables (i.e., biomass, density, biodiversity, and feeding). Cod was the species most commonly studied in community-related (i.e., density, biodiversity, biomass) studies (Table 2).

Fitness-related variables (i.e., condition, mortality, growth, feeding) were studied in 17 articles (38% of the total number of analysed articles). Feeding was studied primarily on cod and saithe, showing a tendency for fish use of hard-bottom habitats for foraging (Fig. 4; e.g., Johannessen, 1986; Malek, Collie & Taylor, 2016; Wennhage & Pihl, 2002). Interestingly, with the exception of cod, little research has been carried out on the remaining fitness-related variables (Table 2).

Reproduction-related variables (i.e., spawning, settling, nursery) were studied in six articles (13.4%). Spawning was studied primarily on herring and cod, with articles on nursery and settling being almost absent for other species in the reviewed literature (Table 2). From the species targeted in our review, hard-bottom substrates appeared to be of particular importance for herring spawning (Aneer & Nellbring, 1982; Johannessen, 1986; Kääriä et al., 1997; Šaškov et al., 2014).

There was a clear predominance in the study of community-related variables in the analysed research (Fig. 3). Broadening the scope of studied variables and targeted species may prove essential to fully understand the relationships between hard-bottom reefs and commercially important species. For example, more quantitative research at the community level reveals the value of hard-bottom habitats for fish populations (essential fish habitat, sensu Valavanis, 2008; or effective habitat, sensu Dahlgren et al., 2006), and integrating this information into assessment and management may provide considerable advances towards ecosystem-based fisheries management (e.g., fish production estimates that can be monetized).

Ecosystems explored

Research was nearly equally distributed across natural and artificial reef structures, with 26 and 27 articles related to each, respectively, and eight targeting both natural and artificial structures simultaneously (Fig. 4, Table 3). Furthermore, 21 articles (46.7%) that resulted from our search performed comparisons of hard-bottom habitats with soft-bottom habitats. The review indicated that different regions of the world have different research foci. All articles on the north-western shore of the Atlantic Ocean focused on natural reefs, with one article also targeting an intentional artificial reef (Fig. 5). Similarly, there was a greater focus on natural reef research in the Baltic Sea, with nine out of eleven articles targeting natural reefs, as well as three articles targeting artificial de facto reefs, and only one article targeting intentional artificial reefs. Conversely, in the North Sea, only three out of 15 articles targeted natural reefs, while 14 targeted de facto artificial reefs (most of which explored the effects of wind farms). Finally, on the north-eastern coast of the Atlantic Ocean, there were nearly equal numbers of articles on natural and artificial reef research (six and five articles, respectively, plus two articles that targeted both reef types simultaneously). However, of the articles targeting artificial reefs, only one studied de facto artificial reefs, while the remaining explored artificial intentional artificial reefs. There was a noticeable paucity of studies in the northern regions. Additionally, most of the analysed research has been conducted near-shore. Broadening the scientific knowledge on the importance of different habitats, areas, and regions along a depth gradient away from shore may provide crucial insights for future management.

Frequently studied species

There has been a consistent research effort towards multiple morphological and behavioural aspects of cod, with this species being researched twice as frequently as herring, the second-most-researched species (25 vs 12 articles, respectively; Table 2). Furthermore, while research on cod and herring is spread across the four groups of ecological variables considered in our review (i.e., species distribution, community, fitness, reproduction), research is lacking for most reviewed species on variables not related with community (e.g., condition, settling, site fidelity; Table 2).

The presence of hard-bottom habitat was associated with enhanced biomass in cod and saithe and, conversely, was negatively associated with biomass in herring and whiting (Merlangius merlangus; Fig. 6). Most articles compared fish densities in multiple habitats, and higher densities of cod on hard-bottom habitats were demonstrated in several articles, with only one article suggesting fewer cod on hard-bottom compared to adjacent soft-bottom habitats. Both positive and negative relationships of hard-bottom habitats with density were demonstrated for herring, saithe, striped red mullet (Mullus surmuletus), whiting, pollock (Pollachius pollachius), plaice (Pleuronectes platessa), and sole (Solea solea). Cod, herring, saithe, striped red mullet, whiting, and pollock appeared to benefit from foraging on hard-bottom habitat (i.e., for plaice feeding, only negative relationships of hard-bottom habitats were reported, as plaice feed less frequently over hard-bottom habitats). Few articles targeted spawning on hard-bottom habitats, and those that did reported associated benefits for cod and herring.

Changes in fish distribution

Research has demonstrated positive effects of the presence of hard-bottom substrates on cod biomass and density (e.g., Krone et al., 2013; Pihl & Wennhage, 2002). With a few exceptions, natural hard-bottom habitats and artificial reefs harbor significantly higher abundances of fish than adjacent unstructured habitats (Rodríguez-Cabello et al., 2008; Fujii, 2016). Fish movement patterns reflected relatively high site fidelity to structures (e.g., Hartman, 1987; Reubens et al., 2014; Reubens, Degraer & Vincx, 2014), and the stomach contents of fish residing on reefs (excluding biogenic reefs) reflected the prey assemblages associated with the reefs (e.g., (Reubens et al., 2014)), suggesting extended residency and foraging in the reef areas. Interestingly, catch per unit effort (CPUE) of cod was greater adjacent to wind farms than in shipwrecks in the Belgian part of the North Sea (Reubens, Degraer & Vincx, 2014), suggesting attraction towards wind farms. It is important to note that the studied wind farms were closed to fisheries, while the shipwreck sites had no fisheries restrictions, possibly biasing the comparison.

Gadoids such as saithe, haddock (Melanogrammus aeglefinus) and cod were the most common species at the “Miller” oil and gas platform in the North Sea (Fujii, 2016). Cod abundance was approximately four times higher at the platform than in nearby shipwrecks or sandy bottom. Using trawl data, Martin et al. (2012) modeled distributions of ten demersal elasmobranchs in the eastern English Channel in relation to the environment, with female small-spotted catshark (Scyliorhinus canicula) occurring at higher densities on hard than soft substrate.

Within four years after the deployment of artificial reefs in an area previously closed to trawl fisheries in the Bay of Biscay, there was a rapid increase in the biomass of small-spotted catshark (Rodríguez-Cabello et al., 2008). Importantly, four to eight years after deployment, axillary sea-bream (Pagellus acarne), horse mackerel (Trachurus trachurus), and striped red mullet demonstrated significant increases in biomass. The observation that horse mackerel responded positively to the addition of artificial reefs demonstrates that schooling pelagic species may also benefit from these hard-bottom habitats.

Reefs (excluding biogenic reefs) may also enhance economic returns. For example, in one location, reefs near Algarve and Faro, Portugal, raised catch rates and economic returns (Neves Santos & Costa Monteiro, 1998). Specifically, CPUE at the reefs exceeded CPUE at the control sites by a factor of 2.03–2.28 for the protected reef and by a factor of 1.11–1.86 for the exploited reef. In some artificial reef systems, the economic return per unit of effort was higher at the reefs than at the control sites. Fishing on artificial reefs would be expected to yield 2.2 times the economic return compared to the control sites (Neves Santos & Costa Monteiro, 1998). The productivity of reefs increased throughout the 15-year period, implying that the interaction between the hard substrate and the marine environment may take many years to produce the full effects on the carrying capacity (Whitmarsh et al., 2008). A distinct spatial pattern of fishing activity around the reefs was detected, with a concentration of fishing gear on the periphery of the reef area and a lower gear density at farther distances. A concentration of fishing gear nearby the reef would suggest that the improved fishing opportunities offered by the reefs have been recognized by fishermen. Fishermen likely take advantage of the spill-over of fish from the reefs, consistent with recent studies (Rosemond et al., 2018).

The inclusion of quantitative data on hard-bottom habitats in the monitoring of fish populations may further resolve a mismatch between fish population assessments and catches in fisheries of several fish species with a patchy distribution due to their high affinity for hard-bottom habitats (e.g., cod; Wieland et al., 2009). The mismatch between fish population assessments and catches of fish species with a patchy distribution may be further exacerbated during stock recovery, where the most suitable habitats (i.e., hard bottoms) will be populated first according to the ideal free distribution theory of MacCall (1990).

Not all fish species benefit from hard-bottom or marine constructed structures. Neutral or negative relationships with biomass and density were recorded for flatfish such as plaice (Pleuronectes platessa) and sole (Solea solea), as well as for whiting (Merlangius merlangus). For example, Støttrup et al. (2014) found decreasing densities of sole post restoration of a natural reef, and Bergström, Sundqvist & Bergström (2013) only found whiting before the implementation of a wind farm. Neutral or negative relationships with hard-bottom habitat for some species reflects the preferred habitat of the different species, with both flatfish and whiting occurring primarily on soft-bottom habitats (Stål, Pihl & Wennhage, 2007).

The rapid growth of offshore wind farms, coupled with the high abundance and biomass of fish associated with wind farms, suggest that these de facto marine protected areas (MPAs) could be relatively easy to enforce, serve as broodstock sanctuaries that benefit fished populations, and optimize use of the seafloor (Fowler et al., 2018). Moreover, the economic impact of artificial reefs also appears high relative to unstructured control sites. Although the potential positive relationships of artificial reefs with fish populations are encouraging, it is important to recognize the attraction effect of artificial reefs and how such a process can facilitate overfishing (Bohnsack, 1989; Smith, Lowry & Suthers, 2015, and references therein). Artificial reefs, however, can also enhance growth and survival, for example by providing hard substrate for sessile fouling communities that provide food to fish (Todd, Lavallin & Macreadie, 2018). Quantitative evidence may also guide modeling efforts to assess the role of artificial reefs on population dynamics of exploited species. Lastly, given that unstructured seafloor (i.e., flat, soft-bottom habitat) can be many orders of magnitude greater in areal footprint than artificial reefs, it is important to recognize that unstructured habitats can serve as effective juvenile habitat (sensu Dahlgren et al., 2006) for exploited fishery species. Unstructured habitat is especially important for species for which ICES gives advice. For example, in comparing coastal habitat use among various structured (e.g., seagrass, kelp, marsh) and unstructured habitats, subtidal soft-bottom was the habitat used as a spawning or nursery area by the largest proportion of ICES species examined, with intertidal soft-bottom habitat as the next most important habitat (Seitz et al., 2014).

Habitat heterogeneity enhances biodiversity

Overall, both natural and artificial hard-bottom or reef structures tend to have a high attraction potential and house densities of fish superior to surrounding habitats (e.g., Bergström, Sundqvist & Bergström, 2013; Støttrup et al., 2014; Whitmarsh et al., 2008). In one study, natural structured habitats (rock, shell, rubble) compared to bare sand had positive effects on density, growth, and survival of juvenile fishes and invertebrates, with a greater positive effect for arthropod invertebrates than vertebrates (Lefcheck et al., 2019). When man-made structures are placed on surrounding soft-bottom areas (e.g., wind turbine foundations, oil platforms), they represent beacons of habitat heterogeneity that are capable of attracting new residents (Gascon & Miller, 1982; Stenberg et al., 2015; Castège et al., 2016; Fujii, 2016). For example, cod appeared to have a clear affinity for hard-bottom substrates, with increased fish densities and/or biomass near natural hard-bottom reefs (Stål, Pihl & Wennhage, 2007; Støttrup et al., 2017; Kristensen et al., 2017), offshore wind farms (Bergström, Sundqvist & Bergström, 2013; Reubens et al., 2013; van Hal, Griffioen & van Keeken, 2017), oil-platforms (Fujii, 2015), shipwrecks (Wieland et al., 2009; Krone et al., 2013) and other man-made structures (Langhamer & Wilhelmsson, 2009; Wehkamp & Fischer, 2013). Similarly, pollock generally restricted their movement to a submerged reef area in Loch Ewe, Scotland, while saithe ranged more widely during daytime and returned to the reef during night-time (Sarno, Glass & Smith, 1994). Specific age groups of cod and pouting (Trisopterus luscus) can have high residency and site fidelity to the bases of marine wind turbines, feeding on the dominant epifaunal prey growing on these introduced habitats (Reubens et al., 2014; Reubens, Degraer & Vincx, 2014).

Different species react differently to the addition of hard-bottom substrates, and recent literature reports on various outcomes even within a species. For example, whiting can occur in relatively high abundance on hard substrates (e.g., shipwrecks and offshore wind farms; Krone et al., 2013; Lindeboom et al., 2011; Pihl & Wennhage, 2002), as well as on soft substrates (Stål, Pihl & Wennhage, 2007; van Hal, Griffioen & van Keeken, 2017). Additionally, in some cases, the attraction of reef-associated demersal fish species such as bogue (Boops boops), horse mackerel, and axillary sea-bream may make them more vulnerable to predation, as the probability of prey–predator encounters increases compared to unstructured habitats (Leitão et al., 2008). Lastly, some commercially important species, such as dab (Limanda limanda) and sole, may avoid rocky substrates, with decreasing abundances whenever there is a shift from soft to hard bottom (van Hal, Griffioen & van Keeken, 2017). Thus, the relationship of hard-bottom structures with fish distribution and abundance patterns in the North Sea and Baltic Sea are species-dependent, a factor that must be considered during management and restoration actions.

Improved fitness and reproduction capacity

Selection by fish for complex, hard-bottom habitats generally improved fitness in those species occupying these habitats. Enhanced foraging opportunities improved condition in cod and pouting when occupying wind farm areas (De Troch et al., 2013). However, this is not always the case. For example, Mathers, Houlihan & Cunningham (1992) examined growth rates in saithe using biochemical indicators and found no differences in growth rates on an oil platform as compared to the open sea. Further, while pouting condition was enhanced in a wind farm area relative to sandy areas, condition of cod was similar in both habitat types (Reubens, Degraer & Vincx, 2014). The similar growth rates or condition of fish on oil platforms or wind farms compared to unstructured areas suggests that although production was evident on a local scale (i.e., wind farm foundations), this may have had little effect on the larger spatial scale (i.e., the Belgian part of the North Sea). In a related example, oil and gas platforms in the North Sea provided unique feeding grounds for commercially important gadoids such as cod, saithe, and haddock (Fujii, 2016). For example, haddock predominantly consumed ophiuroids that occurred in high abundances on platforms at all depths, whereas temporal changes in the presence and abundance of saithe reflected the occurrence and availability of euphausiids (Fujii, 2016). These findings indicate that gadoids utilized the platform for foraging.

A tagging study revealed that spawning activities of cod in the Gulf of Maine occurred around specific bottom features, such as humps and ridges (Siceloff & Howell, 2013). Cod tended to move in groups after spawning, as 14 of 26 tagged and recaptured cod shifted to deep water (>100 m) within 23 to 50 days after tagging, with females tending to leave before males. While these findings indicate that the pelagic spawning of cod may be associated with large offshore hard-bottom structures, relatively shallow, complex habitats with rocks and cobble may enhance survival among the juveniles after benthic settling (Tupper & Boutilier, 1995).

A recent study highlighted the potential role of offshore oil and gas platforms as spawning sites for lumpsucker (Cyclopterus lumpus; Todd, Lavallin & Macreadie, 2018). C. lumpus was observed during different stages of reproduction on an offshore oil and gas structure in the North Sea, providing novel evidence of spawning offshore at depths of around 40–45 m. Examples from the literature also show positive relationships of hard-bottom habitats with spawning and recruitment for several fish species, with much of the relevant literature coming from natural hard-bottom habitats. In a survey of natural hard bottoms in western Norway, spawning of herring generally occurred on hard-bottom substrates down to a depth of 10 m, with little evidence of spawning on soft-bottom habitats (Johannessen, 1986).

In the Baltic Sea, the vegetation usually attached to hard substrates seems to play an important role for successful spawning of Baltic herring (C. harengus membras). In a survey of natural hard bottoms in the northern Baltic Sea (Sweden), herring eggs were only found on filamentous algae (Pilayella littoralis) attached to hard substrates (Aneer & Nellbring, 1982) and not on the surrounding hard or soft bottoms. Likewise, in the southwestern archipelago of Finland, Baltic herring spawning beds occurred in areas with vegetation growing on hard bottoms (down to a depth of 4 m); usually close to the regions’ deepest zones (down to 60 m depth; Kääriä et al., 1997). The complex habitat established by the algae may provide refuge from predators, as birds and other fish are responsible for a majority of the predation of herring eggs (Aneer & Nellbring, 1982). Areas without spawning beds generally had soft sediments with narrow bands of vegetation. Filamentous green algae (Cladophora sp.), brown algae (Ectocarpus sp., Pilayella sp.) and Potamogeton pectinatus made up the most common vegetation where divers observed eggs (Kääriä et al., 1997). For Baltic herring spawning on the Lithuanian coast (eastern coast of the Baltic), eggs were found on red algae Furcellaria lumbricalis and Polysiphonia fucoides, as well as hard habitat covered with mussels at 4–8 m depth (Šaškov et al., 2014). The dependence of Baltic herring on vegetation as substrate for the eggs may render this species vulnerable to pressures such as eutrophication that negatively affect macroalgal growth.

Modeling studies can be informative in assessing the effects of habitat on recruitment success. One model was used to assess the effects of MPAs on juvenile cod recruitment success (Lindholm et al., 2001). MPAs can maintain habitat complexity (i.e., more pebble-cobble substratum) compared to areas that are fished with dredges or trawls that damage the seafloor and reduce complexity (Ocean Studies Board & National Research Council, 2002). Modeled survival of juvenile cod on natural reefs showed that survival was density-dependent (i.e., affected by predation or cannibalism; Wikan & Eide, 2004) and increased with seafloor habitat quality (modeled as increased size of an MPA). Juvenile cod had higher survival rates at low densities and in large MPAs, thus linking juvenile cod survival to seafloor habitat complexity (Lindholm et al., 2001). Modeled improved survival associated with complex sea-floor habitats is consistent with previous laboratory studies and may explain why various fish species prefer habitats with gravel and cobble where they can find shelter in the crevices (Gotceitas & Brown, 1993; Christoffersen et al., 2018).

The review process that we used did not include all commercially important species, but rather was targeted towards species important for management by ICES, and we only examine temperate, non-biogenic reefs. The results from our study are limited to those specific species and the physical structure provided by hard-bottom, non-biogenic habitats. Our search terms were very specific and we acknowledge that this approach may have missed relevant research from habitats that we excluded. Therefore, interpretations of these data apply only to the species and habitats we incorporate.

Recent case studies

Although the literature for this study was not reviewed beyond those articles published in early April 2017, a related study reviewing methods that have examined the nursery role of different habitats included search terms relevant to hard-bottom habitats and fishery species of interest to ICES (Ciotti et al., in prep.). We highlight several recent case studies below that reflect this updated literature review. In the first example (Elliott et al., 2017), stereo baited remote underwater video (SBRUV) surveys were conducted in the nearshore, Scottish waters of the Firth of Clyde during summer 2013 and 2014 to determine the habitat of juvenile Atlantic cod, haddock and whiting. The fishery stocks of all three species declined in the late 20th century, and recruitment and spawning stock biomass remains relatively low in the Firth of Clyde despite efforts to re-build the stocks (Fernandes & Cook, 2013; Holmes et al., 2014; International Council for the Exploration of the Sea , 2016a; International Council for the Exploration of the Sea , 2016b; International Council for the Exploration of the Sea , 2016c). Understanding juvenile gadoid habitat is particularly important given that settlement and post-settlement survival is thought to be the best means to understanding gadoid population regulation (Olsen & Moland, 2011; Laurel, Knoth & Ryer, 2016). Atlantic cod were most abundant in shallow, sheltered areas composed of gravel −pebble containing maerl unattached coralline algae. Haddock and whiting predominated over deeper sand and mud. Relative abundances of Atlantic cod and whiting were positively related to the diversity of epibenthic and demersal fauna. Thus, spatial conservation measures to benefit demersal fish should consider habitat type and diversity (Elliott et al., 2017).

Another case study examined the environmental benefits of leaving de-commissioned offshore infrastructure, such as oil and gas platforms and wind turbines, in the ocean due to benefits such as biodiversity enhancement, provision of reef habitat, and protection from bottom trawling (Fowler et al., 2018). The rationale for this study was that the in situ ecosystem value of platforms and the negative impacts of removal are not factored into decommissioning decisions in regions such as the North Sea, where over 80% of oil structures are more than a decade old (OSPAR Commission, 2017) and are likely integrated to at least some extent into existing ecosystems (Fowler et al., 2018). The removal policy is based on the assumption that “leaving the seabed as you found it” will minimize negative impacts on the marine environment. A total of 200 experts spanning academia, government, and private organizations were sent surveys seeking their opinions on the following: (1) appropriateness of the current removal policy, (2) identification of viable alternatives to complete removal of offshore infrastructure, (3) key environmental considerations and trade-offs for decommissioning decisions, and (4) advice on decommissioning considerations between platforms and wind turbines. Of most relevance to this review of hard-bottom habitats, 78% of the 40 respondents thought that artificial habitats with environmental value should be maintained and protected due to enhanced ecosystem services (Fowler et al., 2018). Evidence includes harboring threatened species (Bell & Smith, 1999), providing reef habitat (Coolen et al., 2020), boosting recruitment of overfished species (Love et al., 2006), producing fish biomass at a greater rate than any other marine ecosystem (by as much as a factor of ten; (Claisse et al., 2014)), and acting as foraging sites for top-order predators (Todd et al., 2009). Fowler et al. (2018) concluded that the traditional view that artificial structures must be removed from marine ecosystems simply because they do not “belong” there has shifted to one of environmental optimization based on comparative assessment (Fowler, Macreadie & Booth, 2015). Each decommissioning option will have positive and negative impacts that must be carefully weighed, while also accounting for site-specific characteristics and the broader environmental context of the disturbance.