Subtidal hard marine substrate supports highly biodiverse assemblages of sessile invertebrates which interact through ecological succession over time. ARMs plate structures deployed at Dunwich, North Stradbroke Island over 12-month and 2-month time periods were extensively photographed and analysed for coverage of sessile invertebrates including porifera, solitary ascidians, colonial ascidians, crustaceans, molluscs, annelids and bryozoan. Multivariate analysis was performed to assess changes in community composition and demography over time which yielded a significant dissimilarity in community assemblages. Further data analysis showed that there were  significant preferences of community assemblage for the downwards facing sides of the ARMs plate and analysis by inveterate group across temporal levels yielded significant differences in mean plate coverage for solitary ascidians, porifera and crustaceans. The community assemblage at the 2-month time period was found to be undergoing primary succession whilst the community assemblage at the 12-month period was found to have a significant difference in biomass and a higher Shannon diversity index, indicating it is undergoing secondary succession. It was interesting to note the decrease in colonial ascidian plate coverage from 2-month time period to the 12-month time period, indicating likely post settlement mortality. Porifera were the most dominant invertebrate group on the 12-month ARMs plates, demonstrating their highly specialised competitive capabilities. Implications of poriferan ecological dominance in the benthic communities are wide and this dominance could pose risks to recovering benthic substrates from disturbances via invertebrate larval settlement inhibition. 

Marine benthic community composition and structure are the focus of many studies aiming to better understand succession and ecosystem responses to both biotic and abiotic disturbances. Subtidal marine benthic communities support extraordinarily biodiverse assemblages of corals, sessile invertebrates and macroalga, many of which are recruited through means of wide scale pelagic larval dispersal (Smale, 2012).This recruitment is the first stage in primary colonisation and succession of the hard substrata and is a process that is well understood. The following secondary succession and later development of sessile marine communities is less so (Nicoletti, Marzialetti, Paganelli, & Ardizzone, 2007) meaning that patterns of traditional succession in benthic assemblages are often disputed (Green & Edmunds, 2011) (Greene & Schoener, 1982) (Chang & Marshall, 2016). Despite several studies attempting to model ecological marine benthic succession ecology, the process is still poorly understood (Antoniadou, Voultsiadou, & Chinitrioglou, 2010).

Primary succession involves the recruitment and colonisation of pioneer species to free space on a substrate.The recruitment and early colonisation of the marine benthos is dominated by taxa which are rapid growing, have high fecundity and recruit easily to the hard surfaces, mainly by invertebrate adults releasing pelagic larvae (Green & Edmunds, 2011). Secondary succession is a continuum which reflects the processes associated survivorship and competitive ability (Greene & Schoener, 1982) (Nicoletti, Marzialetti, Paganelli, & Ardizzone, 2007). Later development phases of benthic communities have been shown to significantly modify the community assemblage and biodiversity (Antoniadou, Voultsiadou, & Chinitrioglou, 2010) beginning with pioneer species and diversifying into climax species (Reiss, et al., 2015). These communities are highly sensitive to disturbances (Roberts, Moloney, Sweatman, & Bridge, 2015) which makes understanding community succession important for management of benthos assemblages under increasing anthropogenic strain.  

Using artificial substrate has proven useful to cultivate benthic communities which provides novel insights into  community succession patterns through both a spatial and temporal lens (Smale, 2012).ARMs (Automated remote monitoring system) are one such artificial substrate which recruit benthic communities through passive collection (NOAA, 2017).The composition of the benthic community is known to change with time through succession, disturbances and environmental factors however quantitative models applied to this process are widely argued. The community succession process is generally characterised by an increase in biodiversity, organism abundance and total biomass (Antoniadou, Voultsiadou, & Chinitrioglou, 2010) (Mazlumyan, 2019). It can be also be described by a drop in the ratio of primary production to total biomass and further, a decrease in biodiversity as the community reaches a climax assemblage (Mazlumyan, 2019). Community equilibrium refers to the persistence of specific taxonomic assemblages of species (Rahel, 1990) however this is difficult to assess without ongoing monitoring of community composition.

For communities assessed that have developed for a short period of time, I predict that there a large area of free space with non-free space primarily colonised by ‘pioneer’ species including molluscs, colonial ascidians, encrusting bryozoans and annelids (Nicoletti, Marzialetti, Paganelli, & Ardizzone, 2007)(Luter, Duckworth, Wolff, Evans-Illidge, & Whalan, 2016). I expect that benthic communities which develop for a longer period of time will exhibit a higher biomass (Antoniadou, Voultsiadou, & Chinitrioglou, 2010) and a higher biodiversity with dominant taxa including porifera, solitary ascidians and encrusting byrozoans (Henschel, Cook, & Branch, 1990). Communities at different developmental stages are expected to be dissimilar in composition and biomass. Whether a community has reached equilibrium or not is difficult to distinguish and requires long term monitoring of changes to assemblage. Orderly succession (following of the defined criteria) would imply that traditional succession models apply to the development of sessile benthic communities (Bivens, 2018).

Field Study

ARMs structures, designed upon an internationally standardised system developed by the US National Oceanographic and Atmospheric Administration (NOAA) were deployed on North Stradbroke Island at One Mile Jetty, Dunwich (Figure 1) (27.49°S, 153.40°E).The ARMs plates (25 x 25 x 1 cm) were structured in stacks of 9 with a tenth larger top plate used to secure the structure to the jetty. One Mile Jetty, Dunwich is floating pontoon meaning the ARMs plates were deployed at a constant depth below sea level for the defined time periods. 3 ARMs structures were deployed for a time period of 12 months from 26^th March 2019 to 11^th March, 2020, referred to as ‘LONG’ plates and 3 ARMs structures deployed for 2 months from 22^nd January 2020 to 18^th March 2020, referred to  as ‘SHORT’ plates. Post structure retrieval, each plate exhibited passsive benthic collection (both upwards and downwards facing sides) was photographed extensively by The University of Queensland BIOL3211 students, tutors and lecturers. These photographs were been uploaded to Invertebrates of the Great Barrier Reef (GBRI) and were the primary data set used in the analysis (Figure 2).

Plate Image Analysis (Image J Fiji Software)

Initial observations of the plates photographed for both time periods showed there was high variability between the upwards and downwards facing side of the plates in terms of visual coverage as well as a stark contrast between the short and long plates (Figure 2).

Figure 2. Example of primary data settlement plates. DUN – refers to Dunwich. S/L refers to ‘Short’ or ‘Long’. 1.2 refers to the structure number and plate position.  B/ Down – Down facing side and T/ Up – Upward facing side. (A) DUNS 1.2 B (B) DUNS 1.2 T  (C) DUNL 1.2 DOWN (D) DUNL 1.2 UP. Clear differences in visible coverage and composition can be seen between short and long plates. (GBRI, 2020)

The photographs of the upwards and downwards facing sides of the settlement plates were analysed using image analysis software Image J, Fiji. One side length of the plate was calibrated to 25 cm and the perimeter of each benthic invertebrate was outlined manually (Figure 3). The area of each outlined shape was automatically calculated through the software. The images were converted to 8-bit gray scale and adjusted for contrast (Figure 3) to detect crustaceans and molluscs not obvious in the coloured photograph. The addition of grayscale also aided visual identification of boarders between overlapping invertebrates. Some issues arose when clear borders between invertebrates could not be discerned (as was expected with the quality of some photographs).

The area data from Image J Fiji was exported into an excel spreadsheet where the number for a specific area (Figure 3) was correlated to an identified invertebrate group.  The invertebrate groups analysed included bryozoan, porifera, colonial ascidians, solitary ascidians, annelids, molluscs and crustaceans. Free space area was also recorded however much of the ‘free space’ on tiles (particularly observed for the Dunwich short plates (Figure 2, B)) was covered in a microalga. These sessile invertebrate phyla were selected based on inferences that the photographs of settlement plates would make it difficult to measure the abundance of sedentary invertebrates accurately (such as echinoderms and some arthropods etc.).

Figure 3. Example of Image J Analysis Outline of Sessile Invertebrates in colour (DUNL3.2 D), 8-bit gray scale (DUNL2.2 D) and 8-bit gray scale with altered contrast  (DUNL 2.3 D) . The numbers (for example on DUNL 3.2 S: 1-21) link the specific area of an invertebrate to the image analysed for identification of the invertebrate group. These numbers are used to identify invertebrate taxa post area image analysis (Author, 2020).

Statistical Analysis

ANOVA was conducted in R Studio to assess interactions between the orientation of the plate and the position of the plate within the ARMs stack. Welch two sample t-tests assessed differences in the mean plate coverage between the upwards and downwards facing surfaces and differences in mean plate coverage between the two time periods. Welch two sample t-tests were also conducted over the mean plate coverage for each invertebrate group analysed.

A method of multivariate analysis was used to compare the similarity of the benthic communities between time periods, similar to a method (Antoniadou, Voultsiadou, & Chinitrioglou, 2010) and (Roberts, Moloney, Sweatman, & Bridge, 2015). Non-metric multidimensional scaling via Bray Curtis distances was applied to log transformed species plate coverage data. Bray-Curtis can be used to quantify the composition dissimilarity of different ecological sites.  A perMANOVA was used to test for differences in composition of the two communities. This assessesed whether the transformed distances differ between groups. This analysis was performed in R Studio using the ‘vegan’package and dependencies. 

Shannon- Weiner’s Index of biodiversity was also calculated across both time periods and averaged within the each time period. Graphs and plots were made in EXCEL and R Studio.

Settlement Preferences and Plate Position

A multiplicative ANOVA was conducted to determine if the settlement plate position within the ARMs stack interacted with the preferred orientation of community assemblage. The interaction term were found to be insignificant (F_(1,96) = 0.108, p > 0.05) and the plate position term insignificant (F_(1,96) = 0.108, p >0.5).The settlement orientation was found to be significant  (F_(1,96) = 71.77, p< 0.001).

A welch two sample t-test showed the benthic community assemblages had a significant preferences for the downwards (ocean floor) facing side of the settlement plate for both temporal variations (df = 70.263 at 95% CI, t =8.55, p<<0.001) seen in Figure 4. The mean plate coverage for the downwards facing side of the settlement plate is 266.64cm² (n =50), approximately 43% of the plate surface, and the mean for the upwards facing side of the settlement plate is 61.62cm² (n = 50), a little under 10% of the ARMs plate.

Figure 4 Boxplot of Mean Area Coverage of ARMs Plates on for downward facing side (mean =266.64cm², sd =151.38 , n = 50) and upward facing sides (mean =61.62cm² , sd = 103.37, n =50 ) (Author, 2020). 

The same analysis repeated for individual temporal variations which both exhibit similar trends. The community assemblage preferences for the downwards facing side of the plate is significant for the 12 month period (df = 42.43 at 95% CI, t =10.63,p<<0.001) and the 3 month period (df = 24.29 at 95% CI, t = 6.24, p <<0.001).The mean for the long time period downwards facing side, is 370.74cm² (sd = 96.73, n = 25) and the upwards facing side 116.35cm² (sd = 66.17, n = 25).These are higher than the short time period averages which had a downwards facing side mean coverage of 162.54cm² (Sd = 121.66, n = 25) and an upwards facing side mean coverage of 6.88cm² (sd = 9.59,  n = 25).

Temporal Variability

There is a significant difference in the mean coverage of the ARMs plates for the long time period and short time period (df = 91.74 for 95% CI, t = 5.81, p <<0.001). The mean plate coverage for the long time period is 243.43cm² (sd = 151.81, n = 50) and the mean plate coverage for the short time period is 84.71cm² (sd= 116.21, n = 50) (Figure 5).

Figure 5 Boxplot of Mean Area Coverage of ARMs plates for the 12 month deployment period (243.43cm², sd = 151.81, n = 50) and 2 month deployment period (84.71cm² , sd= 116.21, n = 50)) (Author, 2020).

The mean plate coverage for each invertebrate group was also analysed via welch two sample t-test between the two different time periods. The mean plate coverage's are plotted into a barplot (Figure 6) where porifera (df = 52.22 at 95% CI, t = 7.68, p<<0.001), solitary ascidians (df = 52.31 at 95% CI, t = 6.86, p<<0.001) and crustaceans (df = 58.89 at 95% CI, t = -3.96, p<<0.001) all showed significant differences. Colonial ascidians, bryozoans, annelids and bivalves did not display significant differences in mean plate coverage between the two ARMs deployment times. Mean free space on the ARMs plates between deployment time was also found to be significant through a welch two sample t-test (df = 91.75 at 95% CI, t = -5.82, p <<0.001).The high standard errors are due to variability between the upwards and downwards facing sides of the plates, low sample sizes (3 ARMs structures for each time period) and introduced error associated with image quality of primary data.

Figure 6 Bar graph of Mean Area Coverage for long immersion period (dark grey) and short immersion period (white) by invertebrate taxa including Porifera (Long: 133.67cm², sd= 108.70  Short: 12.36cm², sd= 19.71 ), Bryozoa (Long: 15.90cm², sd =  27.809 Short: 8.03cm², sd =17.28 ),Colonial ascidians (Long: 30.35cm², sd = 41.23  Short: 45.46cm², sd =91.3 ),Solitary ascidians (Long: 55.05cm², sd =51.0  Short:4.13cm², sd =9.4), Molluscs (Long:6.60cm², sd =12.5  Short:5.4cm², sd =8.3 ), Crustaceans (Long: 1.33cm², sd =3.8  Short: 8.5cm², sd = 12.0 )and Annelids (Long: 0.63cm², sd =2.3  Short: 0.84cm², sd =2.6 ).  (Author, 2020).

Succession Patterns

Analysis of sessile community composition through a perMANOVA showed a significant difference between the community composition of the 12 month ARMs deployment and the 3 month ARMs deployment ( F_(1,82) = 26.50, p<0.05). Plotting the distances of the transformed non-metric dimensions (Figure 7) demonstrates visually the differences in community composition between the two time periods at the one site.

Figure 7.Bray Curtis Distances of log transformed for long (L) and short (S) ARMs plate mean coverage data (method = BRAY, F_(1,82) = 26.50, p<0.05)(Author, 2020).

Shannon-Weiner Diversity Index for the two different time periods – the number of ‘individuals’ was defined as the number of outlined invertebrates – was also calculated. For colonial organisms an ‘individual’ was defined as the community of zooids. The Shannon Diversity Index for both long and short plates are 0.89±0.42 and 0.56±0.58 respectively.

Community Settlement Preferences

Across both ARMs plate deployment time frames, invertebrate recruitment showed a clear preference for the downwards facing side of the plates with analysis yielding a significant difference in the mean plate coverage (Figure 4). This fulfils expected orientation preferences since a large proportion of invertebrate larvae settlement is mediated by negative phototactic and geotactic behaviour. The underside of the ARMs structures provide shelter away from light exposure and is likely to bring exposure to higher nutrient content through the water column via upwelling. This means greater food sources to sustain maturing sessile communities, majority of which are filter or suspension feeders. Several other hypotheses proposed to explain this observation include differences in light intensity or quality, sedimentation and grazing by predators (Adjeroud, Penin, & Carroll, 2007). It was also predicted that the position of the ARMs plate within the structure (1-9 plates) would affect the mean plate coverage however analysis showed this to be an insignificant factor. This indicates that environmental conditions did not vary dramatically between the ARMs plates within a structure. This outcome is not found to be influential in the composition of the benthic communities but moreover an insight into assemblage locations and abiotic factors that can affect them.

Temporal Variation and Community Succession

The temporal variation between the long and short ARMs plates yielded a significant difference in mean plate coverage with the longer time period having a greater mean coverage (Figure 5 and Figure 6). This temporal variation meant, on a base level, a greater probability of invertebrate larvae recruitment from the water column as well as time for growth, metamorphosis and maturation of settled invertebrate species. The higher plate coverage on the long ARMs plates can be linked to a higher biomass compared to the short plates, a trait which is indicative of secondary community succession and a change in the assemblage demographic (Antoniadou, Voultsiadou, & Chinitrioglou, 2010). Further analysis of specific invertebrates present on the long and short  ARMs plates identifies these demographic changes.

The short plates exhibited significantly higher mean plate coverage of crustaceans and a higher mean plate coverages of colonial ascidians (Figure 6). Both of these invertebrates are described as pioneer species (Henschel, Cook, & Branch, 1990) which indicates the benthic community at 2 months on NSI is transitioning through primary succesion. Molluscs (specifically bivalves), encrusting bryozoans and macroalga (Henschel, Cook, & Branch, 1990) are also defined as pinoneer invertebrate groups and are present on the short ARMs plates at NSI, however the mean plate coverages were not significantly different over the temporal variation. The high coverage of colonial ascidians can be attributed to characteristics such as rapid growth rates, early sexual maturity and high fecundity which make them excellent colonisers (Chadwick & Morrow, 2010). The short plates also had significantly greater mean free space – much of which was covered with macroalgae (Figure 2 (B)) – and demonstrates a developing community structure (Nicoletti, Marzialetti, Paganelli, & Ardizzone, 2007) since the free space has not yet been claimed.

The long plates exhibited significantly higher mean plate coverage of solitary ascidians and porifera (Figure 6) whilst still exhibiting higher mean plate coverages of bryozoa, molluscs and annelids (Figure 6). The high coverage of porifera is indicative of the benthic community transitional process from a pioneer species assemblage to intermediate species assemblage since porifera are slow growing  and rely heavily on their competitive ability to claim space on hard substrates (Brandt, Olinger, Chaves-Fonnegra, Olson, & Gochfeld, 2019).The high solitary ascidian coverage is explained in their ability to self-recruit and in doing so form multilayer assemblages of individuals (Hawkins, et al., 2017). The ecological dominance of these two groups is driven by high rates of self-recruitment and overgrowth ability (Hawkins, et al., 2017).

The decrease in mean plate coverage of colonial ascidians and crustaceans between the long and short ARMs plates could indicate post settlement mortality due to the high coverage and dominance of Porifera on the long plates (Henschel, Cook, & Branch, 1990). Colonials ascidians lack defences or means to avoid allelochemical offenses such as those secreted by sponges (Chadwick & Morrow, 2010). This is why invasive colonial ascidian species thrives under minimal competition (short ARMs plates) but not when placed in competition with a higher diversity of sessile invertebrates (long ARMs plates). No significant difference was identified in the mean plate coverage of bryozoans between the long and short plates however this is a species that had been identified as both a dominant pioneer and intermediate invertebrate and as such could not have been dramatically affected by a shift in demography on the plates. This result could also be attributed to difficulty in primary data analysis where some photographs lacked clarity particularly for bryozoan invertebrates. These differences in composition between the long and short settlement plates indicate that there has been ecological succession over time in the marine benthic community at NSI. The significant difference in mean plate coverage indicates an increased biomass and the Shannon-Weiner diversity index shows a higher biodiversity for the long ARMs plates compared to the short. 

Community Composition

Analysis of the settlement plate’s benthic community composition through a perMANOVA showed that the structure and composition of the benthic communities are significantly dissimilar between ARMs plate deployment time periods. This means that the community composition at 2 months is significantly different to the community composition at 12 months in the same location at NSI (Figure 7). This finding, coupled with an increase in biomass between the two time periods and increase in biodiversity, indicates that the benthic community in Moreton Bay follows traditional succession patterns: the community transitions from pioneer species through secondary development succession to intermediate assemblages and then climax assemblages (Green & Edmunds,2011).  Diminished coverage of first colonisers (colonials ascidians and crustaceans) on the long ARMs plates further support  this. Since the community did not exhibit any decreases in biodiversity, it can be tentatively ineferred that the community assemblage had not yet reached a climax community equilibrium (Nicoletti, Marzialetti, Paganelli, & Ardizzone, 2007) which, granted, cannot be  accurately assessed for the ARMs plates on NSI without more extensive monitoring. Benthic communities can be characterised by dynamic structural changes with immigration and mortality throughout time (Antoniadou, Voultsiadou, & Chinitrioglou, 2010) which is echoed in findings at Dunwich, NSI.

Recruitment Rate

Recruitment to settlement plates is a topic well studied at different locations with many studies focusing on both the temporal and spatial variability. The average invertebrate recruitment rate to the ARMs plates substrate over the 12-month period on North Stradbroke Island was found to be 263.36 recruits/m²year ± 132 recruits/m²year (calculated from settlement plates deployed for 12 months).  This is comparable to the recruitment rates found in other studies at different locations. Namely, recruitment rates were averaged to 40.77 recruits/m²year in Moorea, French Polynesia (Adjeroud, Penin, & Carroll, 2007) in a study conducted with similar settlement tiles and another study site, at approximately the same latitude as Moorea, on the Central GBR had recruitment rates up to 4590 recruits/m²year (Hughes, Baird, Dinsdale, & Moltschaniwskyj, 1999). The NSI site is more southern than that of the other study sites and is also not located on an established coral reef where there is a greater abundance of dispersed invertebrate larvae. The reasonable recruitment rate to the ARMs plates at Dunwich, NSI is most likely due to the high number of vessels in Moreton Bay which foul with similar benthic communities or from larvae dispersed from close reefs (such as Flinders Reef, directly north of Moreton Island).

Relevance of Benthic Community Succession

A further application of better understanding the community ecology and succession processes of benthic invertebrate assemblages is their direct influence on reef building coral recruitment. The mean coverage of Porifera on the long ARMs plates was significantly different to the short plates (Figure 6) indicating sponge's advanced capabilites to claim and defend space on the substrata. The specialised ability of sponges to secrete potent allelochemicals and, in some species, toxic mucus via direct contact or into the water column make them highly competitive (Chadwick & Morrow, 2010). In lieu of this, some studies have shown that sponge species can limit or inhibit coral recruitment by pre-empting space for coral larval recruits and aggressively over-growing polyps (Brandt, Olinger, Chaves-Fonnegra, Olson, & Gochfeld, 2019) (Dunstan & Johnson, 1998). As increasing anthropogenic strain is placed on coral reefs, changes to the relationship between 'bottom-up' sessile community development and coral recruitment are unclear.

Some papers addressing shifts towards the dominance of non-reef building invertebrates on the marine benthos (Chadwick & Morrow, 2010) highlight risks this can pose to community structure and its negative impact on coral reef recovery after disturbances (Glassom, Zakai, & Chadwick-Furman, 2004). These benthic communities are similar to those addressed at NSI which show that free substrate is rapidly colonised by pioneer invertebrate species (and algae) which develop into a highly competitive benthic community (McManus & Polsenberg, 2004). This process, if occurring on a wider scale could cause a shift in the equilibrium composition of the marine benthos towards a dominance of porifera on hard substrates. Thus, it is clear that there is rationale to further monitor and investigate the succession processes in sessile marine invertebrate communities.

Adjeroud, M., Penin, L., & Carroll, A. (2007). Spatio-temporal heterogeneity in coral recruitment around Moorea, French Polynesia: Implications for population maintenance. Journal of Experimental Marine Biology and Ecology, 341(2), 204-218. doi:https://doi.org/10.1016/j.jembe.2006.10.048

Antoniadou, C., Voultsiadou, E., & Chinitrioglou, C. (2010). Benthic colonization and succession on temperate sublittoral rocky cliffs. Journal of Experimental Marine Biology and Ecology, 382(21), 145-153. doi:https://doi.org/10.1016/j.jembe.2009.11.004

Bivens, A. (2018). Successional Processes in the Benthic Invertebrate Communities at Gray's Reef National Marine Sanctuary. Georgia Southern University. Retrieved from https://digitalcommons.georgiasouthern.edu/honors-theses/367

Brandt, M., Olinger, L., Chaves-Fonnegra, A., Olson, J., & Gochfeld, D. (2019). Coral recruitment is impacted by the presence of a sponge community. Marine Biology, 166. doi:https://doi.org/10.1007/s00227-019-3493-5

Chadwick, N., & Morrow, K. (2010). Competition Among Sessile Organisms on Coral Reefs. Coral Reefs: An Ecosystem in Transition, 347-371. doi:https://doi-org.ezproxy.library.uq.edu.au/10.1007/978-94-007-0114-4_20

Chang, C.-Y., & Marshall, D. (2016). Spatial Pattern of distribution of marine invertebrates within a subtidal community: do communities vary among pathes or plots? Ecology and Evolution, 6(22), 8330-8337. doi:doi: 10.1002/ece3.2462

Derse Crook, E., Koeker, K., Potts, D., Rebolledo-Vieyra, M., Hernandez-Terrones, L., & Paytan, A. (2016). Recruitment and Succession in a Tropical Benthic Community in Response to In-Situ Ocean Acidification. PLOS ONe. doi:https://doi.org/10.1371/journal.pone.0146707

Dunstan, P., & Johnson, C. (1998). Spatio-temporal variation in coral recruitment at different scales on Heron Reef, southern Great Barrier Reef. Coral Reefs, 17, 71-81. doi:https://doi-org.ezproxy.library.uq.edu.au/10.1007/s003380050098

GBRI. (2020). Great Barrier Reef Invertebrates. Retrieved from ARMS Plates: https://www.gbri.org.au/

Glassom, D., Zakai, D., & Chadwick-Furman, N. (2004). Coral recruitment: a spatio-temporal analysis along the coastline of Eilat, northern Red Sea. Marine Biology, 144, 641-651. doi:https://doi-org.ezproxy.library.uq.edu.au/10.1007/s00227-003-1243-0

Green, D., & Edmunds, P. (2011). Spatio-temporal variability of coral recruitment on shallow reefs in St. John, US Virgin Islands. Journal of Experimental Marine Biology and Ecology, 397(2), 220-229. doi:https://doi.org/10.1016/j.jembe.2010.12.004

Greene, C., & Schoener, A. (1982). Succession on Marine Hard Substrata: A Fixed Lottery. Oecologia, 55, 289-297.

Hawkins, S., Evans, A., Dale, A., Firth, L., Hughes J, & Smith, I. (2017). Ecological Dominance Along Rocky Shores, With Focus on Intertidal Ascidians. 55.

Henschel, J., Cook, P., & Branch, G. (1990). The colonization of artificial substrata by marine sessile organisms in False Bay. 1. Community development. South African Journal of Marine Science, 9(1), 289-297. doi: https://doi.org/10.2989/025776190784378664

Hughes, T. P., Baird, A. H., Dinsdale, E. A., & Moltschaniwskyj, N. A. (1999). Patterns of recruitment and abundance of corals along the Great Barrier Reef. Nature, 397(6714), 59-63. doi:DOI:10.1038/16237

Luter, H., Duckworth, A., Wolff, C., Evans-Illidge, E., & Whalan, S. (2016). Recruitment Variability of Coral Reef Sessile Communities of the Far North Great Barrier Reef. PLOS One, 1-16. doi:DOI:10.1371/journal.pone.0153184

Mazlumyan, S. (2019). Primary and Secondary Succession Modelling in Marine Bottom Biotopes (Black. International Journal of Marine Science, 9(8), 66-76. doi:doi: 10.5376/ijms.2019.09.0008

McManus, J., & Polsenberg, J. (2004). Coral–algal phase shifts on coral reefs: Ecological and environmental aspects. Process in Oceanography, 60(2-4), 263-279. doi:https://doi.org/10.1016/j.pocean.2004.02.014

Nicoletti, L., Marzialetti, S., Paganelli, D., & Ardizzone, G. (2007). Long-term changes in a benthic assemblage associated with artificial reefs. Hydrobiologia(580). doi:https://doi-org.ezproxy.library.uq.edu.au/10.1007/s10750-006-0450-3

NOAA. (2017). Global ARMs Program. Retrieved from National Oceanographic and Atmospheric Administration: https://www.oceanarms.org/

Rahel, F. (1990). The Hierarchial Nature of Community Persistence: A Problem of Scale. The American Naturalist, 136(3), 28-344.

Reiss, H., Birchenough, S., Borja, A., Buhl-Mortensen, L., Craeymeersch, J., Dannheim, J., . . . Neumann, H. (2015). Benthos distribution modelling and its relevance for marine ecosystem management. ICES Journal of Marine Science, 72(2), 297-315. doi:https://doi.org/10.1093/icesjms/fsu107

Roberts, T., Moloney, J., Sweatman, H., & Bridge, T. (2015). Benthic community composition on submerged reefs in the central Great Barreier Reef. Coral Reefs, 569-580.

Smale, D. (2012). Extreme spatial variability in sessile assemblage development in subtidal habitats off southwest Australia (southeast Indian Ocean). Journal of Experimental Marine Biology and Ecology, 438, 76-83. doi:https://doi.org/10.1016/j.jembe.2012.10.002