The morphology of macroalgal substrates can help predict the attachment of juvenile mussels

The morphology of macroalgal substrates can help predict the attachment of juvenile mussels | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article The morphology of macroalgal substrates can help predict the attachment of juvenile mussels Katherine A. Burnham, Jenny R. Hillman, Andrew G. Jeffs This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6745848/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Understanding and addressing substrate availability has been shown to be critical for facilitating recruitment in shellfish reef restoration initiatives globally. For many mussel species, macroalgae is vital for the settlement and establishment of juveniles on reefs, but the factors which determine a macroalga’s effectiveness as an attachment substrate are poorly understood. This study aimed to; 1) develop an index that can be used by restoration practitioners to score the morphological features of macroalgae for their potential to support juvenile mussel recruitment and 2) test the accuracy of this index on the macroalgal substrate attachments of different juvenile size classes of the green-lipped mussel Perna canaliculus on two remnant mussel reefs in northeast New Zealand. Eight morphological features of macroalgae, identified from published studies, were used to create the Macroalgal Morphology Index (MMI). The index scoring criteria for each of these features were able to predict the likelihood of P. canaliculus attachments to macroalgae for juveniles < 10 mm in shell length with 75% accuracy but with only 40–60% accuracy for juveniles 10 – <30 mm. Holdfast complexity and canopy cover were the two most useful features of macroalgae for predicting the attachment of all sizes of juvenile mussels. Meanwhile, the four features of macroalgae that describe their branching morphology were only strong predictors of attachment for juveniles < 10 mm. Overall, these findings suggest that the MMI can aid restoration practitioners in the selection of suitable macroalgal substrates for facilitating juvenile recruitment at mussel reef restoration sites in New Zealand and potentially elsewhere. Figures Figure 1 Figure 2 Figure 3 Figure 4 Introduction In many parts of the world, vast areas of the seafloor that once supported thriving shellfish reefs have been dramatically altered by overharvesting, destructive fishing practices that removed the benthos, and habitat degradation from coastal development that has intensified over the last century (Lotze et al. 2006 ; Beck et al. 2009 ; Ford and Hamer 2016 ). The removal of this food source and structurally complex biogenic habitat can result in not only a loss of biodiversity in the immediate area (McLeod et al. 2014 ; Sea et al. 2022 ; Benjamin 2022a) but also loss of the valuable ecosystem services that these reefs provide, such as water filtration (Suplicy 2020 ; van der Schatte Olivier et al. 2020 ), nitrogen removal (Hillman et al. 2021; Sea et al. 2021), wave energy dissipation (Donker et al. 2013 ), and soft-sediment stabilization (Meadows et al. 1998 ; Brumbaugh et al. 2006 ). Historically in New Zealand or Aotearoa (Māori), reefs of Perna canaliculus , commonly known as green-lipped mussels or kūtai (Māori), were a common and highly valued feature of coastlines, stretching from the intertidal down to a depth of about 30 m (Paul 2012 ). The Hauraki Gulf or Tīkapa Moana / Te Moananui-ā-Toi (Māori), a large coastal embayment bordering the city of Auckland or Tāmaki Makaurau (Māori), once supported a 1,300 km 2 area of P. canaliculus reefs (Greenway 1969 ; Reid 1969 ). However, these mussel reefs, and others around the country, were decimated by an intensive benthic dredge fishery that had collapsed by 1969 (Paul 2012 ), leaving only a tiny portion of primarily intertidal, remnant mussel reefs (McLeod 2009 ; Toone et al. 2023a ). Thirty years later, areas of the Hauraki Gulf seafloor that were once dominated by mussels were predominantly characterised by layers of sandy or calcareous mud (Manighetti and Carter 1999 ), and decades of increased sedimentation from poor management of land-use have sustained these conditions into modern times (Morrison et al. 2012). These fine sediments are highly prone to resuspension during storm events (Hauraki Gulf Forum 2023 ), and low light penetration from higher turbidity and a lack of hard attachment surfaces contribute to a limited presence and diversity of macroalgae that would have been associated with mussel reefs in the past (Rowden et al. 2012 ; Lao 2016 ). Large-scale shellfish reef restoration efforts that return adult shellfish, often raised in aquaculture settings, to natural habitats in their historical ranges show significant promise for the future recovery of these ecosystems and their functional value across the globe (McCay et al. 2003 ; Carranza and Zu Ermgassen 2020 ; Fitzsimons et al. 2020 ), with transplanted adults demonstrating that long-term survival is possible even in degraded environments (Benjamin et al. 2023 ; McAfee et al. 2024 ; Overton et al. 2024 ). Adult P. canaliculus transplanted into soft-sediment intertidal and subtidal habitats show survival at least one to two years post-deployment even under adverse modern-day conditions in the Hauraki Gulf (McLeod et al. 2012 ; Wilcox et al. 2018 ) and elsewhere around New Zealand (Benjamin et al. 2023 ; Toone et al. 2023b ). However, recruitment, or the settlement of larvae and establishment of juveniles at a location, is severely limited within these restored adult P. canaliculus populations (Wilcox et al. 2018 ; Benjamin et al. 2022b , 2023 ; Toone et al. 2023a , b ) and within other areas of the world that supported wild mussel populations prior to their decimation by overharvesting (Beukema and Cadée 1996 ; Eriksson et al. 2010 ), indicating that early juvenile mussels may be negatively impacted, either directly or indirectly, by the degraded conditions (Eriksson et al. 2010 ; McLeod et al. 2012 ; Alder et al. 2020 ; Toone et al. 2023a ). It has been hypothesized that a lack of suitable settlement substrates in these highly sedimented environments is the primary cause of the recruitment bottleneck (McLeod et al. 2012 ; van der Heide et al. 2014 ; Wilcox et al. 2018 ; Banke et al. 2025). Across many regions worldwide, the degradation of coastal habitats following the decline of shellfish populations has made benthic substrate limitation one of the most common issues affecting shellfish restoration today, often placing it at the forefront of restoration priorities (Beck at al. 2009; Brumbaugh and Coen 2009 ; Fitzsimons et al. 2020 ). A challenge in identifying suitable substrates for shellfish reef restoration is the scarcity of healthy reference ecosystems and baseline environmental data available to restoration practitioners that details the type of benthos and associated conditions that historically supported shellfish populations prior to their depletion (Gann et al. 2019 ; McAfee and Connell 2020 ). As a result, in many locations around the world, it remains unclear which substrates were crucial for the functioning of recruitment processes on mussel reefs that may now be absent from restoration sites. Ontogenic changes in the substrate preferences of mussels adds another layer of complexity to overcoming the barrier of substrate limitation at mussel reef restoration sites. The larvae of many species of mussels, including P. canaliculus , are known to initially settle onto filamentous substrates, such as hydroids, seagrasses, and macroalgae (Bayne 1964 ; Seed and Suchanek 1992 ; Buchanan and Babcock 1997 ). Subsequently, the juvenile mussels relocate, by means of pedal crawling and mucous-drifting (Bayne 1964 ; Buchanan and Babcock 1997 ), onto hard substrates located among adult mussel populations, attaching preferentially to conspecifics, and also onto shell and rock (Commito et al. 2014 ; Wilcox and Jeffs 2017 ; Burnham et al. 2025 ). Therefore, while the addition of adult mussels and shell material at restoration sites may be enough to support the establishment of late juvenile and mature mussels, it is likely that facilitating the recruitment of early juvenile mussels requires the addition of filamentous substrates. This can be achieved by provisioning suitable macroalgal substrates at mussel reef restoration sites or by selecting new restoration sites where suitable macroalgal substrates are already present (Fitzsimons et al. 2019 ; Toone et al. 2023c , d ). However, restoration practitioners must firstly determine which species of macroalgae are associated with high levels of recruitment in their target mussel species and whether the suitability of a macroalgal substrate is determined by its species-specific chemical composition or its physical traits. For P. canaliculus juveniles, field research has revealed greater settlement onto red macroalgae species (i.e., phylum Rhodophyta) that exhibit a high degree of branching along their thalli and thinner branch widths (Buchanan and Babcock 1997 ; Alfaro and Jeffs 2002 ; Alfaro et al. 2004 ). However, the degree to which recruitment of P. canaliculus is dependent on these types of macroalgae is unclear, especially in the Hauraki Gulf, where there has been little field research on this topic due to the near-complete loss of wild mussel reefs. Identifying the species and morphological characteristics of macroalgae that most consistently support the highest attachment of juvenile P. canaliculus across different remnant mussel reefs in the Hauraki Gulf would help to clarify which of these aspects of macroalgae plays a more important role in the recruitment of P. canaliculus . While the morphology of the macroalgal substrates appears to be an important determinant of juvenile P. canaliculus attachment (Alfaro and Jeffs 2002 ; Alfaro et al. 2004 ; Loucks 2023 ), the unique chemical signatures of different macroalgae species have also been implicated in promoting mussel attachment (Alfaro et al. 2006 ; Gribben et al. 2011 ). For example, the isolated chemical extracts of species of red and brown macroalgae (i.e., phylum Phaeophyta) have been shown to induce settlement in P. canaliculus (Kelly 2001 ; Alfaro et al. 2006 ; Gribben et al. 2011 ). However, when these algal extracts were applied to substrates with varying surface textures in a laboratory study, substrates with rough surfaces experienced significantly higher settlement of P. canaliculus than substrates with smooth surfaces regardless of the chemical cue (Gribben et al. 2011 ), suggesting that the substrate’s physical structure, rather than its chemical signature, plays a dominant role in juvenile mussel attachment. Furthermore, while the chemical cues of different species of macroalgae may attract varying degrees of larval mussel settlement (Dobretsov 1999 ; Dobretsov and Wahl 2001 , 2008 ), it is the morphology of the substrates that can subsequently determine the post-settlement retention and survival of juvenile mussels on reefs (Moreno 1995 ; Frandsen and Dolmer 2002 ; Brenner and Buck 2010 ). Laboratory experiments that use plastic analogues of macroalgae to isolate the role of substrate morphology in mussel recruitment reveal that certain branching characteristics can influence the size, settlement densities, and small-scale spatial distribution of juvenile P. canaliculus (Kelly 2001 ; Alfaro and Jeffs 2002 ; Wu 2018 ; Wu et al. 2025 ). However, field studies that characterise mussel settlement onto macroalgal substrates often group macroalgae species into functional categories under certain morphological terminology, such as “filamentous” (Bayne 1964 ; Hunt and Scheibling 1995 ; Yang et al. 2007 ), “fine branching/finely branched” (Buchanan and Babcock 1997 ; Alfaro and Jeffs 2002 ; Alfaro at al. 2004), or “densely branching” (Loucks 2023 ), for which the specific physical characteristics that define those categories are not always clearly described. Additionally, the filamentous or finely branching macroalgae species highlighted in these studies often simultaneously exhibit narrow branch widths, a high frequency of branching, and a short distance between branches, making it difficult to isolate which of these individual characteristics promotes mussel attachment (Wu 2018 ). Identifying which morphological features of macroalgae impact mussel recruitment and defining these features under a single set of unifying descriptors and objective measurements is necessary to help guide the selection of effective substrates for future mussel reef restoration efforts. Thus, the macroalgal substrate associations of juvenile P. canaliculus at different stages of development were characterised on two remnant, intertidal mussel reefs in the Hauraki Gulf. This study aimed to; 1) create an index that scores the morphological features of macroalgae for their potential to support juvenile mussel recruitment based on prior published data, 2) compare the relative performance of various macroalgae species with differing morphology as attachment substrates for three size classes of juvenile P. canaliculus on two remnant mussel reefs, and 3) use these findings to evaluate the accuracy with which the index can predict the presence of attached juvenile mussels among macroalgal substrates based on the morphology of the macroalgae. Additionally, to aid an understanding of why certain morphological characteristics of macroalgae support the presence of juvenile P. canaliculus at different ontogenic stages, the location of attachment within each macroalga species was also characterised for each mussel size class. The goal of this study was to develop and test a tool that has the potential to aid restoration practitioners in the selection of suitable natural substrates or the informed design of artificial substrates for overcoming substrate limitation at mussel reef restoration sites in New Zealand, with broader applications for mussel reef restoration efforts in other locations around the globe. Materials and methods Macroalgal Morphology Index A comprehensive search of the literature was conducted using the citation database Scopus (Elsevier 2025 ) to identify previous published studies that analysed or discussed impacts of specific structural properties of natural or artificial substrates on the recruitment, settlement, attachment, growth, or survival of any mussel species. The following keywords were used in the Scopus search string: (“mussel” OR “mussels”) AND (“attachment” OR “recruitment” OR “settlement” OR “survival” OR “growth”) AND (“substrate” OR “macroalgae” OR “algae” OR “macrophyte”) AND (“morphology” OR “structure” OR “features” OR “characteristics” OR “properties” OR “complexity” OR “heterogeneity” OR “cover”). Studies that only tested the structural complexity of live adult mussel substrate, such as mussel patch size and distribution, were excluded from the search because of difficulty isolating the effects of these structural features from the positive effect of conspecific presence on mussel attachment (Commito et al. 2014 ; Wilcox and Jeffs 2017 ). The studies identified from the literature search were screened to identify the physical features of the substrates to which mussels attached. Those studies that discussed a similar physical feature of the attachment substrate, regardless of the variation in the terminology used to describe the feature, were grouped together under the morphological equivalent of that feature on macroalgae (e.g., rope filament thickness was categorised under branch width; Table 1). Each of these overlying morphological features identified from the literature was considered a potential determinant of juvenile mussel attachment, and these were the features used to create the Macroalgal Morphology Index (MMI). Within the MMI, physical variations of each morphological feature of macroalgae were sorted into three levels and assigned a score for their reported ability to support juvenile mussel recruitment, in which a score of one is considered not advantageous, two is moderately advantageous, and three is highly advantageous. Corresponding qualitative criteria were generated for these scores based on the physical descriptions of substrates in the literature. Quantitative criteria were also generated from measurements of each morphological feature collected from at least two individuals of each species of macroalga at each of the two sites sampled in this study (see Fig. 1 for a diagram the specific measurements collected for each morphological feature). For each macroalga species, measurements were averaged to produce one value for each morphological feature, and the mean measurements for each feature were compared among species and cross-referenced with qualitative criteria to generate quantitative ranges for each of the three scored levels of the MMI. For each species of macroalga, scores for each feature were totaled to obtain a single, overall score representing the potential that each macroalgal substrate has to support juvenile mussel recruitment. Sampling sites Reef sampling was conducted at two sites on two of the last known remaining intertidal P. canaliculus reefs in the northern Hauraki Gulf on the North Island of New Zealand. The first site was at Waipū Cove, which is characterised by low levels of sedimentation, moderate levels of human harvesting pressure, and high wave exposure, although ample shelter from wave action can be found among the three-dimensionally complex rock formations. The main mussel reef at this site is approximately 68 m wide by 208 m alongshore, and the majority of adult mussel aggregations extend from approximately 1.5 m above the tide chart datum to 1.5 m below it. The second site was at Pākiri Beach, which is characterised by low levels of sedimentation, protection from harvesting since October 2023, and high wave exposure with minimal shelter from wave action along relatively flat shelves of bedrock. The main mussel reef at this site is approximately 25 m wide by 93 m alongshore, and the majority of adult mussel aggregations extend from approximately 1.5 m above the tide chart datum to 0 m. Both reefs are situated on sandstone bedrock emerging from coarse sand and shell hash, and both sites have limited, adjacent but not contiguous subtidal rocky habitat consisting of emergent ledges populated with macroalgae and some aggregations of mussels. Mussel-macroalgae associations Between May 2023 and June 2024, sampling was conducted in the intertidal during low tide at each site, in which a series of 13 m alongshore transects were laid across the width of the mussel reef to ensure that a range of available macroalgae substrates were sampled for attached juvenile P. canaliculus (< 30 mm in shell length; Alfaro et al. 2001 ). At 0.5 m intervals along each transect, 10 × 10 cm quadrats were placed over the substrate to standardise the area of macroalgae searched for cryptic juvenile mussels (see methods in Erlandsson and McQuaid 2004 ; Burnham et al. 2025 ), resulting in 26 quadrats per transect. Thirteen transects were sampled at Waipū Cove from May to July 2023 and February to June 2024, while only eight transects were sampled at Pākiri Beach from October 2023 to February 2024 because it contains a greater diversity of macroalgae concentrated on a smaller area of reef. The presence of all macroalgal substrates within each quadrat were recorded, except for epiphytic macroalgae growing on mussels or other algae, encrusting algae (e.g., crustose coralline algae), and microalgae films. Macroalgae species were identified according to a reference text (Nelson 2020 ). Only macroalgae species that were observed in at least five quadrats per site were included in the analysis. Each macroalga species with a holdfast located within the quadrat was searched along the entire length of its thallus for any P. canaliculus juveniles attached to it by byssal threads. Once a juvenile mussel was located, its shell length (SL) was measured using callipers. For each macroalga species within a quadrat, the presence or absence of juvenile mussels within three size classes was recorded, i.e., < 10 mm, 10 – <20 mm, and 20 – <30 mm SL. For each macroalga species in a quadrat, once one mussel in a size class was observed, the presence of mussels in that size class was recorded, and any subsequently observed mussels in the same size class were not recorded. The search for mussels attached to each macroalga species continued until all juvenile size classes were found to be present or until all the thalli of each macroalga species within the quadrat had been fully searched. The presence or absence of juvenile mussels in each size class was used in place of mussel counts because of the high densities of juveniles, often in the same size class, in some quadrats (e.g., up to 180 individuals per 0.01 m 2 in the lower intertidal at Waipū Cove), making it impossible to count and measure all of the mussels in situ given tidal time constraints. Destructive sampling was not an option either, as it would have counteracted conservation measures at these locations. A presence/absence measure also ensures that natural variations in the density of juvenile mussels among different areas of the intertidal, where certain species of macroalga are more common than others, were not misinterpreted as higher preference for certain species of macroalga. Location of mussel attachment on macroalgae During the last two of the eight transects sampled at Pākiri Beach during February 2024 and the last four of the 13 transects sampled at Waipū Cove between February and June of 2024, the specific location of juvenile P. canaliculus attachment within macroalgal substrates was recorded as an additional variable to contribute to a comparison of the relative importance of certain morphological features of macroalgae to different developmental stages of juvenile mussels. For each macroalga species within a quadrat, six possible locations were examined for the presence or absence of attached juvenile mussels in each of the three size classes. The six locations within each macroalga were; the holdfast, stipe, lower branch nodes, lower branches, upper branch nodes, and upper branches (Fig. 2 ). A node was defined as the 1 cm 2 area around the point of connection between branches or between a stipe and branch (Alfaro and Jeffs 2002 ; Wu 2018 ). Holdfasts, stipes, and lower and upper branches were defined as 1 cm 2 areas without branch nodes. Lower branches and nodes were considered those that were located below half of the canopy height, while upper branches and nodes were located in the upper half of the canopy height. If the morphology of a macroalga did not have some of these locations, then only the locations it did have were recorded and examined for the presence of attached juvenile mussels. If a macroalga had blades, the blades were categorised as branches, and areas of the thallus with blade proliferation were categorised as nodes. Statistical analyses To test the validity of the MMI, a linear discriminant analysis (LDA; Zhao et al. 2024 ) was applied to juvenile mussel presence data separately for each mussel size class at each site to assess whether scores assigned by the MMI to different morphological characteristics of macroalgae can accurately predict the presence of attached juvenile mussels. An LDA model was chosen because it assumes a linear relationship between the predictors and outcome, and it is less sensitive to small group sizes compared to other discriminant analyses. The predictor variables assessed in the model were the eight morphological features of macroalgae listed in the MMI. The outcome variable used for this analysis was juvenile mussel presence converted from proportional data to binary categorical data (i.e., “high presence” or “low presence”) based on the median for juvenile presence in each size class at each site, allowing relative differences in presence to be assessed without the interference of broader trends in juvenile presence (e.g., lower overall presence of larger juvenile size classes at both sites and higher overall presence of juvenile mussels at the Pākiri Beach site). The LDA was implemented using the lda function in the MASS package in R version 3.5.2 (R Core Team 2023). Prior to running the LDA, the predictors were tested for covariance and multicollinearity. Covariance values among predictors were low to moderate (≤ 0.72), and a test for multicollinearity revealed moderate values on average (0.5) for predictor pairs. However, there were a few predictor pairs that showed moderately high values for multicollinearity (0.77 maximum). Many of the morphological features of macroalgae are inherently correlated. For example, thin branch widths are often found among macroalgae species that also have a high degree of branching and smaller branch spacing (Wu 2018 ), which would result in high scores for each of these predictors. Given that the purpose of the MMI was to assign similar scores to morphological characteristics with a similar capacity to support juvenile mussel recruitment, correlation < 0.8 between predictors was deemed acceptable, and the model was interpreted with the prevailing assumption of potential collinearity. To assess the performance of the LDA and avoid overfitting, a 10-fold cross-validation was employed. The cross-validation partitions the data into 10 subsets, using nine for training and one for testing iteratively, providing a reliable estimate of model accuracy. The model’s performance is summarized using metrics such as accuracy (i.e., the proportion of correctly predicted outcomes) and Cohen’s Kappa (i.e., the level of agreement between observed outcomes and outcomes expected by chance). Afterwards, the relative importance of each morphological feature of macroalgae as a predictor of juvenile mussel presence was quantified using the varImp function in the caret package in R, which evaluates the relative contribution of each predictor based on its influence on the outcome variable. This analysis was useful for determining any morphological features of macroalgae with a low influence on the outcome of juvenile mussel presence that should be excluded from the MMI. To evaluate the relative performance of different macroalgae species of varying morphology as mussel attachment substrates, a 95% confidence interval was calculated for the presence of juvenile P. canaliculus in each size class on each species of macroalga at each site using the binom.test function in R. For each mussel size class, the results of this analysis were used to classify each macroalga species as either a low, moderate, or high-performing attachment substrate for juvenile P. canaliculus , depending on its relation to the interval. To standardise presence against sampling effort, the presence of juvenile mussels on macroalgal substrates is reported as the proportion (%) of each species of macroalga observed during the reef sampling that exhibited juvenile mussel attachments. To evaluate how the location of mussel attachment within substrates varied with mussel size class at each site, the presence/absence of attached juvenile P. canaliculus among macroalgae species was compared using a generalised linear mixed effects model with the binomial family function and logit link, used specifically for data with quasi-binomial distributions, in R. The effect of the factors of site, macroalga species, mussel size class, location of mussel attachment within macroalgae, and their interaction on juvenile mussel presence were assessed with quadrat incorporated as a random effect. Where the interaction of these factors was not significant, the interaction was removed and the model was re-run to assess the significance of the main effects alone using the greater statistical power gained from conserving degrees of freedom. Where the interaction of these factors was significant, post hoc comparisons were performed using Tukey tests to isolate differences in pairs of means through the emmeans function in R. Analysing the data binomially (i.e., presence as 1 and absence as 0) allowed the model to account for differences in sample size among macroalgal substrates. Due to natural variations in the abundance of mussels in each size class, statistical comparisons were not made between size classes but rather, between substrates within size classes. Results Design of the Macroalgal Morphology Index Fifty-nine research publications covering 12 species of mussel from 33 locations around the world were identified from the literature search that tested or discussed impacts of the specific structural properties of natural or artificial substrates on mussel recruitment, settlement, attachment, growth, or survival (see Table 1 for these publications and their relevant findings). Forty-five publications tested the effects of the physical structure of the substrate on mussels, while 14 publications implicated substrate morphology as the cause of their findings. From these publications, eight morphological features of macroalgal substrates were identified as potential determinants of juvenile mussel attachment for the MMI; canopy cover, canopy height, holdfast complexity, clustering frequency, planes of branching, branch spacing, degree of branching, and branch width (Table 2; Fig. 1 ). Substrate surface texture was also identified as a structural feature that can impact mussel attachment in eight of the 59 publications (Bourget et al. 1994 ; Gribben et al. 2011 ; Carl et al. 2012a ; Wu 2018 ; Loucks 2023 ; Wu and Jeffs 2025 ; Wu et al. 2025 ; Lanham et al. 2025 ). However, this feature was not identified in Table 1 or included in the MMI because characterising the surface texture of macroalgae is difficult to do in situ and requires expensive techniques, such as scanning electron microscopy, that would be less accessible for restoration practitioners, for whom the MMI is intended. Nonetheless, these studies were included in Table 1 because there were instances where the descriptions of substrate surface texture were applicable to other morphological features, such as the size of millimetre-scale surface indents on artificial substrates being comparable to the space between branches on macroalgae (Bourget et al. 1994 ; Gribben et al. 2011 ; Carl et al. 2012a ; Wu and Jeffs 2025 ). Similarly, greater surface roughness was comparable to higher macroalgal canopy cover in terms of its effect on flow velocities and particle deposition (Lanham et al. 2025 ). For each morphological feature in the MMI, qualitative criteria for each of the three scores, representing a substrate’s ability to support juvenile mussel attachment and survival, were able to be characterised using the descriptions of substrates in the literature (Table 2). Likewise, mean values of the morphological measurements collected from each species of macroalga in the field allowed for the generation of quantitative ranges for each score (Table S1 ). MMI criteria were used to objectively score the morphological features of each macroalga species, and these scores were summed to produce a total index score for each species intended to represent the potential that each substrate has to support the recruitment of juvenile mussels (Table 3). Since eight morphological features, each subdivided into three scored variations, were identified for the MMI, each macroalga species could have a minimum total index score of eight and a maximum total index score of 24. Mussel-macroalgae associations At Waipū Cove, a total of 17 macroalgae species of varying phyla and morphology were observed during reef sampling (Fig. 3 ). Only one of these species of macroalgae was green (i.e., phylum Chlorophyta), five were brown, and 11 were red, two of which were coralline red macroalgae (i.e., order Corallinales). The presence of juvenile P. canaliculus attached to these macroalgal substrates varied among the three size classes of juvenile mussels. Juveniles < 10 mm SL were present on all macroalgae species at Waipū Cove except for the red macroalga Hymenena variolosa , although juveniles were present on this species at Pākiri Beach. In contrast, juveniles 10 – <20 mm SL and 20 – <30 mm SL were present on only eight species of macroalgae, of which Corallina ferreyrae (prev. Corallina officinalis ; coralline red), Jania sphaeroramosa (coralline red), Psilophycus alveatus (prev. Gigartina alveata ; red), Pterocladiella capillacea (red), Scytothamnus australis (brown), and Hormosira banksii (brown) exhibited attachments by juveniles in both of these larger size classes. Juveniles were only present on the green macroalga Ulva australis up to 20 mm SL. At Waipū Cove, the macroalgae species that demonstrated the highest presence of attached juvenile mussels < 10 mm SL were P. alveatus (87% presence), S. australis (59%), U. australis (53%), and J. sphaeroramosa (47%; Fig. 3 ). The high-performing macroalgae species for juvenile mussels 10 – <20 mm SL were P. alveatus (45% presence), U. australis (20%), S. australis (18%), and H. banksii (18%). For the largest juvenile mussels 20 – <30 mm SL, the high-performing substrates were P. alveatus (30% presence) and S. australis (24%). Additionally, C. ferreyrae and H. banksii (39% presence each) supported a moderate presence of attached juveniles < 10 mm SL, while Ecklonia radiata (brown), C. ferreyrae and P. capillacea (10–14%) were moderate-performing substrates for juveniles 10 – <20 mm SL and C. ferreyrae and P. capillacea (10–12%) performed similarly for juveniles 20 – <30 mm SL. At Pākiri Beach, a total of 16 macroalgae species of varying phyla and morphology were observed during reef sampling (Fig. 3 ). Two of these species of macroalgae were green, four were brown, and 10 were red, two of which were coralline red macroalgae. Similar to Waipū Cove, the presence of attached juvenile mussels varied among mussel size classes. Juveniles < 10 mm SL were present on all macroalgae species at Pākiri Beach. In contrast, juveniles 10 – <20 mm SL were present on only 13 species of macroalgae, and juveniles 20 – <30 mm SL were present on only 11 species of macroalgae, of which C. ferreyrae (coralline red), J. sphaeroramosa (coralline red), Caulacanthus ustulatus (red), Gigartina macrocarpa (red), H. variolosa (red), Carpophyllum maschalocarpum (brown), Sargassum sinclairii (brown), and E. radiata (brown) exhibited attachments by juveniles in both of these larger size classes. Similar to Waipū Cove, juveniles were only present on the two green species of macroalgae, U. australis and Codium fragile subsp. novae-zelandiae , up to 20 mm SL. At Pākiri Beach, there were several more species of red macroalgae that qualified as high-performing attachment substrates for juvenile mussels < 10 mm SL than for juveniles above this size. The macroalgae species with the highest presence of attached juveniles < 10 mm SL were C. maschalocarpum (100% presence), Chondracanthus chapmanii (100%; red), J. sphaeroramosa (96%), C. ferreyrae (92%), Laurencia thyrsifera (92%; red), S. sinclairii (91%), and C. ustulatus (89%; Fig. 3 ). For juvenile mussels 10 – <20 mm SL, the high-performing substrates were C. maschalocarpum (78% presence), C. ferreyrae (48%), and S. sinclairii (38%), while C. ferreyrae (39%) was the only high-performing substrate for juveniles 20 – <30 mm SL. Additionally, Chondria macrocarpa (red; 80% presence) supported a relatively moderate presence of attached juveniles < 10 mm SL, while S. sinclairii (19%) was a moderate-performing substrate for larger juveniles 20 – <30 mm SL and juveniles 10 – <20 mm SL had no moderate-performing substrates. Overall, the high-performing macroalgal substrates for juvenile mussels in all three size classes at both sites shared at least three of the following high-scoring morphological characteristics described in the MMI: a widespread canopy cover, a more complex holdfast that attached to the substrate at multiple points, a higher clustering frequency (i.e., a higher tendency to grow in close proximity to other conspecifics and colonise a larger area of substrate), multiple planes of branching and/or folding, and thinner branch widths at the nodes (Table 3). Shorter branch spacing was also common among the high-performing substrates at Waipū Cove, and a higher degree of branching was common among the high-performing substrates at Pākiri Beach. At both sites, taller canopy heights were not a characteristic of macroalgae that was consistently associated with a higher presence of attached juvenile mussels in all three size classes. Additionally, more of the high-performing macroalgal substrates for smaller juveniles < 10 mm SL received high scores for features that describe aspects of branching morphology than the high-performing substrates for larger juveniles ≥ 10 mm SL at both sites. While there were 11 species of macroalgae that were observed at both Waipū Cove and Pākiri Beach during reef sampling, the only species of macroalga with consistently moderate or high relative presence of attached juvenile mussels for all size classes at both sites was C. ferreyrae . Additionally, J. sphaeroramosa supported high relative presence of attached juveniles in the smallest size class at both sites but not for juveniles ≥ 10 mm SL. Both of these macroalgal substrates are turf-forming, coralline red species that grow throughout rockpools and emergent rock in the intertidal. They both have high-scoring morphological characteristics, such as a widespread canopy cover, a more complex holdfast, a high clustering frequency, very narrow branch widths, a high degree of branching, multiple planes of branching, and shorter branch spacing (Table 3). Consequently, these two species have some of the highest total index scores of 20 for C. ferreyrae and 22 for J. sphaeroramosa out of 24 (Fig. 3 ). Despite a different assemblage of macroalgae species qualifying as high-performing attachment substrates for juvenile mussels at each site, all high-performing macroalgae species scored moderate to high total index scores according to the MMI. At Pākiri Beach, high-performing macroalgae species for all juvenile mussel size classes had a minimum total index score of 14 and a maximum of 22 out of 24, while the minimum score at Waipū Cove was 16 and the maximum score was also 22 (Fig. 3 ). While for many species of macroalgae, moderate to high total index scores were associated with relatively higher presence of attached juvenile mussels, there were a number of other macroalgal substrates with high total index scores that were considered low-performing. For example, Gelidium caulacantheum is a turf-forming, fleshy, red macroalga at Waipū Cove that has a total index score of 21, but it exhibited relatively low presence of attached juvenile mussels (≤ 20% presence; Fig. 3 ). Similarly, U. australis was a high-performing substrate for juvenile mussels < 20 mm SL at Waipū Cove (20–53% presence; Fig. 3 ), but it has a mix of low and high scoring morphological characteristics, resulting in moderate a total index score of 16 (Table 3). Location of mussel attachment on macroalgae During the portion of reef sampling that assessed the location of juvenile mussel attachment on macroalgal substrates, ten species of macroalgae were sampled at Waipū Cove and thirteen species of macroalgae were sampled at Pākiri Beach (Fig. 4 ). Only three of these macroalgae species were observed at both sites during this portion of sampling, i.e., C. ferreyrae , J. sphaeroramosa , and Dictyota kunthii (brown). However, the generalised linear mixed effects model revealed that the species of macroalgae had no significant interactive effects on the presence of attached juvenile mussels among different locations within the substrate for any mussel size class at either site. Therefore, this effect was removed and the model was re-run to interpret the main effects only. Site, mussel size class, and location within the substrate all had significant interactive effects on attached juvenile mussel presence ( P < 0.033). Post hoc comparisons of these factors revealed that while the locations of attachment on macroalgal substrates were fairly consistent for juvenile mussels in the two larger size classes, attachment locations differed between the two study sites for juveniles in the smallest size class. Overall, for larger juveniles ≥ 10 mm SL at both sites, the presence of attached mussels was highest on the holdfasts of macroalgal substrates. For the largest juveniles 20 – <30 mm SL, presence was higher on holdfasts than on any other attachment location within the substrates by at least 11% at Pākiri Beach ( P < 0.038) and 21% at Waipū Cove ( P < 0.0035). The same was true for mid-sized juveniles 10 – <20 mm SL at Waipū Cove by at least 17% ( P < 0.0072), but at Pākiri Beach, the presence of mid-sized juveniles was higher on both the holdfasts and lower branch nodes of macroalgae than in any other attachment location by 20% and 10%, respectively ( P < 0.042). Overall, for the smallest juveniles < 10 mm SL at both sites, the presence of attached mussels was highest on both the holdfasts and branch nodes of macroalgae, while attachments to the branches and stipes of macroalgae were less common. For juveniles < 10 mm SL at Pākiri Beach, the presence of attached mussels was higher on the holdfasts and lower branch nodes of macroalgae than on any other attachment location by at least 59% and 54%, respectively ( P < 0.0001). However, for juveniles < 10 mm SL at Waipū Cove, the presence of attached mussels was higher on the upper branch nodes of macroalgae (32% presence), followed by the holdfasts (23%) and lower branch nodes (18%), than on any other attachment location within the substrates (≤ 12%). The presence of attached juvenile mussels in this size class was significantly higher on the upper branch nodes compared to the stipe and upper and lower branches by at least 20% ( P < 0.0056), and presence was significantly higher on the holdfasts compared to the stipe and upper branches by at least 16% ( P < 0.024). The post hoc comparisons of mussel attachment locations between study sites revealed that at Pākiri Beach, larger juveniles ≥ 10 mm SL were not observed attaching to the lower or upper branches of macroalgae, while larger juveniles at Waipū Cove were not observed attaching to the stipes. For smaller juveniles < 10 mm SL, attachments to the upper portion of macroalgal substrates were more common at Waipū Cove than at Pākiri Beach, while attachments to the lower portion of macroalgae were conversely more common at Pākiri Beach. At Waipū Cove, there was a higher presence of juveniles < 10 mm SL attached to the upper branches and upper branch nodes of macroalgal substrates than at Pākiri Beach by 6% and 30%, respectively ( P < 0.039). Meanwhile, at Pākiri Beach, there was 45% higher presence of juveniles in this size class attached to the holdfasts and lower branch nodes of macroalgae than at Waipū Cove ( P < 0.0001). These site-specific differences in the attachment locations of juvenile mussels on macroalgae were maintained across the same species of macroalgae at both sites. Evaluation of the Macroalgal Morphology Index Assessment of the relative importance of each morphological feature identified in the MMI, as predictor variables of the observed presence of attached juvenile mussels among macroalgae, revealed that the importance of each feature differed among sites and mussel size classes (Table 4). High variability in the relative contribution of each predictor to the observed outcome meant that no one morphological feature was of consistently low enough importance for all mussel size classes and for both sites for it to be eliminated from the MMI. Therefore, all eight morphological features were retained. At Waipū Cove, the morphological features of macroalgae that qualified as strong predictors of attached juvenile presence (i.e., ≥ 0.65 relative importance value on a scale of 0.5 to 1.0) that were shared among all three mussel size classes were holdfast complexity (0.65–0.81 relative importance) and canopy cover (0.67–0.72; Table 4). At Pākiri Beach, canopy cover (0.70–0.75 relative importance) was a strong predictor of the presence of attached juveniles for all three size classes, while holdfast complexity qualified as a strong predictor for all size classes except for juveniles 10 – <20 mm SL. Overall, at both sites, there were a greater number of morphological features in the model that qualified as strong predictors of attached mussel presence for smaller juveniles < 10 mm SL than for larger juveniles ≥ 10 mm SL. Collectively, the MMI scores for each morphological feature were 75% accurate effective at predicting the presence of juvenile mussels < 10 mm SL among macroalgal substrates at both sites, and agreement between observed and predicted juvenile presence for this size class was moderately strong, as indicated by a Cohen’s Kappa of 0.43 at Pākiri Beach and 0.50 at Waipū Cove (Table 4). At both sites, three out of the five strong predictors of the presence of attached juveniles < 10 mm SL were features of macroalgae that describe aspects of branching morphology. At Waipū Cove, these features were branch spacing (0.73 relative importance), planes of branching (0.69), and branch width (0.65), while at Pākiri Beach, these features were degree of branching (0.73), branch width (0.73), and planes of branching (0.69). In contrast, for larger juveniles ≥ 10 mm SL, branching features were generally poor predictors of juvenile presence. At Waipū Cove, there was one branching feature that qualified as a strong predictor of attached mussel presence for juveniles ≥ 10 mm SL, but at Pākiri Beach, all branching features were poor predictors for the presence of larger juveniles. At both sites, canopy cover and holdfast complexity were the only morphological features that consistently acted as strong determinants of attached mussel presence for juveniles ≥ 10 mm SL. As a result, the MMI scores for each morphological feature were collectively able to predict the presence of attached juveniles ≥ 10 mm SL among macroalgal substrates with only 50–60% accuracy (Kappa: 0.11–0.33) at Waipū Cove and 40–45% accuracy (Kappa: -0.13–0) at Pākiri Beach (Table 4). Discussion By creating an index that scores the morphological features of macroalgae for their potential to support recruitment of juvenile mussels, this study has developed a tool for mussel reef restoration practitioners to rank the suitability of macroalgal substrates in an objective and repeatable manner using clearly defined physical descriptions. Furthermore, by testing the accuracy of this index against the presence of P. canaliculus juveniles attached to macroalgae on two remnant mussel reefs in New Zealand, this study demonstrates that the morphology of a macroalgal substrate can be used to accurately predict the substrate’s potential to support juvenile mussel attachments. Eight morphological features of macroalgae were identified from studies on 12 different species of mussel and used to create the Macroalgal Morphology Index (i.e., canopy cover, canopy height, holdfast complexity, clustering frequency, planes of branching, branch spacing, degree of branching, and branch width). Substrates with morphological characteristics that equate to widespread canopy cover, higher holdfast complexity, multiple planes of branching or folding, thinner branch widths at the nodes, and a higher degree of branching on macroalgae are commonly reported to support a higher degree of juvenile attachment in many species of mussel (Bulleri et al. 2006 ; Brenner and Buck 2010 ; Carl et al. 2012b ; Liu et al. 2017 ; Wu and Jeffs 2025 ). In this study, macroalgae species that exhibited a number of these high-scoring physical characteristics supported some of the highest presence of juvenile P . canaliculus . Collectively, the MMI’s scoring criteria was most effective at predicting the likelihood of P. canaliculus attachments to macroalgae for juveniles smaller than 10 mm SL and less effective for juveniles 10 mm SL and larger that relied less on branching morphology for attachment. These findings indicate that the MMI can be used as a tool to assist mussel reef restoration practitioners in both the evaluation of existing macroalgal habitat at future restoration sites and the selection of suitable macroalgal substrates to be transplanted into current restoration sites for the purpose of facilitating juvenile recruitment among restored mussel populations in New Zealand and likely many other locations across the world. Mussel-macroalgae associations At both sites, P. canaliculus juveniles smaller than 10 mm SL attached to almost all species of macroalgae present on the reefs, while larger juveniles between 10 and 20 mm SL and 20 and 30 mm SL were only present on about half of the macroalgae species at Waipū Cove and roughly three quarters of the species at Pākiri Beach. These findings suggest that, regardless of the species or morphology, macroalgae are critical attachment substrates for juvenile P. canaliculus during early development, and the type of macroalgae only begins to exclude the presence of juveniles during later development. Similarly, attached juvenile presence appeared to be unrelated to macroalgal phyla until juveniles reached 20 mm SL, after which point they were no longer present on green macroalgae but still demonstrated attachments to numerous species of red and brown macroalgae. Despite a greater availability of red macroalgae species at both sites, only juveniles smaller than 10 mm SL at Pākiri Beach demonstrated a higher presence on more species of red macroalgae than brown macroalgae. While some studies report higher settlement of juvenile P. canaliculus on a greater number of red macroalgae species than brown macroalgae species at smaller juvenile sizes (< 0.5 mm SL; Alfaro and Jeffs 2002 ; Alfaro et al. 2004 ), high settlement onto numerous species of brown and red macroalgae has been reported for juvenile P. canaliculus as small as ~ 250 µm and as large as ~ 7 mm SL (Buchanan and Babcock 1997 ; Loucks 2023 ; Toone et al. 2023c ). It is likely that differences in the morphology of these macroalgae species are responsible for the variation in juvenile mussel settlement preferences for red or brown macroalgae between studies. Regardless, the results of this study contribute valuable insights into the diversity of macroalgal phyla that P. canaliculus juveniles continue to associate with beyond 7 mm SL. There was a moderate to high presence of attached juvenile mussels on macroalgae species previously reported to support varying degrees of juvenile P. canaliculus attachment (i.e., red macroalgae: P. alveatus , Jania spp., L. thyrsifera , C. ferreyrae , P. capillacea ; brown macroalgae: C. maschalocarpum , S. australis , H. banksii ) but also on macroalgae not previously reported as attachment substrates for P. canaliculus (red macroalgae: C. chapmanii , C. ustulatus , C. macrocarpa ; brown macroalgae: S. sinclairii , E. radiata ; green macroalgae: U. australis ), expanding the current list of macroalgal substrates with strong potential to facilitate the recruitment of juvenile P. canaliculus at restoration sites. In particular, the common, coralline red macroalgae C. ferreyrae and J. sphaeroramosa were the only macroalgae species to support a moderate to high presence of juvenile mussels at both sites, in which J. sphaeroramosa qualified as a high-performing attachment substrate for juveniles in the smallest size class and C. ferreyrae qualified as a moderate- or high-performing substrate for juveniles in all size classes. These findings mirror the high settlement densities of larvae and juvenile mussels supported by species of Jania and C. ferreyrae , especially, throughout numerous laboratory and field studies (Hunt and Scheibling 1995 ; Buchanan and Babcock 1997 ; Alfaro et al. 2004 ; McQuaid and Lindsay 2005 ). Inconsistencies in the performances of the nine other species of macroalgae that were present at both sites are likely the result of differing hydrodynamic conditions between the two sites interacting with macroalgal morphology to positively or negatively influence juvenile mussel attachment. Therefore, these findings indicate that even under variable hydrodynamic conditions, C. ferreyrae and J. sphaeroramosa together have the highest potential to support the attachment of juvenile P. canaliculus throughout their development. While C. ferreyrae and J. sphaeroramosa can grow in the upper intertidal down to a depth of at least 20 m, other high-performing macroalgae species, such as P. alveatus , C. chapmanii, C. ustulatus , S. australis , U. australis , H. banksii , C. maschalocarpum , and S. sinclairii , are not typically found below 6 m depth (Nelson 2020 ), suggesting that many of these species are only suitable for intertidal and shallow subtidal mussel reef restoration sites. Furthermore, thin and highly branched red macroalgae, which describe many of the high-performing species in this study, can be vulnerable to the effects of abrasive, resuspended sediment that characterise areas of the contemporary Hauraki Gulf (Lao 2016 ). Therefore, for deeper or more degraded mussel reef restoration sites, selecting hardier, low-light tolerant species of macroalgae or designing artificial substrates, potentially comprised of natural materials, that exhibit similar morphological characteristics to the high-performing macroalgae species in this study would likely be a more practical and effective choice for future substrate provisioning. Location of mussel attachment on macroalgae Understanding which locations within a macroalgal substrate juvenile mussels attach to throughout different stages of ontogeny is critical for understanding why the impact of different morphological features of macroalgae may change with juvenile size. In this study, for P. canaliculus juveniles smaller than 10 mm SL, attachments to the branches and branch nodes of macroalgae were far more common than for juveniles 10 mm SL and larger, suggesting that attachment by P. canaliculus juveniles is more likely to be influenced by the branching morphology of macroalgae earlier in juvenile development. Additionally, juveniles were more often attached to the branch nodes of macroalgae than to the inter-node areas of the branches regardless of juvenile size, mirroring previously reported higher and more clustered settlement of juvenile P. canaliculus on the nodes of branching macroalgae and plastic mimics (Kelly 2001 ; Alfaro and Jeffs 2002 ). While juvenile mussels are rarely documented attaching to the holdfasts of macroalgae in previous literature, this was the most common location of attachment in this study for juveniles 10 mm SL and larger, and for juveniles smaller than 10 mm SL, it was as common of an attachment location as the branch nodes of macroalgae. It is likely that mussel attachments to holdfasts are virtually unreported for P. canaliculus because the majority of macroalgal and artificial attachment substrates analysed for this species have been drifting or beachcast samples or have been experimentally suspended in the water column (Buchanan and Babcock 1997 ; Alfaro and Jeffs 2002 ; Toone et al. 2023c ). The results of this study highlight the importance of the holdfast morphology of macroalgae for attachment of all sizes of juvenile P. canaliculus but especially for juveniles in later development. For P. canaliculus juveniles smaller than 10 mm SL, attachments to the upper branches and nodes of macroalgae were more common at Waipū Cove than at Pākiri Beach, where smaller juveniles were more often attached to the lower portion of macroalgae. Even when the same species of macroalgae was present at both sites (e.g., C. ferreyrae and J. sphaeroramosa ), these site-specific differences in the location of juvenile mussel attachment persisted, implicating the interaction of the substrate’s morphology with the site’s environmental conditions as the cause of these trends. It is likely that higher wave exposure at Pākiri Beach makes attachment to the upper portions of macroalgae less favourable for juvenile mussels than under the more sheltered conditions at Waipū Cove. As juvenile mussels grow, the risk of dislodgement increases as the impact of hydrodynamic forces like frictional drag and shear stress act more strongly on their larger surface areas (Hunt and Scheibling 2001 ; Donker et al. 2013 ). An increasing risk of dislodgement as juveniles grow would explain why larger juveniles were primarily attached to the holdfasts, where their position close to the substrate places them in an area of reduced water velocity and greater stability beneath the macroalgal canopy (Eckman and Duggins 1991 ; McCook and Chapman 1991 ; Westerbom et al. 2008 ). It would also explain why macroalgal substrates with high-scoring branching morphologies supported a high presence of juveniles in all size classes at Waipū Cove but only supported a high presence of juveniles smaller than 10 mm SL at Pākiri Beach, as attachment to branches likely places larger juveniles at greater risk. This finding highlights the need for mussel reef restoration practitioners that are assessing the suitability of macroalgal substrates to pay close attention to the features in the MMI that score the branching and holdfast morphologies of macroalgae and consider whether environmental conditions at the proposed restoration site are likely to support the continued attachment of juvenile mussels to these locations within the macroalgae. Macroalgal morphology Juvenile P. canaliculus at both mussel reefs were attached to macroalgal substrates with a variety of morphologies, regardless of juvenile size. However, there were a number of key morphological characteristics shared between many of the macroalgal substrates that supported a high presence of attached juvenile mussels in all three size classes. These high-performing macroalgal substrates exhibited not all but a combination of many of the following characteristics: a widespread canopy cover, a higher holdfast complexity, a higher clustering frequency, multiple planes of branching or folding, thinner branch widths at the nodes, shorter branch spacing, and a higher degree of branching. The only morphological feature of macroalgae that consistently qualified as a strong predictor of the presence of attached juvenile mussels in all size classes at both sites was canopy cover, and the same was true of holdfast complexity for all but the 10 – <20 mm SL juvenile size class at Pākiri Beach. This finding suggests that out of all the morphological features described in the MMI, canopy cover and holdfast complexity are likely the most important drivers of the continued attachment of juvenile mussels on macroalgae throughout juvenile ontogeny. Macroalgae species that received high scores from the MMI for having a widespread canopy cover included both large fucoids (e.g., C. maschalocarpum , S. sargassum , S. australis , H. banksii , and E. radiata ) and short, turf-forming rhodophytes (e.g., C. ferreyrae , J. sphaeroramosa , C. chapmanii , and C. ustulatus ) because the wide or continuous canopies exhibited by these species can create similar physical effects on their immediate surroundings that have the potential to benefit juvenile mussels. For example, macroalgal canopies that cover a larger area of substrate can promote the settlement of mussel larvae by reducing water flow as it moves across the canopy, resulting in enhanced particle deposition (Eckman and Duggins 1991 ; Bégin et al. 2004 ), and by providing shaded surfaces preferred by settling mussel larvae (Marsden and Lansky 1999; Kobak 2001 ; Holthuis et al. 2015 ). Greater canopy cover can also enhance the survival of juvenile mussels by providing shade and moisture that protect against desiccation and mediate temperature stress in the intertidal (McCook and Chapman 1991 ; Bulleri et al. 2006 ; de Nesnera 2016 ), and reduced water velocities across areas of greater continuous macroalgal cover have also been hypothesized to reduce the dislodgement of intertidal mussels (McCook and Chapman 1991 ). Therefore, macroalgae with widespread canopy covers have the potential to enhance both the initial attachment and the longer-term retention of juvenile mussels. Ameliorated temperature extremes and reduced hydrodynamic forces created by broader macroalgal canopies can also be associated with a higher density of predators in their understory, sometimes leading to higher mussel mortality (Eckman and Duggins 1991 ; Menge 1978 ). However, the three-dimensional complexity of macroalgal holdfasts can further mediate juvenile mussel predation by regulating predator access (Moreno 1995 ). On artificial surfaces with higher three-dimensional complexity, such as corrugated panels and shells embedded in concrete, Mytilus edulis juveniles demonstrated greater survival from predation through the formation of protective aggregations and a higher number of byssal attachments to the substrate that increased their stability (Frandsen and Dolmer 2002 ; Christensen et al. 2015 ). In this study, macroalgae that exhibited more complex holdfasts secured to the substrate at multiple points, such as P. alveatus , H. banksii , and the majority of the turf-forming red macroalgae, were often observed supporting densely-packed aggregations of juvenile mussels interwoven amongst the holdfasts in a manner that appeared to limit the amount of surface area on their shells that was exposed to predators and hydrodynamic drag. Refuge-seeking behaviour from predators and hydrodynamic forces was implicated in a previous study as the reason that Mytilus galloprovincialis larvae initially settled deep inside of structurally complex culture ropes, only migrating to the rope exteriors as larger juveniles (Carl et al. 2012b ). These findings suggest that like canopy cover, holdfast complexity can influence both the settlement behaviour of mussel larvae and the post-settlement survival of juvenile mussels, especially considering that in this study, holdfasts were one of the most common attachment locations for P. canaliculus juveniles smaller than 10 mm SL and the most common area of attachment for juveniles 10 mm SL and larger. Given the commonality in the substrate attachment preferences reported for other species of mussel, these features of macroalgal morphology are likely to have similar effects on other mussel species in other locations. Therefore, mussel reef restoration practitioners evaluating new restoration sites or provisioning existing sites with macroalgal substrates in an attempt to facilitate juvenile recruitment should look for macroalgae with three-dimensionally complex holdfasts and widespread canopy covers to improve the chances of supporting the establishment of juvenile mussels within the reef. Unlike holdfast complexity and canopy cover, the majority of macroalgal features that describe aspects of branching morphology qualified as strong predictors of the presence of attached juvenile P. canaliculus in the smallest size class but were poor predictors for the two larger juvenile size classes at both sites. This finding, along with the observation that juveniles smaller than 10 mm SL were attached to the branches and branch nodes of macroalgae much more often than juveniles 10 mm SL and larger, suggests that the branching features of macroalgal morphology are only drivers of the attachment of juvenile mussels in their earlier stages of development. For juveniles smaller than 10 mm SL, branch width and planes of branching qualified as strong predictors of presence among macroalgae at both sites, while degree of branching was a strong predictor only at Waipū Cove and branch spacing was only a strong predictor at Pākiri Beach. In many studies, juvenile mussels smaller than 1 mm SL settled in higher densities on macroalgae characterised primarily by thinner branch widths but also by a higher degree of branching (Buchanan and Babcock 1997 ; Dobretsov and Wahl 2001 ; Alfaro et al. 2004 ; Yang et al. 2007 ). However, in many of these studies, juveniles began to attach in higher densities to macroalgae with characteristically broader and less numerous branches starting at just 1 or 2 mm SL (Hunt and Scheibling 1995 ; Kelly 2001 ; Alfaro and Jeffs 2002 ; Alfaro et al. 2004 ), and some studies showed higher settlement onto these types of macroalgae in juveniles below this size (Loucks 2023 ; Toone et al. 2023c ). Therefore, while thinner, highly branched macroalgae species, such as J. sphaeroramosa , C. ferreyrae , Caulacanthus ustulatus , Laurencia thyrsifera , and Chondria macrocarpa , typified many of the moderate- to high-performing attachment substrates for juvenile P. canaliculus smaller than 10 mm SL in this study, it is not unusual for wider, sparsely branched macroalgae, such as U. australis, C. maschalocarpum , and S. sinclairii , to have also supported a higher presence of smaller juveniles. It is likely a limitation of this study that the 10 mm size classes used were potentially too large to capture any variation in preference for different branch widths or degrees of branching that juvenile mussels exhibit during the earliest stages of development. Higher presence of smaller P. canaliculus juveniles on wider, sparsely branched U. australis and S. sinclairii may have also been the result of higher scores for branch spacing and/or planes of branching due to tightly overlapping branches that twisted or folded onto multiple planes. There is less known about the influence of these two other branching features on juvenile mussel attachment to macroalgae. Studies on mussel settlement often sort macroalgae species into loosely defined functional groups that focus on a single branching characteristic, usually the branch width or degree of branching, when for many species of macroalgae, thinner branch widths, a higher degree of branching, multiple planes of branching, and shorter branch spacing are all frequently associated with each other. These commonly overlapping morphological traits make it difficult to determine which of these characteristics is primarily responsible for observed patterns in mussel attachment among macroalgae (Wu 2018 ). Moreover, studies utilising plastic analogues of macroalgae to isolate and test the influence of each branching feature on the settlement of juvenile mussels smaller than 1 mm SL revealed mixed results for all features except the number of branching planes, which remains to be tested in this manner. In these studies, settlement densities were unaffected by changing the distance between branches (Kelly 2001 ; Wu 2018 ), either remain unchanged or declined with increasing degree of branching (Harvey et al. 1995 ; Kelly 2001 ; Wu 2018 ), and generally increased as branch width decreased–except in one study that reported no change (Harvey et al. 1995 ; Kelly 2001 ; Wu 2018 ; Wu and Jeffs 2025 ; Wu et al. 2025 ). Despite these mixed results, artificial substrates designed to collect mussel larvae and grow them for aquaculture that employ the use of thinner filaments, a higher degree of three-dimensional complexity across multiple planes, and interwoven or looped filaments that provide small refuge spaces within the substrate each support higher densities of early juvenile mussel settlement (Walter and Liebezeit 2003 ; Filgueira et al. 2007 ; Brenner and Buck 2010 ; Protopopescu and Beal 2015 ). These structurally complex morphologies allow juvenile mussels to increase their stability on the substrate through a greater surface area for the attachment of more byssal threads and the ability to surround themselves with the substrate, which can help protect against dislodgement and predation (Brenner and Buck 2010 ; Filgueira et al. 2007 ). In this study, juvenile P. canaliculus attached to thinner, more highly branched macroalgal substrates were observed forming densely packed mussel-substrate conglomerates, in which juveniles were tightly attached to and surrounded by both neighbouring juveniles and the branches of the macroalgae. Similar observations were made for Mytilus edulis juveniles attached to three-dimensionally complex, filamentous artificial substrates, where the formation of these mussel-substrate conglomerates was associated with decreased dislodgement of juveniles from these substrates in areas of greater hydrodynamic exposure (Brenner and Buck 2010 ). However, it is possible that at sites with particularly high hydrodynamic exposure, such as Pākiri Beach, these more structurally complex branching characteristics may not be beneficial enough for juveniles larger than 10 mm SL to overcome the shear stress generated across their larger surface areas by wave action (Donker et al. 2013 ). This would help to explain why fewer larger juveniles were found in the branches of macroalgae at Pākiri Beach compared to Waipū Cove as well as why no branching features qualified as strong predictors for the presence of larger juveniles at Pākiri Beach. While all four branching features of macroalgae are likely important determinants of the attachment of juvenile mussels during early development, the influence of branching features on larger juvenile attachments to macroalgae appears to be reduced at sites with higher hydrodynamic exposure, although further testing would be needed to confirm if wave action is the primary cause of the observed patterns. Therefore, these findings suggest that restoration practitioners seeking to assess the branching morphology of macroalgal substrates for their potential to support not only the initial settlement, but also the continued attachment, of juvenile mussels should place greater value on high-scoring branching characteristics if the substrates are found at, or intended for, restoration sites that are characterised by lower levels of hydrodynamic stress. However, additional experimentation is needed to determine the level at which hydrodynamic stress begins to have a significant negative impact on juvenile attachments to the branches of macroalgae. Unlike canopy cover, holdfast complexity, and the features of branching morphology, canopy height varied greatly among the high-performing macroalgal substrates for each size class of juvenile P. canaliculus . While canopy height qualified as the strongest predictor of the presence of attached juvenile mussels between 10 and 20 mm SL on macroalgae at Waipū Cove, it was a poor predictor of juvenile presence for the other two size classes at this site and all juvenile size classes at Pākiri Beach. This inconsistency is likely due to a trade-off between the costs and benefits that taller macroalgal canopy heights offer juvenile mussels. Some studies suggest that taller canopies can enhance the survival of juvenile mussels attached to their branches by lifting them away from scouring sediments, predators, and competition with adult mussels (Seed and Suchanek 1992 ; Westerbom et al. 2008 ) and up into the area above the seabed, where higher water flow and food availability can support higher growth rates (Fréchette and MMI. However, macroalgae with taller canopies are also more susceptible to ‘whiplash effects’ under higher wave action, in which the movement of the thallus can dislodge both epibiotic organisms attached to the canopies and understory organisms that may be attached to the holdfasts (Leonard 1999 ; O’Connor et al. 2006 ; Erlandsson et al. 2008 ). Juvenile mussels may be able to tolerate the more dynamic movement of taller macroalgae if they also exhibit morphological characteristics that allow for multiple, firm attachment points, such as species with more three-dimensionally complex branching or holdfast morphologies, as greater structural complexity has been shown to reduce the dislodgement of juvenile mussels attached to artificial substrates (Brenner and Buck 2010 ). Variable morphologies like these would help to explain why large but more highly branched species, such as S. sinclairii and C. maschalocarpum , or taller species with more complex holdfasts, such as P. alveatus and H. banksii , supported a higher presence of juvenile mussels than tall but less structurally complex species, such as E. radiata or Dictyota kunthii . Additionally, greater shelter from wave action among the more physically complex rock formations at Waipū Cove may have reduced the dislodging effects of whiplash on mid-sized juveniles attached to taller macroalgae, allowing the advantages of taller canopy heights to drive greater presence of juveniles between 10 and 20 mm SL in more sheltered areas of the reef. Confirmation that the interaction of macroalgal morphology and wave action is responsible for these observed patterns in the distribution of juvenile mussels will need to come from further research. These findings once again highlight the importance of considering the hydrodynamic conditions at restoration sites when selecting suitable macroalgal substrates for the facilitation of juvenile mussel recruitment. However, while canopy height may have a stronger influence on the presence of juvenile mussels among macroalgae at some sites, supporting its inclusion in the MMI, the overall inconsistency of its impact suggests that this feature one of the least important drivers of juvenile mussel attachment. Clustering frequency was another morphological feature that only once qualified as a strong predictor of P. canaliculus presence among macroalgae in the case of juveniles smaller than 10 mm SL at Pākiri Beach. This outcome can most likely be attributed to the large number of both high- and low-performing macroalgal substrates for each juvenile size class that received high scores for their tendency to frequently cluster in patches or continuous turfs comprised of the same species, especially at Waipū Cove. Macroalgae with high clustering frequencies are awarded high scores in the MMI because a greater number macroalgal individuals across a smaller area is more likely to produce a stronger localised chemical cue, and previous studies show that chemical cues from algal extracts and macroalgae placed in close proximity can promote the settlement and retention of juvenile mussels (Dobretsov 1999 ; Alfaro et al. 2006 ; Skelton and Jeffs 2021 ). While the MMI does not score the biochemical effects of macroalgae on juvenile mussel attachment, scoring the clustering frequency has the benefit of incorporating some of the influence of these effects on mussel settlement, justifying the preservation of this feature within the MMI. However, since a higher clustering frequency was not more common in the high-performing macroalgae than in the low-performing macroalgae, this feature is only likely to be useful in the assessment of macroalgae as attachment substrates when a substrate exhibits a particularly low clustering frequency, suggesting a lower suitability for use in restoration. Evaluation of the Macroalgal Morphology Index At both sites, the species of macroalgae that supported the highest presence of attached P. canaliculus juveniles in each size class received moderate to high total scores according to the criteria in the MMI. This finding suggests that the MMI has a high capability of predicting the relative presence of juvenile mussels among different macroalgae species based on their morphology. Patterns in the macroalgal associations of P. canaliculus juveniles on two remnant mussel reefs indicate that macroalgae with morphological characteristics that receive a total index score of at least 14 out of 24 likely have a high potential to facilitate juvenile recruitment at mussel reef restoration sites. However, the observed macroalgal association patterns only partially validate the collective accuracy of the MMI’s scoring criteria because there were a number of other macroalgae species with total index scores of 14 or higher that supported a relatively low presence of P. canaliculus juveniles in each size class but especially in the two larger size classes. The MMI is limited in its ability to capture all aspects of a macroalgal substrate that can positively or negatively affect the attachment of juvenile mussels because in addition to a macroalga’s morphology, the chemical cues that it produces (Kelly 2001 ; Alfaro et al. 2006 ; Gribben et al. 2011 ), the microscale texture of its surfaces (Gribben et al. 2011 ; Loucks 2023 ; Wu and Jeffs 2025 ), and the location and environmental conditions in which it is normally found all have the potential to impact juvenile presence on that substrate (Hunt and Scheibling 1995 , 1996 , 2001 ). The scores and performances of G. caulacantheum and U. australis provide good examples of the limitations of the MMI. The red macroalga G. caulacantheum is a highly and thinly branched, turf-forming species that received a high total index score of 21, but it was a low-performing attachment substrate for juveniles in all size classes at Waipū Cove. Its total index score does not reflect the tendency for this species to grow exclusively on vertical rockfaces in the upper intertidal, where higher exposure to wave action and desiccation stress would make this substrate a much less suitable location for juvenile mussels to establish, especially as their susceptibility to these stressors increases with size. In contrast to G. caulacantheum , the sparsely and broadly branched green macroalga U. australis was a high-performing attachment substrate for juveniles up to 20 mm SL at Waipū Cove, but it only received a moderate total index score of 14. This score does not reflect the tendency for U. australis to grow in close proximity to available freshwater outputs (Nelson 2020 ), such as the stream that empties into the ocean on the north side of the mussel reef at Waipū Cove. These locations tend to experience a high input of allochthonous organic nutrients that can benefit the growth of both U. australis and juvenile mussels, which may explain why U. australis supported such a high presence of juvenile mussels at Waipū Cove but not at Pākiri Beach, where there are no nearby freshwater outputs. These findings underscore the need for mussel reef restoration practitioners using the MMI to interpret the total index score that a substrate receives with a reasonable amount of caution and consideration for whether that species of macroalgae is likely to grow best under environmental conditions that also benefit the growth and survival of juvenile mussels. The MMI’s criteria could potentially be improved by testing and modifying its scores based on additional observations of juvenile mussel presence among macroalgae at more sites that experience a variety of environmental conditions and support different species of macroalgae than those observed in this study. In particular, the MMI would benefit from being tested on macroalgal association data collected from subtidal mussel reefs, where the lack of desiccation stress and reduced wave action may promote different morphological features of macroalgae as the primary drivers of juvenile mussel attachment than those identified for intertidal mussel reefs. Additionally, more studies are needed to improve our understanding of the mechanisms through which each of these morphological features influences the settlement, survival, and establishment of juvenile mussels on macroalgae under a range of physical and biological stressors. Importantly, while the morphological features and scoring criteria used in the MMI were identified from literature on a wide range of mussel species, observations of the macroalgal associations of other species of juvenile mussels on remnant reefs would be valuable for confirming that the MMI is applicable for mussel species other than P. canaliculus . Conclusions When totalled, the scores assigned to each species of macroalgae for each of the eight morphological features in the MMI were able to predict the presence of attached juvenile P. canaliculus with a relatively high accuracy of 75% for juveniles smaller than 10 mm SL but a low accuracy of 40–60% for juveniles 10 mm SL and larger. The majority of morphological features used to create the MMI were identified from studies that focus on the impact of substrate morphology on larval and early juvenile mussel settlement, while very few studies have been conducted on how substrate morphology influences the retention of juveniles throughout later development. Therefore, many of these features, in particular the features of macroalgae that describe their branching morphology, while useful in predicting a macroalga’s potential to support the initial attachment of settling juveniles, become largely irrelevant as juveniles attach less frequently to the branches of macroalgae and more frequently to the holdfasts as they grow. Above 10 mm SL, juveniles likely become more susceptible to sources of post-settlement detachment, stress, or mortality that appear to be primarily mediated by just two of the eight features in the MMI, macroalgal canopy cover and holdfast complexity. Therefore, the total index score appears to be collectively useful for predicting a macroalga’s potential to support the settlement of larvae and smaller juvenile mussels. However, evaluating the likelihood that a substrate will aid the continued attachment of larger juvenile mussels may require users of the MMI to give greater weight to scores for canopy cover and holdfast complexity as well as consider whether environmental conditions at the proposed restoration site are likely to have a positive or negative interaction with canopy height. Additionally, it is possible that if multiple macroalgae species with different sets of high-scoring morphological characteristics are transplanted into restoration sites together, such as species that exhibit thinner branch widths, a higher degree of branching, shorter branch spacing, and multiple planes of branching being paired with other species that exhibit more complex holdfasts and widespread canopy covers, these substrates may work together to secure the attachment of juvenile mussels throughout all stages of their development. Overall, the results of this study suggest that the MMI can be used to accurately rank various macroalgal substrates for their potential to support the recruitment of juvenile mussels based on the morphology of the substrate. The quantitative and qualitative criteria outlined in the MMI are designed to ensure that macroalgal substrata are assessed and scored in a manner that is objective and repeatable for its users. Additionally, the MMI can begin to remove the ambiguity associated with the wide variety of terms for differing macroalgal morphology that have been used throughout the literature to describe the settlement preferences of juvenile mussels, instead condensing these descriptions into just three scored variations of eight overarching structural features of macroalgae. Ultimately, the MMI can serve as a useful tool for mussel reef restoration practitioners seeking to: 1) determine the most suitable species of macroalgae to transplant into current restoration sites, 2) assess new restoration sites with existing macroalgal habitat for their potential to support juvenile recruitment, and 3) apply the highest scoring morphological characteristics outlined in the MMI to the design of artificial substrata for particularly degraded restoration sites that are unlikely to support the growth of macroalgal substrates. While the MMI’s scoring criteria were only tested on the attachment substrates of P. canaliculus juveniles, its application extends to all mussel species that are the focus of restoration efforts. The MMI and the methods used in this study to validate its accuracy provide a solid foundation for future studies evaluating the effective use of macroalgal substrates in mussel reef restoration projects across the globe. Declarations Acknowledgements We thank and acknowledge the tangata whenua (indigenous people of the land) of Waipū Cove and Pākiri Beach as the traditional owners and kaitiaki (guardians) of the land, coast, and oceans within which this study was undertaken. We would also like to thank The Nature Conservancy Aotearoa New Zealand for providing the funds to make this research possible. Funding Funding for this research was provided by The Nature Conservancy Aotearoa New Zealand. Conflict of Interest All authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Authors’ Contribution Katherine A. Burnham : Writing – original draft, Visualization,Validation, Methodology, Investigation, Formal analysis, Data curation,Conceptualization. Jenny R. Hillman : Writing – review & editing,Validation, Supervision, Resources, Project administration, Methodology,Funding acquisition, Conceptualization. Andrew G. Jeffs : Writing– review & editing, Validation, Supervision, Resources, Project administration,Methodology, Funding acquisition, Conceptualization. Ethics Approval This study was conducted in compliance with ethical standards and all applicable guidelines for sampling of organisms have been followed. Data Availability The datasets generated and analysed during the current study are available in the Mendeley data repository, [10.17632/2dy4hgn9fw.1, 10.17632/fd2fhwxfdp.1]. References Alder A, Jeffs A, Hillman JR (2020) Considering the use of subadult and juvenile mussels for mussel reef restoration. Restor Ecol 29(3):e13322. https://doi.org/10.1111/rec.13322 Alfaro AC, Jeffs AG, Hooker SH (2001) Reproductive behavior of the green-lipped mussel, Perna canaliculus , in Northern New Zealand. Bull Mar Sci 69(3). https://www.researchgate.net/publication/233520444 Alfaro AC, Jeffs AG (2002) Small-scale mussel settlement patterns within morphologically distinct substrates at Ninety Mile Beach, northern New Zealand. Malacologia 44:1–16. https://www.researchgate.net/publication/285776689 Alfaro AC, Jeffs AG, Creese RG (2004) Bottom-drifting algal/mussel spat associations along a sandy coastal region in northern New Zealand. Aquaculture 241(1–4):269–290. https://doi.org/10.1016/j.aquaculture.2004.07.029 Alfaro AC, Copp BR, Appleton DR, Kelly S, Jeffs AG (2006) Chemical cues promote settlement in larvae of the green-lipped mussel, Perna canaliculus . Aquacult Int 14:405–412. https://doi.org/10.1007/s10499-005-9041-y Aghzar A, Talbaoui M, Benajiba MH, Presa P (2012) Influence of depth and diameter of rope collectors on settlement density of Mytilus galloprovincialis spat in Baie de M’diq (Alboran Sea). Mar Freshw Behav Physiol 45(1):51–61. https://doi.org/10.1080/10236244.2012.678759 Banke TL, Steinfurth RC, Lange T, Canal‐Vergés P, Svane N, Flindt MR (2024) Dislodgement and mortality challenges when restoring shallow mussel beds ( Mytilus edulis ) in a Danish estuary. Restor Ecol 32(5):e14160. https://doi.org/10.1111/rec.14160 Bayne BL (1964) Primary and secondary settlement in Mytilus edulis L. (Mollusca). J Anim Ecol 33(3):513. https://doi.org/10.2307/2569 Beck MW, Brumbaugh RD, Airoldi L, Carranza A, Coen LD, Crawford C, Defeo O, Edgar GJ, Hancock B, Kay M, Lenihan H, Luckenbach MW, Toropova CL, Zhang G (2009) Shellfish Reefs at Risk: A Global Analysis of Problems and Solutions. The Nature Conservancy, Arlington, Virginia, United States of America. https://www.researchgate.net/publication/265238947 Bégin C, Johnson LE, Himmelman JH (2004) Macroalgal canopies: distribution and diversity of associated invertebrates and effects on the recruitment and growth of mussels. Mar Ecol Prog Ser 271:121–132. https://doi.org/10.3354/meps271121 Benjamin ED, Handley SJ, Hale R, Toone TA, Jeffs A, Hillman JR (2022a) Biodiversity associated with restored small-scale mussel habitats has restoration decision implications. Biodivers Conserv 31(11):2833–2855. https://doi.org/10.1007/s10531-022-02462-1 Benjamin ED, Hillman JR, Handley SJ, Toone TA, Jeffs A (2022b) The effectiveness of providing shell substrate for the restoration of adult mussel reefs. Sustainability 14(23):15746. https://doi.org/10.3390/su142315746 Benjamin E, Jeffs A, Handley S, Toone T, Hillman J (2023) Determining restoration potential by transplanting mussels of different size classes over a range of aerial exposures. Mar Ecol Prog Ser 713:71–81. https://doi.org/10.3354/meps14337 Beukema JJ, Cadée GC (1996) Consequences of the sudden removal of nearly all mussels and cockles from the Dutch Wadden Sea. Mar Ecol 17(1-3):279–289. https://doi.org/10.1111/j.1439-0485.1996.tb00508.x van den Bogaart LA, Schotanus J, Capelle JJ, Bouma TJ (2023) Using a biodegradable substrate to increase transplantation success: effect of density and sediment on aggregation behavior of mussels. Ecol Eng 196:107096. https://doi.org/10.1016/j.ecoleng.2023.107096 Bourget E, DeGuise J, Daigle G (1994) Scales of substratum heterogeneity, structural complexity, and the early establishment of a marine epibenthic community. J Exp Mar Biol Ecol 181(1):31–51. https://doi.org/10.1016/0022-0981(94)90102-3 Brenner M, Buck BH (2010) Attachment properties of blue mussel ( Mytilus edulis L.) byssus threads on culture-based artificial collector substrates. Aquac Eng 42(3):128–139. https://doi.org/10.1016/j.aquaeng.2010.02.001 Brumbaugh RD, Beck MW, Coen LD, Craig L, Hicks P (2006) A practitioners’ guide to the design and monitoring of shellfish: an ecosystem services approach. MRD Education Report, The Nature Conservancy, Arlington, Virginia, United States of America. https://repository.library.noaa.gov/view/noaa/3470 Brumbaugh RD, Coen LD (2009) Contemporary approaches for small-scale oyster reef restoration to address substrate versus recruitment limitation: a review and comments relevant for the Olympia oyster, Ostrea lurida Carpenter 1864. J Shellfish Res 28(1):147–161. https://doi.org/10.2983/035.028.0105 Buchanan S, Babcock R (1997) Primary and secondary settlement by the green-shell mussel Perna canaliculus . Oceanogr Lit Rev 12:1500. https://www.researchgate.net/publication/279937722 Bulleri F, Airoldi L, Branca GM, Abbiati M (2006) Positive effects of the introduced green alga, Codium fragile ssp. tomentosoides , on recruitment and survival of mussels. Mar Biol 148:1213–1220. https://doi.org/10.1007/s00227-005-0181-4 Burnham KA, Hillman JR, Jeffs AG (2025) Macroalgae and conspecifics serve as critical attachment substrates for juvenile green-lipped mussels ( Perna canaliculus ) during different stages of development. J Exp Mar Biol Ecol 588:152108. https://doi.org/10.1016/j.jembe.2025.152108 Capelle JJ, Leuchter L, de Wit M, Hartog E, Bouma TJ (2019) Creating a window of opportunity for establishing ecosystem engineers by adding substratum: a case study on mussels. Ecosphere 10(4):e02688. https://doi.org/10.1002/ecs2.2688 Carl C, Poole AJ, Sexton BA, Glenn FL, Vucko MJ, Williams MR, Whalan S, de Nys R (2012a) Enhancing the settlement and attachment strength of pediveligers of Mytilus galloprovincialis by changing surface wettability and microtopography. Biofouling 28(2):175–186. https://doi.org/10.1080/08927014.2012.662676 Carl C, Poole AJ, Williams MR, de Nys R (2012b) Where to settle – settlement preferences of Mytilus galloprovincialis and choice of habitat at a micro spatial scale. PLoS ONE 7(12):e52358. https://doi.org/10.1371/journal.pone.0052358 Carranza A, Zu Ermgassen PS (2020) A global overview of restorative shellfish mariculture. Front Mar Sci 7:722. https://doi.org/10.3389/fmars.2020.00722 Christensen HT, Dolmer P, Hansen BW, Holmer M, Kristensen LD, Poulsen LK, Sternberg C, Albertsen CM, Støttrup JG (2015) Aggregation and attachment responses of blue mussels, Mytilus edulis – impact of substrate composition, time scale and source of mussel seed. Aquaculture 435:245–251. https://doi.org/10.1016/j.aquaculture.2014.09.043 Commito JA, Commito AE, Platt RV, Grupe BM, Piniak WED, Gownaris NJ, Reeves KA, Vissichelli AM (2014) Recruitment facilitation and spatial pattern formation in soft‐bottom mussel beds. Ecosphere 5(12):1–26. https://doi.org/10.1890/ES14-00200.1 de Nesnera KL (2016) Stress, ontogeny, and movement determine the relative importance of facilitation for juvenile mussels. Ecol 97(9):2199–2205. https://doi.org/10.1002/ecy.1505 de Paoli H, van de Koppel J, van der Zee E, Kangeri A, van Belzen J, Holthuijsen S, van den Berg A, Herman P, Olff H, van der Heide T (2015) Processes limiting mussel bed restoration in the Wadden Sea. J Sea Res 103:42-49. https://doi.org/10.1016/j.seares.2015.05.008 Dobretsov SV (1999) Effects of macroalgae and biofilm on settlement of blue mussel ( Mytilus edulis L.) larvae. Biofouling 14(2):153–165. https://doi.org/10.1080/08927019909378406 Dobretsov S, Wahl M (2001) Recruitment preferences of blue mussel spat ( Mytilus edulis ) for different substrates and microhabitats in the White Sea (Russia). Hydrobiologia 445:27–35. https://doi.org/10.1023/A:1017502126707 Dobretsov S, Wahl M (2008) Larval recruitment of the blue mussel Mytilus edulis : the effect of flow and algae. J Exp Mar Biol Ecol 355(2):137–144. https://doi.org/10.1016/j.jembe.2007.12.018 Donker JJA, Van Der Vegt M, Hoekstra P (2013) Wave forcing over an intertidal mussel bed. J Sea Res 82:54–66. https://doi.org/10.1016/j.seares.2012.08.010 Eckman JE, Duggins DO (1991) Life and death beneath macrophyte canopies: effects of understory kelps on growth rates and survival of marine, benthic suspension feeders. Oecologia 87:473–487. https://doi.org/10.1007/BF00320409 Elsevier (2025) Scopus. https://www.scopus.com Erlandsson J, McQuaid C (2004) Spatial structure of recruitment in the mussel Perna perna at local scales: effects of adults, algae and recruit size. Mar Ecol Prog Ser 267:173–185. https://doi.org/10.3354/meps267173 Erlandsson J, McQuaid CD, Kostylev VE (2005) Contrasting spatial heterogeneity of sessile organisms within mussel ( Perna perna L.) beds in relation to topographic variability. J Exp Mar Biol Ecol 314(1):79–97. https://doi.org/10.1016/j.jembe.2004.09.010 Filgueira R, Peteiro LG, Labarta U, Fernández-Reiriz MJ (2007) Assessment of spat collector ropes in Galician mussel farming. Aquac Eng 37(3):195–201. https://doi.org/10.1016/j.aquaeng.2007.06.001 Eriksson BK, van der Heide T, van de Koppel J, Piersma T, van der Veer HW, Olff H (2010) Major changes in the ecology of the Wadden Sea: human impacts, ecosystem engineering and sediment dynamics. Ecosyst 13:752–764. https://doi.org/10.1007/s10021-010-9352-3 Erlandsson J, Porri F, McQuaid CD (2008) Ontogenetic changes in small-scale movement by recruits of an exploited mussel: implications for the fate of larvae settling on algae. Mar Biol 153(3):365–373. https://doi.org/10.1007/s00227-007-0812-z Fitzsimons J, Branigan S, Brumbaugh RD, McDonald T, Zu Ermgassen PSE (2019) Restoration Guidelines for Shellfish Reefs. The Nature Conservancy, Arlington, Virginia, United States of America. https://doi.org/10.13140/RG.2.2.22511.38563 Fitzsimons JA, Branigan S, Gillies CL, Brumbaugh RD, Cheng J, DeAngelis BM, Geselbracht L, Hancock B, Jeffs A, McDonald T, McLeod IM, Pogoda B, Theuerkauf SJ, Thomas M, Westby S, Zu Ermgassen PSE (2020) Restoring shellfish reefs: global guidelines for practitioners and scientists. Conserv Sci Pract 2(6):e198. https://doi.org/10.1111/csp2.198 Ford JR, Hamer P (2016) The forgotten shellfish reefs of coastal Victoria: documenting the loss of a marine ecosystem over 200 years since European settlement. Proceedings of the Royal Society of Victoria 128(1):87. https://doi.org/10.1071/RS16008 Frandsen R, Dolmer P (2002) Effects of substrate type on growth and mortality of blue mussels ( Mytilus edulis ) exposed to the predator Carcinus maenas . Mar Biol 141:253–262. https://doi.org/10.1007/s00227-002-0831-8 Fréchette M, Bourget E (1985) Food-limited growth of Mytilus edulis L. in relation to the benthic boundary layer. Can J Fish Aquat Sci 42(6):1166–1170. https://doi.org/10.1139/f85-144 Gagnon K, Hartvig C, Didderen K, Fagerli CW, Govers LL, Gräfnings MLE, Heusinkveld JHT, Kaljurand K, Lengkeek W, Martin G, Meysick L, Pajusalu L, Rinde E, van der Heide T, Boström C (2021) Incorporating facilitative interactions into small‐scale eelgrass restoration—challenges and opportunities. Restor Ecol 29(5):e13398. https://doi.org/10.1111/rec.13398 Gann GD, McDonald T, Walder B, Aronson J, Nelson CR, Jonson J, Hallett JG, Eisenberg C, Guariguata MR, Liu J, Hua F, Echeverría C, Gonzales E, Shaw N, Decleer K, Dixon KW (2019) International principles and standards for the practice of ecological restoration. Restor Ecol 27(S1):S1–S46. https://doi.org/10.1111/rec.13035 Greenway JPC (1969) Survey of mussels (Mollusca: Lamellibranchia) in the firth of Thames, 1961–1967. N Z J Mar Freshw Res 3:304–317. https://ref.coastalrestorationtrust.org.nz/site/assets/files/9479/surveys_of_mussels_mollusca_lamellibranchia_in_the_firth_of_thames_1961_67.pdf Gribben PE, Jeffs AG, de Nys R, Steinberg PD (2011) Relative importance of natural cues and substrate morphology for settlement of the New Zealand Greenshell ™ mussel, Perna canaliculus . Aquaculture 319(1–2):240–246. https://doi.org/10.1016/j.aquaculture.2011.06.026 Harvey M, Bourget E, Ingram RG (1995) Experimental evidence of passive accumulation of marine bivalve larvae on filamentous epibenthic structures. Limnol Oceanogr 40(1):94–104. https://doi.org/10.4319/lo.1995.40.1.0094 Hauraki Gulf Forum (2023) State of our Gulf Report. https://gulfjournal.org.nz/wp- content/uploads/2023/08/SOER-sum-online.pdf Holthuis TD, Bergström P, Lindegarth M, Lindegarth S (2015) Monitoring recruitment patterns of mussels and fouling tunicates in mariculture. J Shellfish Res 34(3):1007–1018. https://doi.org/10.2983/035.034.0327 Hunt HL, Scheibling R (1995) Structure and dynamics of mussel patches in tidepools on a rocky shore in Nova Scotia, Canada. Mar Ecol Prog Ser 124:105–115. https://doi.org/10.3354/meps124105 Hunt H, Scheibling R (1996) Physical and biological factors influencing mussel ( Mytilus trossulus, M. edulis) settlement on a wave-exposed rocky shore. Mar Ecol Prog Ser 142:135–145. https://doi.org/10.3354/meps142135 Hunt H, Scheibling R (1998) Spatial and temporal variability of patterns of colonization by mussels ( Mytilus trossulus, M. edulis ) on a wave-exposed rocky shore. Mar Ecol Prog Ser 167:155–169. https://doi.org/10.3354/meps167155 Hunt HL, Scheibling RE (2001) Patch dynamics of mussels on rocky shores: integrating process to understand pattern. Ecol 82(11):3213–3231. https://www.jstor.org/stable/2679845 Kelly S (2001) Improving the supply of mussel spat. Kinnears Ropes Ltd. and Kaitaia Spat Ltd. Technical Report. Kobak J (2001) Light, gravity and conspecifics as cues to site selection and attachment behaviour of juvenile and adult Dreissena polymorpha Pallas, 1771. J Molluscan Stud 67(2):183–189. https://doi.org/10.1093/mollus/67.2.183 Lanham BS, Pomeroy AW, Swearer SE, Marusic I, Javaherchian J, Morris RL (2025) Substrate-mediated alterations to hydrodynamic conditions enhances shellfish larval settlement: implications for artificial reef restoration. Ecol Eng 212:107474. https://doi.org/10.1016/j.ecoleng.2024.107474 Lao NCH (2016) Rehabilitation of Red Macroalgae in the Hauraki Gulf. Master’s Thesis, University of Auckland. https://hdl.handle.net/2292/31829 Leonard GH (1999) Positive and negative effects of intertidal algal canopies on recruitment and survival of barnacles. Mar Ecol Prog Ser 178:241–249. https://www.jstor.org/stable/24852065 Lipcius RN, Burke RP (2018) Successful recruitment, survival and long-term persistence of eastern oyster and hooked mussel on a subtidal, artificial restoration reef system in Chesapeake Bay. PLoS ONE 13(10):e0204329. https://doi.org/10.1371/journal.pone.0204329 Liu W, Xu M, Wang Z, Yu D, Zhour X (2017) Experimental study of attachment characteristics of golden mussel larvae. Acta Ecol Sin 37(8):2779–2787. https://doi.org/10.5846/stxb201601030010 Lotze HK, Lenihan HS, Bourque BJ, Bradbury RH, Cooke RG, Kay MC, Kidwell SM, Kirby MX, Peterson CH, Jackson JBC (2006) Depletion, degradation, and recovery potential of estuaries and coastal seas. Science 312(5781):1806–1809. https://www.jstor.org/stable/3846557 Loucks AJ (2023) Settlement preferences of green-lipped mussel larvae on key seaweed species from Hauraki Gulf. Honours Thesis, University of Auckland. Manighetti B, Carter L (1999) Across-shelf sediment dispersal, Hauraki Gulf, New Zealand. Mar Geol 160(3):271–300. https://doi.org/10.1016/S0025-3227(99)00024-9 Marsden JE, Lansky DM (2000) Substrate selection by settling zebra mussels, Dreissena polymorpha , relative to material, texture, orientation, and sunlight. Can J Zool 78(5):787–793. https://doi.org/10.1139/z00-004 McAfee D, Connell SD (2020) Cuing oyster recruitment with shell and rock: implications for timing reef restoration. Restor Ecol 28(3):506–511. https://doi.org/10.1111/rec.13134 McAfee D, McLeod LD, Carruthers S, Connell SD (2024) Reversing functional extinction: successful restoration of eradicated oyster reefs. Restor Ecol 32(5):e14169. https://doi.org/10.1111/rec.14169 McCay DPF, Peterson CH, DeAlteris JT, Catena J (2003) Restoration that targets function as opposed to structure: replacing lost bivalve production and filtration. Mar Ecol Prog Ser 264:197–212. https://www.jstor.org/stable/24867511 McCook LJ, Chapman ARO (1991) Community succession following massive ice-scour on an exposed rocky shore: effects of Fucus canopy algae and of mussels during late succession. J Exp Mar Bio Ecol 154(2):137–169. https://doi.org/10.1016/0022-0981(91)90161-O McLeod IM (2009) Green-lipped mussels, Perna canaliculus , in soft-sediment systems in northeastern New Zealand. Master’s Thesis, University of Auckland. https://doi.org/10.13140/2.1.2822.3046 McLeod IM, Parsons DM, Morrison MA, Le Port A, Taylor RB (2012) Factors affecting the recovery of soft-sediment mussel reefs in the Firth of Thames, New Zealand. Mar Freshwater Res 63(1):78. https://doi.org/10.1071/MF11083 McLeod IM, Parsons DM, Morrison MA, Van Dijken SG, Taylor RB (2014) Mussel reefs on soft sediments: a severely reduced but important habitat for macroinvertebrates and fishes in New Zealand. N Z J Mar Freshw Res 48(1):48–59. https://doi.org/10.1080/00288330.2013.834831 McQuaid CD, Lindsay JR (2005) Interacting effects of wave exposure, tidal height and substrate on spatial variation in densities of mussel Perna perna plantigrades. Mar Ecol Prog Ser 301:173–184. https://doi.org/10.3354/meps301173 Meadows PS, Meadows A, West FJC, Shand PS, Shaikh MA (1998) Mussels and mussel beds ( Mytilus edulis ) as stabilizers of sedimentary environments in the intertidal zone. Geological Society, London, Special Publications 139(1):331–347. https://doi.org/10.1144/GSL.SP.1998.139.01.26 Menge BA (1978) Predation intensity in a rocky intertidal community. Oecologia 34:1–16. https://doi.org/10.1007/BF00346237 Moody JA, Gentry MJ, Bouboulis SA, Kreeger DA (2020) Effects of substrate (protection and type) on ribbed mussel ( Geukensia demissa ) recruitment for living shoreline applications. J Coast Res 36(3):619–627. https://doi.org/10.2112/JCOASTRES-D-19-00062.1 Moreno CA (1995) Macroalgae as a refuge from predation for recruits of the mussel Choromytilus chorus (Molina, 1782) in Southern Chile. J Exp Mar Bio Ecol 191(2):181–193. https://doi.org/10.1016/0022-0981(95)00050-2 Morrison MA, Lowe M, Parsons D, Usmar N, McLeod I (2009) A review of land-based effects on coastal fisheries and supporting biodiversity in New Zealand. Wellington, New Zealand: Ministry of Fisheries. https://ref.coastalrestorationtrust.org.nz/documents/a-review-of-land-based-effects-on-coastal-fisheries-and-supporting-biodiversity-in-new-zealand-1/ Nelson WA (2020) New Zealand seaweeds: an illustrated guide. Te Papa Press. O’Connor NE, Crowe TP, McGrath D (2006) Effects of epibiotic algae on the survival, biomass and recruitment of mussels, Mytilus L. (Bivalvia: Mollusca). J Exp Mar Bio Ecol 328(2):265–276. https://doi.org/10.1016/j.jembe.2005.07.013 Overton K, Dempster T, Swearer SE, Morris RL, Barrett LT (2024) Predictors of outplanted marine bivalve survival in restoration: a review and synthesis. J Appl Ecol 61(12):2884–2896. https://doi.org/10.1111/1365-2664.14795 Paul LJ (2012) A history of the Firth of Thames dredge fishery for mussels: use and abuse of a coastal resource. New Zealand Aquatic Environment and Biodiversity Report No. 94. NIWA. https://www.mpi.govt.nz/dmsdocument/4016/direct Protopopescu GC, Beal BF (2015) Settlement response to various rope substrates in blue mussels ( Mytilus edulis Linnaeus) in a hatchery setting. J Shellfish Res 34(2):383–391. https://doi.org/10.2983/035.034.0221 R Core Team (2024) R: a language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. https://www.R-project.org/ Reid B (1969) Mussel survey Hauraki Gulf and Firth of Thames. 1958 Fisheries Technical Report No. 34, New Zealand Marine Department, Wellington, New Zealand. Reusch TB, Williams SL (1999) Macrophyte canopy structure and the success of an invasive marine bivalve. Oikos 398–416. https://www.jstor.org/stable/3546420?seq=1andcid=pdf- Rowden AA, Berkenbusch K, Brewin PE, Dalen J, Neill KF, Nelson WA, Oliver MD, Probert PK, Schwarz A, Sui PH, Sutherland D (2012) A review of the marine soft-sediment assemblages of New Zealand. Wellington, New Zealand: Ministry for Primary Industries. https://www.researchgate.net/publication/306394493 Schotanus J, Capelle JJ, Paree E, Fivash GS, Van De Koppel J, Bouma TJ (2020) Restoring mussel beds in highly dynamic environments by lowering environmental stressors. Restor Ecol 28(5):1124–1134. https://doi.org/10.1111/rec.13168 Sea MA, Hillman JR, Thrush SF (2022) Enhancing multiple scales of seafloor biodiversity with mussel restoration. Sci Rep 12(1):5027. https://doi.org/10.1038/s41598-022-09132-w Seed R, Suchanek TH (1992) Population and community ecology of Mytilus. In E. Gosling (Ed.), The mussel Mytilus : ecology, physiology, genetics and culture (pp. 87–169). New York, NY: Elsevier. https://www.scirp.org/reference/referencespapers?referenceid= 1139224 Skelton BM, Jeffs AG (2020) The importance of physical characteristics of settlement substrate to the retention and fine-scale movements of Perna canaliculus spat in suspended longline aquaculture. Aquaculture 521:735054. https://doi.org/10.1016/j.aquaculture.2020.735054 Skelton BM, Jeffs AG (2021) An assessment of the use of macroalgae to improve the retention of Greenshell ™ mussel ( Perna canaliculus ) spat in longline culture. Aquac Int 29(4):1683–1695. https://doi.org/10.1007/s10499-021-00710-9 Suplicy FM (2020) A review of the multiple benefits of mussel farming. Rev Aquacult 12(1):204–223. https://doi.org/10.1111/raq.12313 Temmink RJM, Angelini C, Fivash GS, Swart L, Nouta R, Teunis M, Lengkeek W, Didderen K, Lamers LPM, Bouma TJ, Van Der Heide T (2021) Life cycle informed restoration: engineering settlement substrate material characteristics and structural complexity for reef formation. J Appl Ecol 58(10):2158–2170. https://doi.org/10.1111/1365-2664.13968 Temmink RJM, Fivash GS, Govers LL, Nauta J, Marin-Diaz B, Cruijsen PMJM, Didderen K, Penning E, Olff H, Heusinkveld JHT, Lamers LPM, Lengkeek W, Christianen MJA, Reijers VC, Bouma TJ, Van Der Heide T (2022) Initiating and upscaling mussel reef establishment with life cycle informed restoration: successes and future challenges. Ecol Eng 175:106496. https://doi.org/10.1016/j.ecoleng.2021.106496 Toone TA, Benjamin ED, Hillman JR, Handley S, Jeffs A (2023a) Multidisciplinary baselines quantify a drastic decline of mussel reefs and reveal an absence of natural recovery. Ecosphere 14(3):e4390. https://doi.org/10.1002/ecs2.4390 Toone TA, Hillman JR, Benjamin ED, Handley S, Jeffs AG (2023b) Out of their depth: the successful use of cultured subtidal mussels for intertidal restoration. Conserv Sci Prac 5(4):e12914. https://doi.org/10.1111/csp2.12914 Toone TA, Hillman JR, South PM, Benjamin ED, Handley S, Jeffs AG (2023c) Bottlenecks and barriers: patterns of abundance in early mussel life stages reveal a potential obstacle to reef recovery. Aquat Conserv 33(8):810–821. https://doi.org/10.1002/aqc.3979 Toone TA, Hillman JR, Benjamin ED, Handley S, Jeffs AG (2023d) Provision of early mussel life stages via macroalgae enhances recruitment and uncovers a novel restoration technique. J Exp Mar Bio Ecol 566:151919. https://doi.org/10.1016/j.jembe.2023.151919 van der Heide T, Tielens E, van der Zee EM, Weerman EJ, Holthuijsen S, Eriksson BK, Piersma T, van de Koppel J, Olff H (2014) Predation and habitat modification synergistically interact to control bivalve recruitment on intertidal mudflats. Biol Conserv 172:163–169. https://doi.org/10.1016/j.biocon.2014.02.036 van der Schatte Olivier A, Jones L, Vay LL, Christie M, Wilson J, Malham SK (2020) A global review of the ecosystem services provided by bivalve aquaculture. Rev Aquacult 12(1):3–25. https://doi.org/10.1111/raq.12301 Walter U, Liebezeit G (2003) Efficiency of blue mussel ( Mytilus edulis ) spat collectors in highly dynamic tidal environments of the Lower Saxonian coast (southern North Sea). Biomol Eng 20(4–6):407–411. https://doi.org/10.1016/S1389-0344(03)00064-9 Westerbom M, Mustonen O, Kilpi M (2008) Distribution of a marginal population of Mytilus edulis : responses to biotic and abiotic processes at different spatial scales. Mar Biol 153(6):1153–1164. https://doi.org/10.1007/s00227-007-0886-7 Wilcox M, Jeffs A (2017) Is attachment substrate a prerequisite for mussels to establish on soft-sediment substrate? J Exp Mar Bio Ecol 495:83–88. https://doi.org/10.1016/j.jembe.2017.07.004 Wilcox M, Kelly S, Jeffs A (2018) Ecological restoration of mussel beds onto soft-sediment using transplanted adults. Restor Ecol 26(3):581–590. https://doi.org/10.1111/rec.12607 Wu W (2018) The role of physical structure in the attachment of juvenile green-lipped mussels. Master’s Thesis, University of Auckland. https://hdl.handle.net/2292/45020 Wu W, Jeffs AG (2025) Influence of microstructure of substrate surface on the attachment of juvenile mussels. Fishes 10(3):135. https://doi.org/10.3390/fishes10030135 Wu W, Anderson I, Jeffs AG (2025) The role of substrates width and millimeter scale surface micro-structure in the attachment of juvenile mussels. Aquaculture 595:741593. https://doi.org/10.1016/j.aquaculture.2024.741593 Yang JL, Satuito CG, Bao WY, Kitamura H (2007) Larval settlement and metamorphosis of the mussel Mytilus galloprovincialis on different macroalgae. Mar Biol 152:1121–1132. https://doi.org/10.1007/s00227-007-0759-0 Zhao S, Zhang B, Yang J, Zhou J, Xu Y (2024) Linear discriminant analysis. Nature Reviews Methods Primers 4(1):70. https://doi.org/10.1038/s43586-024-00346-y Tables Tables 1 - 4 are available in the Supplementary Files section. Supplementary Files Table1.pptx Table2.pptx Table3.pptx Table4.pptx Burnhametal.SupplementaryMaterial.pdf Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-6745848","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":466747078,"identity":"c669ebb3-ae34-4b74-ac54-ff0cf858cba8","order_by":0,"name":"Katherine A. Burnham","email":"data:image/png;base64,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","orcid":"https://orcid.org/0009-0006-5039-5137","institution":"University of Auckland","correspondingAuthor":true,"prefix":"","firstName":"Katherine","middleName":"A.","lastName":"Burnham","suffix":""},{"id":466747079,"identity":"bed94ce9-ff67-4e5f-ba9e-36dc8410b050","order_by":1,"name":"Jenny R. Hillman","email":"","orcid":"","institution":"University of Auckland","correspondingAuthor":false,"prefix":"","firstName":"Jenny","middleName":"R.","lastName":"Hillman","suffix":""},{"id":466747080,"identity":"1d0f3161-9c24-4263-ae34-c177c1c57273","order_by":2,"name":"Andrew G. Jeffs","email":"","orcid":"","institution":"University of Auckland","correspondingAuthor":false,"prefix":"","firstName":"Andrew","middleName":"G.","lastName":"Jeffs","suffix":""}],"badges":[],"createdAt":"2025-05-25 22:49:03","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6745848/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6745848/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":84205256,"identity":"8719856f-2781-453f-8994-f3e4e82fbe6e","added_by":"auto","created_at":"2025-06-09 09:02:30","extension":"jpg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":3122610,"visible":true,"origin":"","legend":"\u003cp\u003eThe eight morphological features of macroalgae identified as potential determinants of juvenile mussel attachment were canopy cover, canopy height, holdfast complexity, clustering frequency, planes of branching, branch spacing, degree of branching, and branch width. The morphological measurements collected for each species of macroalgae to generate index scores for each of these features are shown (in bold text) alongside the calculations used (in plain text).\u003c/p\u003e","description":"","filename":"Fig1.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6745848/v1/3c713983f9818e017f706d3f.jpg"},{"id":84205268,"identity":"85124b82-e013-4801-a68e-ef951183c987","added_by":"auto","created_at":"2025-06-09 09:02:31","extension":"jpg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":2576525,"visible":true,"origin":"","legend":"\u003cp\u003eDiagram of the six possible locations of juvenile mussel attachment on macroalgae recorded during reef sampling. The potential attachment locations were the holdfast, stipe, lower branch nodes, lower branches, upper branch nodes, and upper branches of the macroalga. Lower branches and nodes were considered those located below half of the canopy height, while upper branches and nodes were those located in the upper half of the canopy height.\u003c/p\u003e","description":"","filename":"Fig2.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6745848/v1/65e74f74f89399449031f3f5.jpg"},{"id":84205254,"identity":"20d44577-c154-4c00-a12e-43b901693ed2","added_by":"auto","created_at":"2025-06-09 09:02:30","extension":"jpg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":1547936,"visible":true,"origin":"","legend":"\u003cp\u003eJuvenile mussel presence, or the proportion (%) of each macroalgal substrate sampled that exhibited attachments by juvenile \u003cem\u003eP. canaliculus\u003c/em\u003e, plotted against the total index score assigned to each species of macroalga by the Macroalgal Morphology Index. The results are presented for three 10 mm (shell length; SL) size classes of juvenile mussels at two sites (Waipū Cove and Pākiri Beach). Abbreviations next to each data point denote the species of macroalgal substrate they represent. The solid black trendlines show the linear relationship between the total index score of each macroalgal substrate and attached juvenile mussel presence in each size class. The dotted grey lines indicate the upper and lower bounds of the 95% confidence interval calculated for the presence of attached juvenile mussels in each size class. Macroalgal substrates were considered low-performing attachment substrates if they exhibited juvenile presence lower than the interval, moderate-performing attachment substrates if they exhibited presence within the interval, and high-performing attachment substrates if they exhibited presence higher than the interval.\u003c/p\u003e","description":"","filename":"Fig3.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6745848/v1/ef694f969afedaa2fc0c21b7.jpg"},{"id":84205252,"identity":"e603e348-6ac0-4372-8f16-285f6c15c08f","added_by":"auto","created_at":"2025-06-09 09:02:30","extension":"jpg","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":960434,"visible":true,"origin":"","legend":"\u003cp\u003eJuvenile mussel presence, or the proportion (%) of each macroalgal substrate sampled that exhibited attachments by juvenile \u003cem\u003eP. canaliculus\u003c/em\u003e, for each possible location of attachment on the macroalga. Juvenile mussels in three 10 mm (shell length; SL) size classes across two sites (Waipū Cove and Pākiri Beach) are shown. The six possible locations of mussel attachment on a macroalga were the holdfast (HF), stipe (ST), lower branch node (LBN), lower branch (LB), upper branch node (UBN), and upper branch (UB). Only statistical comparisons made between attachment locations within size classes, not between size classes, are displayed. Letters above the columns indicate between which attachment locations differences in juvenile presence are significant (α = 0.05), where different letters indicate a statistical difference, and similar letters indicate statistical similarity.\u003c/p\u003e","description":"","filename":"Fig4.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6745848/v1/93c2c8b5ab8e83c2f9230f3a.jpg"},{"id":85600559,"identity":"8f3ccbbe-63b6-471b-962e-5b8b387c2d6f","added_by":"auto","created_at":"2025-06-28 21:04:29","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":8913037,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6745848/v1/183572eb-4b64-491c-a431-33031a3846fa.pdf"},{"id":84205251,"identity":"f5b45330-2f3d-481d-8656-8220bd87668d","added_by":"auto","created_at":"2025-06-09 09:02:30","extension":"pptx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":73542,"visible":true,"origin":"","legend":"","description":"","filename":"Table1.pptx","url":"https://assets-eu.researchsquare.com/files/rs-6745848/v1/374775f99340820334b90d2c.pptx"},{"id":84205272,"identity":"810894e1-2fde-4993-9982-7f8683898b3d","added_by":"auto","created_at":"2025-06-09 09:02:31","extension":"pptx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":61432,"visible":true,"origin":"","legend":"","description":"","filename":"Table2.pptx","url":"https://assets-eu.researchsquare.com/files/rs-6745848/v1/c0fbdbdd3af170dbd8b527ec.pptx"},{"id":84206310,"identity":"c12655d5-1e94-4de3-a3da-136be9ba2348","added_by":"auto","created_at":"2025-06-09 09:10:30","extension":"pptx","order_by":3,"title":"","display":"","copyAsset":false,"role":"supplement","size":54395,"visible":true,"origin":"","legend":"","description":"","filename":"Table3.pptx","url":"https://assets-eu.researchsquare.com/files/rs-6745848/v1/28860c2d453c275b5558546b.pptx"},{"id":84205264,"identity":"ccb1b81f-d7e1-47e1-971c-01ecb909f54e","added_by":"auto","created_at":"2025-06-09 09:02:31","extension":"pptx","order_by":4,"title":"","display":"","copyAsset":false,"role":"supplement","size":45175,"visible":true,"origin":"","legend":"","description":"","filename":"Table4.pptx","url":"https://assets-eu.researchsquare.com/files/rs-6745848/v1/25ac28f459bc804d9c16bc85.pptx"},{"id":84206315,"identity":"01999184-14e6-4de7-b088-c6acfdd6c5dc","added_by":"auto","created_at":"2025-06-09 09:10:31","extension":"pdf","order_by":12,"title":"","display":"","copyAsset":false,"role":"supplement","size":284026,"visible":true,"origin":"","legend":"","description":"","filename":"Burnhametal.SupplementaryMaterial.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6745848/v1/8d4457161d61b8110b32cb63.pdf"}],"financialInterests":"","formattedTitle":"The morphology of macroalgal substrates can help predict the attachment of juvenile mussels","fulltext":[{"header":"Introduction","content":"\u003cp\u003eIn many parts of the world, vast areas of the seafloor that once supported thriving shellfish reefs have been dramatically altered by overharvesting, destructive fishing practices that removed the benthos, and habitat degradation from coastal development that has intensified over the last century (Lotze et al. \u003cspan citationid=\"CR65\" class=\"CitationRef\"\u003e2006\u003c/span\u003e; Beck et al. \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2009\u003c/span\u003e; Ford and Hamer \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). The removal of this food source and structurally complex biogenic habitat can result in not only a loss of biodiversity in the immediate area (McLeod et al. \u003cspan citationid=\"CR75\" class=\"CitationRef\"\u003e2014\u003c/span\u003e; Sea et al. \u003cspan citationid=\"CR92\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Benjamin 2022a) but also loss of the valuable ecosystem services that these reefs provide, such as water filtration (Suplicy \u003cspan citationid=\"CR96\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; van der Schatte Olivier et al. \u003cspan citationid=\"CR104\" class=\"CitationRef\"\u003e2020\u003c/span\u003e), nitrogen removal (Hillman et al. 2021; Sea et al. 2021), wave energy dissipation (Donker et al. \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2013\u003c/span\u003e), and soft-sediment stabilization (Meadows et al. \u003cspan citationid=\"CR77\" class=\"CitationRef\"\u003e1998\u003c/span\u003e; Brumbaugh et al. \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2006\u003c/span\u003e). Historically in New Zealand or Aotearoa (Māori), reefs of \u003cem\u003ePerna canaliculus\u003c/em\u003e, commonly known as green-lipped mussels or kūtai (Māori), were a common and highly valued feature of coastlines, stretching from the intertidal down to a depth of about 30 m (Paul \u003cspan citationid=\"CR85\" class=\"CitationRef\"\u003e2012\u003c/span\u003e). The Hauraki Gulf or Tīkapa Moana / Te Moananui-ā-Toi (Māori), a large coastal embayment bordering the city of Auckland or Tāmaki Makaurau (Māori), once supported a 1,300 km\u003csup\u003e2\u003c/sup\u003e area of \u003cem\u003eP. canaliculus\u003c/em\u003e reefs (Greenway \u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e1969\u003c/span\u003e; Reid \u003cspan citationid=\"CR88\" class=\"CitationRef\"\u003e1969\u003c/span\u003e). However, these mussel reefs, and others around the country, were decimated by an intensive benthic dredge fishery that had collapsed by 1969 (Paul \u003cspan citationid=\"CR85\" class=\"CitationRef\"\u003e2012\u003c/span\u003e), leaving only a tiny portion of primarily intertidal, remnant mussel reefs (McLeod \u003cspan citationid=\"CR73\" class=\"CitationRef\"\u003e2009\u003c/span\u003e; Toone et al. \u003cspan citationid=\"CR99\" class=\"CitationRef\"\u003e2023a\u003c/span\u003e). Thirty years later, areas of the Hauraki Gulf seafloor that were once dominated by mussels were predominantly characterised by layers of sandy or calcareous mud (Manighetti and Carter \u003cspan citationid=\"CR67\" class=\"CitationRef\"\u003e1999\u003c/span\u003e), and decades of increased sedimentation from poor management of land-use have sustained these conditions into modern times (Morrison et al. 2012). These fine sediments are highly prone to resuspension during storm events (Hauraki Gulf Forum \u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e2023\u003c/span\u003e), and low light penetration from higher turbidity and a lack of hard attachment surfaces contribute to a limited presence and diversity of macroalgae that would have been associated with mussel reefs in the past (Rowden et al. \u003cspan citationid=\"CR90\" class=\"CitationRef\"\u003e2012\u003c/span\u003e; Lao \u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e2016\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eLarge-scale shellfish reef restoration efforts that return adult shellfish, often raised in aquaculture settings, to natural habitats in their historical ranges show significant promise for the future recovery of these ecosystems and their functional value across the globe (McCay et al. \u003cspan citationid=\"CR71\" class=\"CitationRef\"\u003e2003\u003c/span\u003e; Carranza and Zu Ermgassen \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Fitzsimons et al. \u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e2020\u003c/span\u003e), with transplanted adults demonstrating that long-term survival is possible even in degraded environments (Benjamin et al. \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; McAfee et al. \u003cspan citationid=\"CR70\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Overton et al. \u003cspan citationid=\"CR84\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Adult \u003cem\u003eP. canaliculus\u003c/em\u003e transplanted into soft-sediment intertidal and subtidal habitats show survival at least one to two years post-deployment even under adverse modern-day conditions in the Hauraki Gulf (McLeod et al. \u003cspan citationid=\"CR74\" class=\"CitationRef\"\u003e2012\u003c/span\u003e; Wilcox et al. \u003cspan citationid=\"CR108\" class=\"CitationRef\"\u003e2018\u003c/span\u003e) and elsewhere around New Zealand (Benjamin et al. \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Toone et al. \u003cspan citationid=\"CR100\" class=\"CitationRef\"\u003e2023b\u003c/span\u003e). However, recruitment, or the settlement of larvae and establishment of juveniles at a location, is severely limited within these restored adult \u003cem\u003eP. canaliculus\u003c/em\u003e populations (Wilcox et al. \u003cspan citationid=\"CR108\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Benjamin et al. \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2022b\u003c/span\u003e, \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Toone et al. \u003cspan citationid=\"CR99\" class=\"CitationRef\"\u003e2023a\u003c/span\u003e, \u003cspan citationid=\"CR100\" class=\"CitationRef\"\u003eb\u003c/span\u003e) and within other areas of the world that supported wild mussel populations prior to their decimation by overharvesting (Beukema and Cad\u0026eacute;e \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e1996\u003c/span\u003e; Eriksson et al. \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e2010\u003c/span\u003e), indicating that early juvenile mussels may be negatively impacted, either directly or indirectly, by the degraded conditions (Eriksson et al. \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e2010\u003c/span\u003e; McLeod et al. \u003cspan citationid=\"CR74\" class=\"CitationRef\"\u003e2012\u003c/span\u003e; Alder et al. \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Toone et al. \u003cspan citationid=\"CR99\" class=\"CitationRef\"\u003e2023a\u003c/span\u003e). It has been hypothesized that a lack of suitable settlement substrates in these highly sedimented environments is the primary cause of the recruitment bottleneck (McLeod et al. \u003cspan citationid=\"CR74\" class=\"CitationRef\"\u003e2012\u003c/span\u003e; van der Heide et al. \u003cspan citationid=\"CR103\" class=\"CitationRef\"\u003e2014\u003c/span\u003e; Wilcox et al. \u003cspan citationid=\"CR108\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Banke et al. 2025). Across many regions worldwide, the degradation of coastal habitats following the decline of shellfish populations has made benthic substrate limitation one of the most common issues affecting shellfish restoration today, often placing it at the forefront of restoration priorities (Beck at al. 2009; Brumbaugh and Coen \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2009\u003c/span\u003e; Fitzsimons et al. \u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). A challenge in identifying suitable substrates for shellfish reef restoration is the scarcity of healthy reference ecosystems and baseline environmental data available to restoration practitioners that details the type of benthos and associated conditions that historically supported shellfish populations prior to their depletion (Gann et al. \u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; McAfee and Connell \u003cspan citationid=\"CR69\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). As a result, in many locations around the world, it remains unclear which substrates were crucial for the functioning of recruitment processes on mussel reefs that may now be absent from restoration sites.\u003c/p\u003e \u003cp\u003eOntogenic changes in the substrate preferences of mussels adds another layer of complexity to overcoming the barrier of substrate limitation at mussel reef restoration sites. The larvae of many species of mussels, including \u003cem\u003eP. canaliculus\u003c/em\u003e, are known to initially settle onto filamentous substrates, such as hydroids, seagrasses, and macroalgae (Bayne \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e1964\u003c/span\u003e; Seed and Suchanek \u003cspan citationid=\"CR93\" class=\"CitationRef\"\u003e1992\u003c/span\u003e; Buchanan and Babcock \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e1997\u003c/span\u003e). Subsequently, the juvenile mussels relocate, by means of pedal crawling and mucous-drifting (Bayne \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e1964\u003c/span\u003e; Buchanan and Babcock \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e1997\u003c/span\u003e), onto hard substrates located among adult mussel populations, attaching preferentially to conspecifics, and also onto shell and rock (Commito et al. \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2014\u003c/span\u003e; Wilcox and Jeffs \u003cspan citationid=\"CR107\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; Burnham et al. \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). Therefore, while the addition of adult mussels and shell material at restoration sites may be enough to support the establishment of late juvenile and mature mussels, it is likely that facilitating the recruitment of early juvenile mussels requires the addition of filamentous substrates. This can be achieved by provisioning suitable macroalgal substrates at mussel reef restoration sites or by selecting new restoration sites where suitable macroalgal substrates are already present (Fitzsimons et al. \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Toone et al. \u003cspan citationid=\"CR101\" class=\"CitationRef\"\u003e2023c\u003c/span\u003e, \u003cspan citationid=\"CR102\" class=\"CitationRef\"\u003ed\u003c/span\u003e). However, restoration practitioners must firstly determine which species of macroalgae are associated with high levels of recruitment in their target mussel species and whether the suitability of a macroalgal substrate is determined by its species-specific chemical composition or its physical traits. For \u003cem\u003eP. canaliculus\u003c/em\u003e juveniles, field research has revealed greater settlement onto red macroalgae species (i.e., phylum Rhodophyta) that exhibit a high degree of branching along their thalli and thinner branch widths (Buchanan and Babcock \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e1997\u003c/span\u003e; Alfaro and Jeffs \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2002\u003c/span\u003e; Alfaro et al. \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2004\u003c/span\u003e). However, the degree to which recruitment of \u003cem\u003eP. canaliculus\u003c/em\u003e is dependent on these types of macroalgae is unclear, especially in the Hauraki Gulf, where there has been little field research on this topic due to the near-complete loss of wild mussel reefs. Identifying the species and morphological characteristics of macroalgae that most consistently support the highest attachment of juvenile \u003cem\u003eP. canaliculus\u003c/em\u003e across different remnant mussel reefs in the Hauraki Gulf would help to clarify which of these aspects of macroalgae plays a more important role in the recruitment of \u003cem\u003eP. canaliculus\u003c/em\u003e.\u003c/p\u003e \u003cp\u003eWhile the morphology of the macroalgal substrates appears to be an important determinant of juvenile \u003cem\u003eP. canaliculus\u003c/em\u003e attachment (Alfaro and Jeffs \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2002\u003c/span\u003e; Alfaro et al. \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2004\u003c/span\u003e; Loucks \u003cspan citationid=\"CR66\" class=\"CitationRef\"\u003e2023\u003c/span\u003e), the unique chemical signatures of different macroalgae species have also been implicated in promoting mussel attachment (Alfaro et al. \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2006\u003c/span\u003e; Gribben et al. \u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e2011\u003c/span\u003e). For example, the isolated chemical extracts of species of red and brown macroalgae (i.e., phylum Phaeophyta) have been shown to induce settlement in \u003cem\u003eP. canaliculus\u003c/em\u003e (Kelly \u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e2001\u003c/span\u003e; Alfaro et al. \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2006\u003c/span\u003e; Gribben et al. \u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e2011\u003c/span\u003e). However, when these algal extracts were applied to substrates with varying surface textures in a laboratory study, substrates with rough surfaces experienced significantly higher settlement of \u003cem\u003eP. canaliculus\u003c/em\u003e than substrates with smooth surfaces regardless of the chemical cue (Gribben et al. \u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e2011\u003c/span\u003e), suggesting that the substrate\u0026rsquo;s physical structure, rather than its chemical signature, plays a dominant role in juvenile mussel attachment. Furthermore, while the chemical cues of different species of macroalgae may attract varying degrees of larval mussel settlement (Dobretsov \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e1999\u003c/span\u003e; Dobretsov and Wahl \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2001\u003c/span\u003e, \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2008\u003c/span\u003e), it is the morphology of the substrates that can subsequently determine the post-settlement retention and survival of juvenile mussels on reefs (Moreno \u003cspan citationid=\"CR80\" class=\"CitationRef\"\u003e1995\u003c/span\u003e; Frandsen and Dolmer \u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e2002\u003c/span\u003e; Brenner and Buck \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2010\u003c/span\u003e). Laboratory experiments that use plastic analogues of macroalgae to isolate the role of substrate morphology in mussel recruitment reveal that certain branching characteristics can influence the size, settlement densities, and small-scale spatial distribution of juvenile \u003cem\u003eP. canaliculus\u003c/em\u003e (Kelly \u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e2001\u003c/span\u003e; Alfaro and Jeffs \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2002\u003c/span\u003e; Wu \u003cspan citationid=\"CR109\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Wu et al. \u003cspan citationid=\"CR111\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). However, field studies that characterise mussel settlement onto macroalgal substrates often group macroalgae species into functional categories under certain morphological terminology, such as \u0026ldquo;filamentous\u0026rdquo; (Bayne \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e1964\u003c/span\u003e; Hunt and Scheibling \u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e1995\u003c/span\u003e; Yang et al. \u003cspan citationid=\"CR112\" class=\"CitationRef\"\u003e2007\u003c/span\u003e), \u0026ldquo;fine branching/finely branched\u0026rdquo; (Buchanan and Babcock \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e1997\u003c/span\u003e; Alfaro and Jeffs \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2002\u003c/span\u003e; Alfaro at al. 2004), or \u0026ldquo;densely branching\u0026rdquo; (Loucks \u003cspan citationid=\"CR66\" class=\"CitationRef\"\u003e2023\u003c/span\u003e), for which the specific physical characteristics that define those categories are not always clearly described. Additionally, the filamentous or finely branching macroalgae species highlighted in these studies often simultaneously exhibit narrow branch widths, a high frequency of branching, and a short distance between branches, making it difficult to isolate which of these individual characteristics promotes mussel attachment (Wu \u003cspan citationid=\"CR109\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). Identifying which morphological features of macroalgae impact mussel recruitment and defining these features under a single set of unifying descriptors and objective measurements is necessary to help guide the selection of effective substrates for future mussel reef restoration efforts.\u003c/p\u003e \u003cp\u003eThus, the macroalgal substrate associations of juvenile \u003cem\u003eP. canaliculus\u003c/em\u003e at different stages of development were characterised on two remnant, intertidal mussel reefs in the Hauraki Gulf. This study aimed to; 1) create an index that scores the morphological features of macroalgae for their potential to support juvenile mussel recruitment based on prior published data, 2) compare the relative performance of various macroalgae species with differing morphology as attachment substrates for three size classes of juvenile \u003cem\u003eP. canaliculus\u003c/em\u003e on two remnant mussel reefs, and 3) use these findings to evaluate the accuracy with which the index can predict the presence of attached juvenile mussels among macroalgal substrates based on the morphology of the macroalgae. Additionally, to aid an understanding of why certain morphological characteristics of macroalgae support the presence of juvenile \u003cem\u003eP. canaliculus\u003c/em\u003e at different ontogenic stages, the location of attachment within each macroalga species was also characterised for each mussel size class. The goal of this study was to develop and test a tool that has the potential to aid restoration practitioners in the selection of suitable natural substrates or the informed design of artificial substrates for overcoming substrate limitation at mussel reef restoration sites in New Zealand, with broader applications for mussel reef restoration efforts in other locations around the globe.\u003c/p\u003e"},{"header":"Materials and methods","content":"\u003cp\u003eMacroalgal Morphology Index\u003c/p\u003e \u003cp\u003eA comprehensive search of the literature was conducted using the citation database Scopus (Elsevier \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e2025\u003c/span\u003e) to identify previous published studies that analysed or discussed impacts of specific structural properties of natural or artificial substrates on the recruitment, settlement, attachment, growth, or survival of any mussel species. The following keywords were used in the Scopus search string: (\u0026ldquo;mussel\u0026rdquo; OR \u0026ldquo;mussels\u0026rdquo;) AND (\u0026ldquo;attachment\u0026rdquo; OR \u0026ldquo;recruitment\u0026rdquo; OR \u0026ldquo;settlement\u0026rdquo; OR \u0026ldquo;survival\u0026rdquo; OR \u0026ldquo;growth\u0026rdquo;) AND (\u0026ldquo;substrate\u0026rdquo; OR \u0026ldquo;macroalgae\u0026rdquo; OR \u0026ldquo;algae\u0026rdquo; OR \u0026ldquo;macrophyte\u0026rdquo;) AND (\u0026ldquo;morphology\u0026rdquo; OR \u0026ldquo;structure\u0026rdquo; OR \u0026ldquo;features\u0026rdquo; OR \u0026ldquo;characteristics\u0026rdquo; OR \u0026ldquo;properties\u0026rdquo; OR \u0026ldquo;complexity\u0026rdquo; OR \u0026ldquo;heterogeneity\u0026rdquo; OR \u0026ldquo;cover\u0026rdquo;). Studies that only tested the structural complexity of live adult mussel substrate, such as mussel patch size and distribution, were excluded from the search because of difficulty isolating the effects of these structural features from the positive effect of conspecific presence on mussel attachment (Commito et al. \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2014\u003c/span\u003e; Wilcox and Jeffs \u003cspan citationid=\"CR107\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). The studies identified from the literature search were screened to identify the physical features of the substrates to which mussels attached. Those studies that discussed a similar physical feature of the attachment substrate, regardless of the variation in the terminology used to describe the feature, were grouped together under the morphological equivalent of that feature on macroalgae (e.g., rope filament thickness was categorised under branch width; Table\u0026nbsp;1). Each of these overlying morphological features identified from the literature was considered a potential determinant of juvenile mussel attachment, and these were the features used to create the Macroalgal Morphology Index (MMI).\u003c/p\u003e \u003cp\u003eWithin the MMI, physical variations of each morphological feature of macroalgae were sorted into three levels and assigned a score for their reported ability to support juvenile mussel recruitment, in which a score of one is considered not advantageous, two is moderately advantageous, and three is highly advantageous. Corresponding qualitative criteria were generated for these scores based on the physical descriptions of substrates in the literature. Quantitative criteria were also generated from measurements of each morphological feature collected from at least two individuals of each species of macroalga at each of the two sites sampled in this study (see Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e for a diagram the specific measurements collected for each morphological feature). For each macroalga species, measurements were averaged to produce one value for each morphological feature, and the mean measurements for each feature were compared among species and cross-referenced with qualitative criteria to generate quantitative ranges for each of the three scored levels of the MMI. For each species of macroalga, scores for each feature were totaled to obtain a single, overall score representing the potential that each macroalgal substrate has to support juvenile mussel recruitment.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eSampling sites\u003c/p\u003e \u003cp\u003eReef sampling was conducted at two sites on two of the last known remaining intertidal \u003cem\u003eP. canaliculus\u003c/em\u003e reefs in the northern Hauraki Gulf on the North Island of New Zealand. The first site was at Waipū Cove, which is characterised by low levels of sedimentation, moderate levels of human harvesting pressure, and high wave exposure, although ample shelter from wave action can be found among the three-dimensionally complex rock formations. The main mussel reef at this site is approximately 68 m wide by 208 m alongshore, and the majority of adult mussel aggregations extend from approximately 1.5 m above the tide chart datum to 1.5 m below it. The second site was at Pākiri Beach, which is characterised by low levels of sedimentation, protection from harvesting since October 2023, and high wave exposure with minimal shelter from wave action along relatively flat shelves of bedrock. The main mussel reef at this site is approximately 25 m wide by 93 m alongshore, and the majority of adult mussel aggregations extend from approximately 1.5 m above the tide chart datum to 0 m. Both reefs are situated on sandstone bedrock emerging from coarse sand and shell hash, and both sites have limited, adjacent but not contiguous subtidal rocky habitat consisting of emergent ledges populated with macroalgae and some aggregations of mussels.\u003c/p\u003e \u003cp\u003eMussel-macroalgae associations\u003c/p\u003e \u003cp\u003eBetween May 2023 and June 2024, sampling was conducted in the intertidal during low tide at each site, in which a series of 13 m alongshore transects were laid across the width of the mussel reef to ensure that a range of available macroalgae substrates were sampled for attached juvenile \u003cem\u003eP. canaliculus\u003c/em\u003e (\u0026lt;\u0026thinsp;30 mm in shell length; Alfaro et al. \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2001\u003c/span\u003e). At 0.5 m intervals along each transect, 10 \u0026times; 10 cm quadrats were placed over the substrate to standardise the area of macroalgae searched for cryptic juvenile mussels (see methods in Erlandsson and McQuaid \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e2004\u003c/span\u003e; Burnham et al. \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2025\u003c/span\u003e), resulting in 26 quadrats per transect. Thirteen transects were sampled at Waipū Cove from May to July 2023 and February to June 2024, while only eight transects were sampled at Pākiri Beach from October 2023 to February 2024 because it contains a greater diversity of macroalgae concentrated on a smaller area of reef.\u003c/p\u003e \u003cp\u003eThe presence of all macroalgal substrates within each quadrat were recorded, except for epiphytic macroalgae growing on mussels or other algae, encrusting algae (e.g., crustose coralline algae), and microalgae films. Macroalgae species were identified according to a reference text (Nelson \u003cspan citationid=\"CR82\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). Only macroalgae species that were observed in at least five quadrats per site were included in the analysis. Each macroalga species with a holdfast located within the quadrat was searched along the entire length of its thallus for any \u003cem\u003eP. canaliculus\u003c/em\u003e juveniles attached to it by byssal threads. Once a juvenile mussel was located, its shell length (SL) was measured using callipers. For each macroalga species within a quadrat, the presence or absence of juvenile mussels within three size classes was recorded, i.e., \u0026lt;\u0026thinsp;10 mm, 10 \u0026ndash; \u0026lt;20 mm, and 20 \u0026ndash; \u0026lt;30 mm SL. For each macroalga species in a quadrat, once one mussel in a size class was observed, the presence of mussels in that size class was recorded, and any subsequently observed mussels in the same size class were not recorded. The search for mussels attached to each macroalga species continued until all juvenile size classes were found to be present or until all the thalli of each macroalga species within the quadrat had been fully searched. The presence or absence of juvenile mussels in each size class was used in place of mussel counts because of the high densities of juveniles, often in the same size class, in some quadrats (e.g., up to 180 individuals per 0.01 m\u003csup\u003e2\u003c/sup\u003e in the lower intertidal at Waipū Cove), making it impossible to count and measure all of the mussels in situ given tidal time constraints. Destructive sampling was not an option either, as it would have counteracted conservation measures at these locations. A presence/absence measure also ensures that natural variations in the density of juvenile mussels among different areas of the intertidal, where certain species of macroalga are more common than others, were not misinterpreted as higher preference for certain species of macroalga.\u003c/p\u003e \u003cp\u003eLocation of mussel attachment on macroalgae\u003c/p\u003e \u003cp\u003eDuring the last two of the eight transects sampled at Pākiri Beach during February 2024 and the last four of the 13 transects sampled at Waipū Cove between February and June of 2024, the specific location of juvenile \u003cem\u003eP. canaliculus\u003c/em\u003e attachment within macroalgal substrates was recorded as an additional variable to contribute to a comparison of the relative importance of certain morphological features of macroalgae to different developmental stages of juvenile mussels. For each macroalga species within a quadrat, six possible locations were examined for the presence or absence of attached juvenile mussels in each of the three size classes. The six locations within each macroalga were; the holdfast, stipe, lower branch nodes, lower branches, upper branch nodes, and upper branches (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). A node was defined as the 1 cm\u003csup\u003e2\u003c/sup\u003e area around the point of connection between branches or between a stipe and branch (Alfaro and Jeffs \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2002\u003c/span\u003e; Wu \u003cspan citationid=\"CR109\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). Holdfasts, stipes, and lower and upper branches were defined as 1 cm\u003csup\u003e2\u003c/sup\u003e areas without branch nodes. Lower branches and nodes were considered those that were located below half of the canopy height, while upper branches and nodes were located in the upper half of the canopy height. If the morphology of a macroalga did not have some of these locations, then only the locations it did have were recorded and examined for the presence of attached juvenile mussels. If a macroalga had blades, the blades were categorised as branches, and areas of the thallus with blade proliferation were categorised as nodes.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eStatistical analyses\u003c/p\u003e \u003cp\u003eTo test the validity of the MMI, a linear discriminant analysis (LDA; Zhao et al. \u003cspan citationid=\"CR113\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) was applied to juvenile mussel presence data separately for each mussel size class at each site to assess whether scores assigned by the MMI to different morphological characteristics of macroalgae can accurately predict the presence of attached juvenile mussels. An LDA model was chosen because it assumes a linear relationship between the predictors and outcome, and it is less sensitive to small group sizes compared to other discriminant analyses. The predictor variables assessed in the model were the eight morphological features of macroalgae listed in the MMI. The outcome variable used for this analysis was juvenile mussel presence converted from proportional data to binary categorical data (i.e., \u0026ldquo;high presence\u0026rdquo; or \u0026ldquo;low presence\u0026rdquo;) based on the median for juvenile presence in each size class at each site, allowing relative differences in presence to be assessed without the interference of broader trends in juvenile presence (e.g., lower overall presence of larger juvenile size classes at both sites and higher overall presence of juvenile mussels at the Pākiri Beach site). The LDA was implemented using the lda function in the MASS package in R version 3.5.2 (R Core Team 2023). Prior to running the LDA, the predictors were tested for covariance and multicollinearity. Covariance values among predictors were low to moderate (\u0026le;\u0026thinsp;0.72), and a test for multicollinearity revealed moderate values on average (0.5) for predictor pairs. However, there were a few predictor pairs that showed moderately high values for multicollinearity (0.77 maximum). Many of the morphological features of macroalgae are inherently correlated. For example, thin branch widths are often found among macroalgae species that also have a high degree of branching and smaller branch spacing (Wu \u003cspan citationid=\"CR109\" class=\"CitationRef\"\u003e2018\u003c/span\u003e), which would result in high scores for each of these predictors. Given that the purpose of the MMI was to assign similar scores to morphological characteristics with a similar capacity to support juvenile mussel recruitment, correlation\u0026thinsp;\u0026lt;\u0026thinsp;0.8 between predictors was deemed acceptable, and the model was interpreted with the prevailing assumption of potential collinearity.\u003c/p\u003e \u003cp\u003eTo assess the performance of the LDA and avoid overfitting, a 10-fold cross-validation was employed. The cross-validation partitions the data into 10 subsets, using nine for training and one for testing iteratively, providing a reliable estimate of model accuracy. The model\u0026rsquo;s performance is summarized using metrics such as accuracy (i.e., the proportion of correctly predicted outcomes) and Cohen\u0026rsquo;s Kappa (i.e., the level of agreement between observed outcomes and outcomes expected by chance). Afterwards, the relative importance of each morphological feature of macroalgae as a predictor of juvenile mussel presence was quantified using the varImp function in the caret package in R, which evaluates the relative contribution of each predictor based on its influence on the outcome variable. This analysis was useful for determining any morphological features of macroalgae with a low influence on the outcome of juvenile mussel presence that should be excluded from the MMI.\u003c/p\u003e \u003cp\u003eTo evaluate the relative performance of different macroalgae species of varying morphology as mussel attachment substrates, a 95% confidence interval was calculated for the presence of juvenile \u003cem\u003eP. canaliculus\u003c/em\u003e in each size class on each species of macroalga at each site using the binom.test function in R. For each mussel size class, the results of this analysis were used to classify each macroalga species as either a low, moderate, or high-performing attachment substrate for juvenile \u003cem\u003eP. canaliculus\u003c/em\u003e, depending on its relation to the interval. To standardise presence against sampling effort, the presence of juvenile mussels on macroalgal substrates is reported as the proportion (%) of each species of macroalga observed during the reef sampling that exhibited juvenile mussel attachments.\u003c/p\u003e \u003cp\u003eTo evaluate how the location of mussel attachment within substrates varied with mussel size class at each site, the presence/absence of attached juvenile \u003cem\u003eP. canaliculus\u003c/em\u003e among macroalgae species was compared using a generalised linear mixed effects model with the binomial family function and logit link, used specifically for data with quasi-binomial distributions, in R. The effect of the factors of site, macroalga species, mussel size class, location of mussel attachment within macroalgae, and their interaction on juvenile mussel presence were assessed with quadrat incorporated as a random effect. Where the interaction of these factors was not significant, the interaction was removed and the model was re-run to assess the significance of the main effects alone using the greater statistical power gained from conserving degrees of freedom. Where the interaction of these factors was significant, post hoc comparisons were performed using Tukey tests to isolate differences in pairs of means through the emmeans function in R. Analysing the data binomially (i.e., presence as 1 and absence as 0) allowed the model to account for differences in sample size among macroalgal substrates. Due to natural variations in the abundance of mussels in each size class, statistical comparisons were not made between size classes but rather, between substrates within size classes.\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003eDesign of the Macroalgal Morphology Index\u003c/p\u003e \u003cp\u003eFifty-nine research publications covering 12 species of mussel from 33 locations around the world were identified from the literature search that tested or discussed impacts of the specific structural properties of natural or artificial substrates on mussel recruitment, settlement, attachment, growth, or survival (see Table\u0026nbsp;1 for these publications and their relevant findings). Forty-five publications tested the effects of the physical structure of the substrate on mussels, while 14 publications implicated substrate morphology as the cause of their findings. From these publications, eight morphological features of macroalgal substrates were identified as potential determinants of juvenile mussel attachment for the MMI; canopy cover, canopy height, holdfast complexity, clustering frequency, planes of branching, branch spacing, degree of branching, and branch width (Table\u0026nbsp;2; Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). Substrate surface texture was also identified as a structural feature that can impact mussel attachment in eight of the 59 publications (Bourget et al. \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e1994\u003c/span\u003e; Gribben et al. \u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e2011\u003c/span\u003e; Carl et al. \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e2012a\u003c/span\u003e; Wu \u003cspan citationid=\"CR109\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Loucks \u003cspan citationid=\"CR66\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Wu and Jeffs \u003cspan citationid=\"CR110\" class=\"CitationRef\"\u003e2025\u003c/span\u003e; Wu et al. \u003cspan citationid=\"CR111\" class=\"CitationRef\"\u003e2025\u003c/span\u003e; Lanham et al. \u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). However, this feature was not identified in Table\u0026nbsp;1 or included in the MMI because characterising the surface texture of macroalgae is difficult to do \u003cem\u003ein situ\u003c/em\u003e and requires expensive techniques, such as scanning electron microscopy, that would be less accessible for restoration practitioners, for whom the MMI is intended. Nonetheless, these studies were included in Table\u0026nbsp;1 because there were instances where the descriptions of substrate surface texture were applicable to other morphological features, such as the size of millimetre-scale surface indents on artificial substrates being comparable to the space between branches on macroalgae (Bourget et al. \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e1994\u003c/span\u003e; Gribben et al. \u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e2011\u003c/span\u003e; Carl et al. \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e2012a\u003c/span\u003e; Wu and Jeffs \u003cspan citationid=\"CR110\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). Similarly, greater surface roughness was comparable to higher macroalgal canopy cover in terms of its effect on flow velocities and particle deposition (Lanham et al. \u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e2025\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eFor each morphological feature in the MMI, qualitative criteria for each of the three scores, representing a substrate\u0026rsquo;s ability to support juvenile mussel attachment and survival, were able to be characterised using the descriptions of substrates in the literature (Table\u0026nbsp;2). Likewise, mean values of the morphological measurements collected from each species of macroalga in the field allowed for the generation of quantitative ranges for each score (Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e). MMI criteria were used to objectively score the morphological features of each macroalga species, and these scores were summed to produce a total index score for each species intended to represent the potential that each substrate has to support the recruitment of juvenile mussels (Table\u0026nbsp;3). Since eight morphological features, each subdivided into three scored variations, were identified for the MMI, each macroalga species could have a minimum total index score of eight and a maximum total index score of 24.\u003c/p\u003e \u003cp\u003eMussel-macroalgae associations\u003c/p\u003e \u003cp\u003eAt Waipū Cove, a total of 17 macroalgae species of varying phyla and morphology were observed during reef sampling (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). Only one of these species of macroalgae was green (i.e., phylum Chlorophyta), five were brown, and 11 were red, two of which were coralline red macroalgae (i.e., order Corallinales). The presence of juvenile \u003cem\u003eP. canaliculus\u003c/em\u003e attached to these macroalgal substrates varied among the three size classes of juvenile mussels. Juveniles\u0026thinsp;\u0026lt;\u0026thinsp;10 mm SL were present on all macroalgae species at Waipū Cove except for the red macroalga \u003cem\u003eHymenena variolosa\u003c/em\u003e, although juveniles were present on this species at Pākiri Beach. In contrast, juveniles 10 \u0026ndash; \u0026lt;20 mm SL and 20 \u0026ndash; \u0026lt;30 mm SL were present on only eight species of macroalgae, of which \u003cem\u003eCorallina ferreyrae\u003c/em\u003e (prev. \u003cem\u003eCorallina officinalis\u003c/em\u003e; coralline red), \u003cem\u003eJania sphaeroramosa\u003c/em\u003e (coralline red), \u003cem\u003ePsilophycus alveatus\u003c/em\u003e (prev. \u003cem\u003eGigartina alveata\u003c/em\u003e; red), \u003cem\u003ePterocladiella capillacea\u003c/em\u003e (red), \u003cem\u003eScytothamnus australis\u003c/em\u003e (brown), and \u003cem\u003eHormosira banksii\u003c/em\u003e (brown) exhibited attachments by juveniles in both of these larger size classes. Juveniles were only present on the green macroalga \u003cem\u003eUlva australis\u003c/em\u003e up to 20 mm SL.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eAt Waipū Cove, the macroalgae species that demonstrated the highest presence of attached juvenile mussels\u0026thinsp;\u0026lt;\u0026thinsp;10 mm SL were \u003cem\u003eP. alveatus\u003c/em\u003e (87% presence), \u003cem\u003eS. australis\u003c/em\u003e (59%), \u003cem\u003eU. australis\u003c/em\u003e (53%), and \u003cem\u003eJ. sphaeroramosa\u003c/em\u003e (47%; Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). The high-performing macroalgae species for juvenile mussels 10 \u0026ndash; \u0026lt;20 mm SL were \u003cem\u003eP. alveatus\u003c/em\u003e (45% presence), \u003cem\u003eU. australis\u003c/em\u003e (20%), \u003cem\u003eS. australis\u003c/em\u003e (18%), and \u003cem\u003eH. banksii\u003c/em\u003e (18%). For the largest juvenile mussels 20 \u0026ndash; \u0026lt;30 mm SL, the high-performing substrates were \u003cem\u003eP. alveatus\u003c/em\u003e (30% presence) and \u003cem\u003eS. australis\u003c/em\u003e (24%). Additionally, \u003cem\u003eC. ferreyrae\u003c/em\u003e and \u003cem\u003eH. banksii\u003c/em\u003e (39% presence each) supported a moderate presence of attached juveniles\u0026thinsp;\u0026lt;\u0026thinsp;10 mm SL, while \u003cem\u003eEcklonia radiata\u003c/em\u003e (brown), \u003cem\u003eC. ferreyrae\u003c/em\u003e and \u003cem\u003eP. capillacea\u003c/em\u003e (10\u0026ndash;14%) were moderate-performing substrates for juveniles 10 \u0026ndash; \u0026lt;20 mm SL and \u003cem\u003eC. ferreyrae\u003c/em\u003e and \u003cem\u003eP. capillacea\u003c/em\u003e (10\u0026ndash;12%) performed similarly for juveniles 20 \u0026ndash; \u0026lt;30 mm SL.\u003c/p\u003e \u003cp\u003eAt Pākiri Beach, a total of 16 macroalgae species of varying phyla and morphology were observed during reef sampling (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). Two of these species of macroalgae were green, four were brown, and 10 were red, two of which were coralline red macroalgae. Similar to Waipū Cove, the presence of attached juvenile mussels varied among mussel size classes. Juveniles\u0026thinsp;\u0026lt;\u0026thinsp;10 mm SL were present on all macroalgae species at Pākiri Beach. In contrast, juveniles 10 \u0026ndash; \u0026lt;20 mm SL were present on only 13 species of macroalgae, and juveniles 20 \u0026ndash; \u0026lt;30 mm SL were present on only 11 species of macroalgae, of which \u003cem\u003eC. ferreyrae\u003c/em\u003e (coralline red), \u003cem\u003eJ. sphaeroramosa\u003c/em\u003e (coralline red), \u003cem\u003eCaulacanthus ustulatus\u003c/em\u003e (red), \u003cem\u003eGigartina macrocarpa\u003c/em\u003e (red), \u003cem\u003eH. variolosa\u003c/em\u003e (red), \u003cem\u003eCarpophyllum maschalocarpum\u003c/em\u003e (brown), \u003cem\u003eSargassum sinclairii\u003c/em\u003e (brown), and \u003cem\u003eE. radiata\u003c/em\u003e (brown) exhibited attachments by juveniles in both of these larger size classes. Similar to Waipū Cove, juveniles were only present on the two green species of macroalgae, \u003cem\u003eU. australis\u003c/em\u003e and \u003cem\u003eCodium fragile\u003c/em\u003e subsp. \u003cem\u003enovae-zelandiae\u003c/em\u003e, up to 20 mm SL.\u003c/p\u003e \u003cp\u003eAt Pākiri Beach, there were several more species of red macroalgae that qualified as high-performing attachment substrates for juvenile mussels\u0026thinsp;\u0026lt;\u0026thinsp;10 mm SL than for juveniles above this size. The macroalgae species with the highest presence of attached juveniles\u0026thinsp;\u0026lt;\u0026thinsp;10 mm SL were \u003cem\u003eC. maschalocarpum\u003c/em\u003e (100% presence), \u003cem\u003eChondracanthus chapmanii\u003c/em\u003e (100%; red), \u003cem\u003eJ. sphaeroramosa\u003c/em\u003e (96%), \u003cem\u003eC. ferreyrae\u003c/em\u003e (92%), \u003cem\u003eLaurencia thyrsifera\u003c/em\u003e (92%; red), \u003cem\u003eS. sinclairii\u003c/em\u003e (91%), and \u003cem\u003eC. ustulatus\u003c/em\u003e (89%; Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). For juvenile mussels 10 \u0026ndash; \u0026lt;20 mm SL, the high-performing substrates were \u003cem\u003eC. maschalocarpum\u003c/em\u003e (78% presence), \u003cem\u003eC. ferreyrae\u003c/em\u003e (48%), and \u003cem\u003eS. sinclairii\u003c/em\u003e (38%), while \u003cem\u003eC. ferreyrae\u003c/em\u003e (39%) was the only high-performing substrate for juveniles 20 \u0026ndash; \u0026lt;30 mm SL. Additionally, \u003cem\u003eChondria macrocarpa\u003c/em\u003e (red; 80% presence) supported a relatively moderate presence of attached juveniles\u0026thinsp;\u0026lt;\u0026thinsp;10 mm SL, while \u003cem\u003eS. sinclairii\u003c/em\u003e (19%) was a moderate-performing substrate for larger juveniles 20 \u0026ndash; \u0026lt;30 mm SL and juveniles 10 \u0026ndash; \u0026lt;20 mm SL had no moderate-performing substrates.\u003c/p\u003e \u003cp\u003eOverall, the high-performing macroalgal substrates for juvenile mussels in all three size classes at both sites shared at least three of the following high-scoring morphological characteristics described in the MMI: a widespread canopy cover, a more complex holdfast that attached to the substrate at multiple points, a higher clustering frequency (i.e., a higher tendency to grow in close proximity to other conspecifics and colonise a larger area of substrate), multiple planes of branching and/or folding, and thinner branch widths at the nodes (Table\u0026nbsp;3). Shorter branch spacing was also common among the high-performing substrates at Waipū Cove, and a higher degree of branching was common among the high-performing substrates at Pākiri Beach. At both sites, taller canopy heights were not a characteristic of macroalgae that was consistently associated with a higher presence of attached juvenile mussels in all three size classes. Additionally, more of the high-performing macroalgal substrates for smaller juveniles\u0026thinsp;\u0026lt;\u0026thinsp;10 mm SL received high scores for features that describe aspects of branching morphology than the high-performing substrates for larger juveniles\u0026thinsp;\u0026ge;\u0026thinsp;10 mm SL at both sites.\u003c/p\u003e \u003cp\u003eWhile there were 11 species of macroalgae that were observed at both Waipū Cove and Pākiri Beach during reef sampling, the only species of macroalga with consistently moderate or high relative presence of attached juvenile mussels for all size classes at both sites was \u003cem\u003eC. ferreyrae\u003c/em\u003e. Additionally, \u003cem\u003eJ. sphaeroramosa\u003c/em\u003e supported high relative presence of attached juveniles in the smallest size class at both sites but not for juveniles\u0026thinsp;\u0026ge;\u0026thinsp;10 mm SL. Both of these macroalgal substrates are turf-forming, coralline red species that grow throughout rockpools and emergent rock in the intertidal. They both have high-scoring morphological characteristics, such as a widespread canopy cover, a more complex holdfast, a high clustering frequency, very narrow branch widths, a high degree of branching, multiple planes of branching, and shorter branch spacing (Table\u0026nbsp;3). Consequently, these two species have some of the highest total index scores of 20 for \u003cem\u003eC. ferreyrae\u003c/em\u003e and 22 for \u003cem\u003eJ. sphaeroramosa\u003c/em\u003e out of 24 (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eDespite a different assemblage of macroalgae species qualifying as high-performing attachment substrates for juvenile mussels at each site, all high-performing macroalgae species scored moderate to high total index scores according to the MMI. At Pākiri Beach, high-performing macroalgae species for all juvenile mussel size classes had a minimum total index score of 14 and a maximum of 22 out of 24, while the minimum score at Waipū Cove was 16 and the maximum score was also 22 (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). While for many species of macroalgae, moderate to high total index scores were associated with relatively higher presence of attached juvenile mussels, there were a number of other macroalgal substrates with high total index scores that were considered low-performing. For example, \u003cem\u003eGelidium caulacantheum\u003c/em\u003e is a turf-forming, fleshy, red macroalga at Waipū Cove that has a total index score of 21, but it exhibited relatively low presence of attached juvenile mussels (\u0026le;\u0026thinsp;20% presence; Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). Similarly, \u003cem\u003eU. australis\u003c/em\u003e was a high-performing substrate for juvenile mussels\u0026thinsp;\u0026lt;\u0026thinsp;20 mm SL at Waipū Cove (20\u0026ndash;53% presence; Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e), but it has a mix of low and high scoring morphological characteristics, resulting in moderate a total index score of 16 (Table\u0026nbsp;3).\u003c/p\u003e \u003cp\u003eLocation of mussel attachment on macroalgae\u003c/p\u003e \u003cp\u003eDuring the portion of reef sampling that assessed the location of juvenile mussel attachment on macroalgal substrates, ten species of macroalgae were sampled at Waipū Cove and thirteen species of macroalgae were sampled at Pākiri Beach (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e). Only three of these macroalgae species were observed at both sites during this portion of sampling, i.e., \u003cem\u003eC. ferreyrae\u003c/em\u003e, \u003cem\u003eJ. sphaeroramosa\u003c/em\u003e, and \u003cem\u003eDictyota kunthii\u003c/em\u003e (brown). However, the generalised linear mixed effects model revealed that the species of macroalgae had no significant interactive effects on the presence of attached juvenile mussels among different locations within the substrate for any mussel size class at either site. Therefore, this effect was removed and the model was re-run to interpret the main effects only.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eSite, mussel size class, and location within the substrate all had significant interactive effects on attached juvenile mussel presence (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.033). Post hoc comparisons of these factors revealed that while the locations of attachment on macroalgal substrates were fairly consistent for juvenile mussels in the two larger size classes, attachment locations differed between the two study sites for juveniles in the smallest size class. Overall, for larger juveniles\u0026thinsp;\u0026ge;\u0026thinsp;10 mm SL at both sites, the presence of attached mussels was highest on the holdfasts of macroalgal substrates. For the largest juveniles 20 \u0026ndash; \u0026lt;30 mm SL, presence was higher on holdfasts than on any other attachment location within the substrates by at least 11% at Pākiri Beach (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.038) and 21% at Waipū Cove (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.0035). The same was true for mid-sized juveniles 10 \u0026ndash; \u0026lt;20 mm SL at Waipū Cove by at least 17% (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.0072), but at Pākiri Beach, the presence of mid-sized juveniles was higher on both the holdfasts and lower branch nodes of macroalgae than in any other attachment location by 20% and 10%, respectively (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.042). Overall, for the smallest juveniles\u0026thinsp;\u0026lt;\u0026thinsp;10 mm SL at both sites, the presence of attached mussels was highest on both the holdfasts and branch nodes of macroalgae, while attachments to the branches and stipes of macroalgae were less common. For juveniles\u0026thinsp;\u0026lt;\u0026thinsp;10 mm SL at Pākiri Beach, the presence of attached mussels was higher on the holdfasts and lower branch nodes of macroalgae than on any other attachment location by at least 59% and 54%, respectively (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.0001). However, for juveniles\u0026thinsp;\u0026lt;\u0026thinsp;10 mm SL at Waipū Cove, the presence of attached mussels was higher on the upper branch nodes of macroalgae (32% presence), followed by the holdfasts (23%) and lower branch nodes (18%), than on any other attachment location within the substrates (\u0026le;\u0026thinsp;12%). The presence of attached juvenile mussels in this size class was significantly higher on the upper branch nodes compared to the stipe and upper and lower branches by at least 20% (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.0056), and presence was significantly higher on the holdfasts compared to the stipe and upper branches by at least 16% (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.024).\u003c/p\u003e \u003cp\u003eThe post hoc comparisons of mussel attachment locations between study sites revealed that at Pākiri Beach, larger juveniles\u0026thinsp;\u0026ge;\u0026thinsp;10 mm SL were not observed attaching to the lower or upper branches of macroalgae, while larger juveniles at Waipū Cove were not observed attaching to the stipes. For smaller juveniles\u0026thinsp;\u0026lt;\u0026thinsp;10 mm SL, attachments to the upper portion of macroalgal substrates were more common at Waipū Cove than at Pākiri Beach, while attachments to the lower portion of macroalgae were conversely more common at Pākiri Beach. At Waipū Cove, there was a higher presence of juveniles\u0026thinsp;\u0026lt;\u0026thinsp;10 mm SL attached to the upper branches and upper branch nodes of macroalgal substrates than at Pākiri Beach by 6% and 30%, respectively (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.039). Meanwhile, at Pākiri Beach, there was 45% higher presence of juveniles in this size class attached to the holdfasts and lower branch nodes of macroalgae than at Waipū Cove (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.0001). These site-specific differences in the attachment locations of juvenile mussels on macroalgae were maintained across the same species of macroalgae at both sites.\u003c/p\u003e \u003cp\u003eEvaluation of the Macroalgal Morphology Index\u003c/p\u003e \u003cp\u003eAssessment of the relative importance of each morphological feature identified in the MMI, as predictor variables of the observed presence of attached juvenile mussels among macroalgae, revealed that the importance of each feature differed among sites and mussel size classes (Table\u0026nbsp;4). High variability in the relative contribution of each predictor to the observed outcome meant that no one morphological feature was of consistently low enough importance for all mussel size classes and for both sites for it to be eliminated from the MMI. Therefore, all eight morphological features were retained.\u003c/p\u003e \u003cp\u003eAt Waipū Cove, the morphological features of macroalgae that qualified as strong predictors of attached juvenile presence (i.e., \u0026ge;\u0026thinsp;0.65 relative importance value on a scale of 0.5 to 1.0) that were shared among all three mussel size classes were holdfast complexity (0.65\u0026ndash;0.81 relative importance) and canopy cover (0.67\u0026ndash;0.72; Table\u0026nbsp;4). At Pākiri Beach, canopy cover (0.70\u0026ndash;0.75 relative importance) was a strong predictor of the presence of attached juveniles for all three size classes, while holdfast complexity qualified as a strong predictor for all size classes except for juveniles 10 \u0026ndash; \u0026lt;20 mm SL. Overall, at both sites, there were a greater number of morphological features in the model that qualified as strong predictors of attached mussel presence for smaller juveniles\u0026thinsp;\u0026lt;\u0026thinsp;10 mm SL than for larger juveniles\u0026thinsp;\u0026ge;\u0026thinsp;10 mm SL. Collectively, the MMI scores for each morphological feature were 75% accurate effective at predicting the presence of juvenile mussels\u0026thinsp;\u0026lt;\u0026thinsp;10 mm SL among macroalgal substrates at both sites, and agreement between observed and predicted juvenile presence for this size class was moderately strong, as indicated by a Cohen\u0026rsquo;s Kappa of 0.43 at Pākiri Beach and 0.50 at Waipū Cove (Table\u0026nbsp;4). At both sites, three out of the five strong predictors of the presence of attached juveniles\u0026thinsp;\u0026lt;\u0026thinsp;10 mm SL were features of macroalgae that describe aspects of branching morphology. At Waipū Cove, these features were branch spacing (0.73 relative importance), planes of branching (0.69), and branch width (0.65), while at Pākiri Beach, these features were degree of branching (0.73), branch width (0.73), and planes of branching (0.69). In contrast, for larger juveniles\u0026thinsp;\u0026ge;\u0026thinsp;10 mm SL, branching features were generally poor predictors of juvenile presence. At Waipū Cove, there was one branching feature that qualified as a strong predictor of attached mussel presence for juveniles\u0026thinsp;\u0026ge;\u0026thinsp;10 mm SL, but at Pākiri Beach, all branching features were poor predictors for the presence of larger juveniles. At both sites, canopy cover and holdfast complexity were the only morphological features that consistently acted as strong determinants of attached mussel presence for juveniles\u0026thinsp;\u0026ge;\u0026thinsp;10 mm SL. As a result, the MMI scores for each morphological feature were collectively able to predict the presence of attached juveniles\u0026thinsp;\u0026ge;\u0026thinsp;10 mm SL among macroalgal substrates with only 50\u0026ndash;60% accuracy (Kappa: 0.11\u0026ndash;0.33) at Waipū Cove and 40\u0026ndash;45% accuracy (Kappa: -0.13\u0026ndash;0) at Pākiri Beach (Table\u0026nbsp;4).\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eBy creating an index that scores the morphological features of macroalgae for their potential to support recruitment of juvenile mussels, this study has developed a tool for mussel reef restoration practitioners to rank the suitability of macroalgal substrates in an objective and repeatable manner using clearly defined physical descriptions. Furthermore, by testing the accuracy of this index against the presence of \u003cem\u003eP. canaliculus\u003c/em\u003e juveniles attached to macroalgae on two remnant mussel reefs in New Zealand, this study demonstrates that the morphology of a macroalgal substrate can be used to accurately predict the substrate\u0026rsquo;s potential to support juvenile mussel attachments. Eight morphological features of macroalgae were identified from studies on 12 different species of mussel and used to create the Macroalgal Morphology Index (i.e., canopy cover, canopy height, holdfast complexity, clustering frequency, planes of branching, branch spacing, degree of branching, and branch width). Substrates with morphological characteristics that equate to widespread canopy cover, higher holdfast complexity, multiple planes of branching or folding, thinner branch widths at the nodes, and a higher degree of branching on macroalgae are commonly reported to support a higher degree of juvenile attachment in many species of mussel (Bulleri et al. \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2006\u003c/span\u003e; Brenner and Buck \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2010\u003c/span\u003e; Carl et al. \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2012b\u003c/span\u003e; Liu et al. \u003cspan citationid=\"CR64\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; Wu and Jeffs \u003cspan citationid=\"CR110\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). In this study, macroalgae species that exhibited a number of these high-scoring physical characteristics supported some of the highest presence of juvenile \u003cem\u003eP\u003c/em\u003e. \u003cem\u003ecanaliculus\u003c/em\u003e. Collectively, the MMI\u0026rsquo;s scoring criteria was most effective at predicting the likelihood of \u003cem\u003eP. canaliculus\u003c/em\u003e attachments to macroalgae for juveniles smaller than 10 mm SL and less effective for juveniles 10 mm SL and larger that relied less on branching morphology for attachment. These findings indicate that the MMI can be used as a tool to assist mussel reef restoration practitioners in both the evaluation of existing macroalgal habitat at future restoration sites and the selection of suitable macroalgal substrates to be transplanted into current restoration sites for the purpose of facilitating juvenile recruitment among restored mussel populations in New Zealand and likely many other locations across the world.\u003c/p\u003e \u003cp\u003eMussel-macroalgae associations\u003c/p\u003e \u003cp\u003eAt both sites, \u003cem\u003eP. canaliculus\u003c/em\u003e juveniles smaller than 10 mm SL attached to almost all species of macroalgae present on the reefs, while larger juveniles between 10 and 20 mm SL and 20 and 30 mm SL were only present on about half of the macroalgae species at Waipū Cove and roughly three quarters of the species at Pākiri Beach. These findings suggest that, regardless of the species or morphology, macroalgae are critical attachment substrates for juvenile \u003cem\u003eP. canaliculus\u003c/em\u003e during early development, and the type of macroalgae only begins to exclude the presence of juveniles during later development. Similarly, attached juvenile presence appeared to be unrelated to macroalgal phyla until juveniles reached 20 mm SL, after which point they were no longer present on green macroalgae but still demonstrated attachments to numerous species of red and brown macroalgae. Despite a greater availability of red macroalgae species at both sites, only juveniles smaller than 10 mm SL at Pākiri Beach demonstrated a higher presence on more species of red macroalgae than brown macroalgae. While some studies report higher settlement of juvenile \u003cem\u003eP. canaliculus\u003c/em\u003e on a greater number of red macroalgae species than brown macroalgae species at smaller juvenile sizes (\u0026lt;\u0026thinsp;0.5 mm SL; Alfaro and Jeffs \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2002\u003c/span\u003e; Alfaro et al. \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2004\u003c/span\u003e), high settlement onto numerous species of brown and red macroalgae has been reported for juvenile \u003cem\u003eP. canaliculus\u003c/em\u003e as small as ~\u0026thinsp;250 \u0026micro;m and as large as ~\u0026thinsp;7 mm SL (Buchanan and Babcock \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e1997\u003c/span\u003e; Loucks \u003cspan citationid=\"CR66\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Toone et al. \u003cspan citationid=\"CR101\" class=\"CitationRef\"\u003e2023c\u003c/span\u003e). It is likely that differences in the morphology of these macroalgae species are responsible for the variation in juvenile mussel settlement preferences for red or brown macroalgae between studies. Regardless, the results of this study contribute valuable insights into the diversity of macroalgal phyla that \u003cem\u003eP. canaliculus\u003c/em\u003e juveniles continue to associate with beyond 7 mm SL.\u003c/p\u003e \u003cp\u003eThere was a moderate to high presence of attached juvenile mussels on macroalgae species previously reported to support varying degrees of juvenile \u003cem\u003eP. canaliculus\u003c/em\u003e attachment (i.e., red macroalgae: \u003cem\u003eP. alveatus\u003c/em\u003e, \u003cem\u003eJania\u003c/em\u003e spp., \u003cem\u003eL. thyrsifera\u003c/em\u003e, \u003cem\u003eC. ferreyrae\u003c/em\u003e, \u003cem\u003eP. capillacea\u003c/em\u003e; brown macroalgae: \u003cem\u003eC. maschalocarpum\u003c/em\u003e, \u003cem\u003eS. australis\u003c/em\u003e, \u003cem\u003eH. banksii\u003c/em\u003e) but also on macroalgae not previously reported as attachment substrates for \u003cem\u003eP. canaliculus\u003c/em\u003e (red macroalgae: \u003cem\u003eC. chapmanii\u003c/em\u003e, \u003cem\u003eC. ustulatus\u003c/em\u003e, \u003cem\u003eC. macrocarpa\u003c/em\u003e; brown macroalgae: \u003cem\u003eS. sinclairii\u003c/em\u003e, \u003cem\u003eE. radiata\u003c/em\u003e; green macroalgae: \u003cem\u003eU. australis\u003c/em\u003e), expanding the current list of macroalgal substrates with strong potential to facilitate the recruitment of juvenile \u003cem\u003eP. canaliculus\u003c/em\u003e at restoration sites. In particular, the common, coralline red macroalgae \u003cem\u003eC. ferreyrae\u003c/em\u003e and \u003cem\u003eJ. sphaeroramosa\u003c/em\u003e were the only macroalgae species to support a moderate to high presence of juvenile mussels at both sites, in which \u003cem\u003eJ. sphaeroramosa\u003c/em\u003e qualified as a high-performing attachment substrate for juveniles in the smallest size class and \u003cem\u003eC. ferreyrae\u003c/em\u003e qualified as a moderate- or high-performing substrate for juveniles in all size classes. These findings mirror the high settlement densities of larvae and juvenile mussels supported by species of \u003cem\u003eJania\u003c/em\u003e and \u003cem\u003eC. ferreyrae\u003c/em\u003e, especially, throughout numerous laboratory and field studies (Hunt and Scheibling \u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e1995\u003c/span\u003e; Buchanan and Babcock \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e1997\u003c/span\u003e; Alfaro et al. \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2004\u003c/span\u003e; McQuaid and Lindsay \u003cspan citationid=\"CR76\" class=\"CitationRef\"\u003e2005\u003c/span\u003e). Inconsistencies in the performances of the nine other species of macroalgae that were present at both sites are likely the result of differing hydrodynamic conditions between the two sites interacting with macroalgal morphology to positively or negatively influence juvenile mussel attachment. Therefore, these findings indicate that even under variable hydrodynamic conditions, \u003cem\u003eC. ferreyrae\u003c/em\u003e and \u003cem\u003eJ. sphaeroramosa\u003c/em\u003e together have the highest potential to support the attachment of juvenile \u003cem\u003eP. canaliculus\u003c/em\u003e throughout their development.\u003c/p\u003e \u003cp\u003eWhile \u003cem\u003eC. ferreyrae\u003c/em\u003e and \u003cem\u003eJ. sphaeroramosa\u003c/em\u003e can grow in the upper intertidal down to a depth of at least 20 m, other high-performing macroalgae species, such as \u003cem\u003eP. alveatus\u003c/em\u003e, \u003cem\u003eC. chapmanii, C. ustulatus\u003c/em\u003e, \u003cem\u003eS. australis\u003c/em\u003e, \u003cem\u003eU. australis\u003c/em\u003e, \u003cem\u003eH. banksii\u003c/em\u003e, \u003cem\u003eC. maschalocarpum\u003c/em\u003e, and \u003cem\u003eS. sinclairii\u003c/em\u003e, are not typically found below 6 m depth (Nelson \u003cspan citationid=\"CR82\" class=\"CitationRef\"\u003e2020\u003c/span\u003e), suggesting that many of these species are only suitable for intertidal and shallow subtidal mussel reef restoration sites. Furthermore, thin and highly branched red macroalgae, which describe many of the high-performing species in this study, can be vulnerable to the effects of abrasive, resuspended sediment that characterise areas of the contemporary Hauraki Gulf (Lao \u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). Therefore, for deeper or more degraded mussel reef restoration sites, selecting hardier, low-light tolerant species of macroalgae or designing artificial substrates, potentially comprised of natural materials, that exhibit similar morphological characteristics to the high-performing macroalgae species in this study would likely be a more practical and effective choice for future substrate provisioning.\u003c/p\u003e \u003cp\u003eLocation of mussel attachment on macroalgae\u003c/p\u003e \u003cp\u003eUnderstanding which locations within a macroalgal substrate juvenile mussels attach to throughout different stages of ontogeny is critical for understanding why the impact of different morphological features of macroalgae may change with juvenile size. In this study, for \u003cem\u003eP. canaliculus\u003c/em\u003e juveniles smaller than 10 mm SL, attachments to the branches and branch nodes of macroalgae were far more common than for juveniles 10 mm SL and larger, suggesting that attachment by \u003cem\u003eP. canaliculus\u003c/em\u003e juveniles is more likely to be influenced by the branching morphology of macroalgae earlier in juvenile development. Additionally, juveniles were more often attached to the branch nodes of macroalgae than to the inter-node areas of the branches regardless of juvenile size, mirroring previously reported higher and more clustered settlement of juvenile \u003cem\u003eP. canaliculus\u003c/em\u003e on the nodes of branching macroalgae and plastic mimics (Kelly \u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e2001\u003c/span\u003e; Alfaro and Jeffs \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2002\u003c/span\u003e). While juvenile mussels are rarely documented attaching to the holdfasts of macroalgae in previous literature, this was the most common location of attachment in this study for juveniles 10 mm SL and larger, and for juveniles smaller than 10 mm SL, it was as common of an attachment location as the branch nodes of macroalgae. It is likely that mussel attachments to holdfasts are virtually unreported for \u003cem\u003eP. canaliculus\u003c/em\u003e because the majority of macroalgal and artificial attachment substrates analysed for this species have been drifting or beachcast samples or have been experimentally suspended in the water column (Buchanan and Babcock \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e1997\u003c/span\u003e; Alfaro and Jeffs \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2002\u003c/span\u003e; Toone et al. \u003cspan citationid=\"CR101\" class=\"CitationRef\"\u003e2023c\u003c/span\u003e). The results of this study highlight the importance of the holdfast morphology of macroalgae for attachment of all sizes of juvenile \u003cem\u003eP. canaliculus\u003c/em\u003e but especially for juveniles in later development.\u003c/p\u003e \u003cp\u003eFor \u003cem\u003eP. canaliculus\u003c/em\u003e juveniles smaller than 10 mm SL, attachments to the upper branches and nodes of macroalgae were more common at Waipū Cove than at Pākiri Beach, where smaller juveniles were more often attached to the lower portion of macroalgae. Even when the same species of macroalgae was present at both sites (e.g., \u003cem\u003eC. ferreyrae\u003c/em\u003e and \u003cem\u003eJ. sphaeroramosa\u003c/em\u003e), these site-specific differences in the location of juvenile mussel attachment persisted, implicating the interaction of the substrate\u0026rsquo;s morphology with the site\u0026rsquo;s environmental conditions as the cause of these trends. It is likely that higher wave exposure at Pākiri Beach makes attachment to the upper portions of macroalgae less favourable for juvenile mussels than under the more sheltered conditions at Waipū Cove. As juvenile mussels grow, the risk of dislodgement increases as the impact of hydrodynamic forces like frictional drag and shear stress act more strongly on their larger surface areas (Hunt and Scheibling \u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e2001\u003c/span\u003e; Donker et al. \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2013\u003c/span\u003e). An increasing risk of dislodgement as juveniles grow would explain why larger juveniles were primarily attached to the holdfasts, where their position close to the substrate places them in an area of reduced water velocity and greater stability beneath the macroalgal canopy (Eckman and Duggins \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e1991\u003c/span\u003e; McCook and Chapman \u003cspan citationid=\"CR72\" class=\"CitationRef\"\u003e1991\u003c/span\u003e; Westerbom et al. \u003cspan citationid=\"CR106\" class=\"CitationRef\"\u003e2008\u003c/span\u003e). It would also explain why macroalgal substrates with high-scoring branching morphologies supported a high presence of juveniles in all size classes at Waipū Cove but only supported a high presence of juveniles smaller than 10 mm SL at Pākiri Beach, as attachment to branches likely places larger juveniles at greater risk. This finding highlights the need for mussel reef restoration practitioners that are assessing the suitability of macroalgal substrates to pay close attention to the features in the MMI that score the branching and holdfast morphologies of macroalgae and consider whether environmental conditions at the proposed restoration site are likely to support the continued attachment of juvenile mussels to these locations within the macroalgae.\u003c/p\u003e \u003cp\u003eMacroalgal morphology\u003c/p\u003e \u003cp\u003eJuvenile \u003cem\u003eP. canaliculus\u003c/em\u003e at both mussel reefs were attached to macroalgal substrates with a variety of morphologies, regardless of juvenile size. However, there were a number of key morphological characteristics shared between many of the macroalgal substrates that supported a high presence of attached juvenile mussels in all three size classes. These high-performing macroalgal substrates exhibited not all but a combination of many of the following characteristics: a widespread canopy cover, a higher holdfast complexity, a higher clustering frequency, multiple planes of branching or folding, thinner branch widths at the nodes, shorter branch spacing, and a higher degree of branching. The only morphological feature of macroalgae that consistently qualified as a strong predictor of the presence of attached juvenile mussels in all size classes at both sites was canopy cover, and the same was true of holdfast complexity for all but the 10 \u0026ndash; \u0026lt;20 mm SL juvenile size class at Pākiri Beach. This finding suggests that out of all the morphological features described in the MMI, canopy cover and holdfast complexity are likely the most important drivers of the continued attachment of juvenile mussels on macroalgae throughout juvenile ontogeny.\u003c/p\u003e \u003cp\u003eMacroalgae species that received high scores from the MMI for having a widespread canopy cover included both large fucoids (e.g., \u003cem\u003eC. maschalocarpum\u003c/em\u003e, \u003cem\u003eS. sargassum\u003c/em\u003e, \u003cem\u003eS. australis\u003c/em\u003e, \u003cem\u003eH. banksii\u003c/em\u003e, and \u003cem\u003eE. radiata\u003c/em\u003e) and short, turf-forming rhodophytes (e.g., \u003cem\u003eC. ferreyrae\u003c/em\u003e, \u003cem\u003eJ. sphaeroramosa\u003c/em\u003e, \u003cem\u003eC. chapmanii\u003c/em\u003e, and \u003cem\u003eC. ustulatus\u003c/em\u003e) because the wide or continuous canopies exhibited by these species can create similar physical effects on their immediate surroundings that have the potential to benefit juvenile mussels. For example, macroalgal canopies that cover a larger area of substrate can promote the settlement of mussel larvae by reducing water flow as it moves across the canopy, resulting in enhanced particle deposition (Eckman and Duggins \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e1991\u003c/span\u003e; B\u0026eacute;gin et al. \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2004\u003c/span\u003e), and by providing shaded surfaces preferred by settling mussel larvae (Marsden and Lansky 1999; Kobak \u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e2001\u003c/span\u003e; Holthuis et al. \u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e2015\u003c/span\u003e). Greater canopy cover can also enhance the survival of juvenile mussels by providing shade and moisture that protect against desiccation and mediate temperature stress in the intertidal (McCook and Chapman \u003cspan citationid=\"CR72\" class=\"CitationRef\"\u003e1991\u003c/span\u003e; Bulleri et al. \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2006\u003c/span\u003e; de Nesnera \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2016\u003c/span\u003e), and reduced water velocities across areas of greater continuous macroalgal cover have also been hypothesized to reduce the dislodgement of intertidal mussels (McCook and Chapman \u003cspan citationid=\"CR72\" class=\"CitationRef\"\u003e1991\u003c/span\u003e). Therefore, macroalgae with widespread canopy covers have the potential to enhance both the initial attachment and the longer-term retention of juvenile mussels.\u003c/p\u003e \u003cp\u003eAmeliorated temperature extremes and reduced hydrodynamic forces created by broader macroalgal canopies can also be associated with a higher density of predators in their understory, sometimes leading to higher mussel mortality (Eckman and Duggins \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e1991\u003c/span\u003e; Menge \u003cspan citationid=\"CR78\" class=\"CitationRef\"\u003e1978\u003c/span\u003e). However, the three-dimensional complexity of macroalgal holdfasts can further mediate juvenile mussel predation by regulating predator access (Moreno \u003cspan citationid=\"CR80\" class=\"CitationRef\"\u003e1995\u003c/span\u003e). On artificial surfaces with higher three-dimensional complexity, such as corrugated panels and shells embedded in concrete, \u003cem\u003eMytilus edulis\u003c/em\u003e juveniles demonstrated greater survival from predation through the formation of protective aggregations and a higher number of byssal attachments to the substrate that increased their stability (Frandsen and Dolmer \u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e2002\u003c/span\u003e; Christensen et al. \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2015\u003c/span\u003e). In this study, macroalgae that exhibited more complex holdfasts secured to the substrate at multiple points, such as \u003cem\u003eP. alveatus\u003c/em\u003e, \u003cem\u003eH. banksii\u003c/em\u003e, and the majority of the turf-forming red macroalgae, were often observed supporting densely-packed aggregations of juvenile mussels interwoven amongst the holdfasts in a manner that appeared to limit the amount of surface area on their shells that was exposed to predators and hydrodynamic drag. Refuge-seeking behaviour from predators and hydrodynamic forces was implicated in a previous study as the reason that \u003cem\u003eMytilus galloprovincialis\u003c/em\u003e larvae initially settled deep inside of structurally complex culture ropes, only migrating to the rope exteriors as larger juveniles (Carl et al. \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2012b\u003c/span\u003e). These findings suggest that like canopy cover, holdfast complexity can influence both the settlement behaviour of mussel larvae and the post-settlement survival of juvenile mussels, especially considering that in this study, holdfasts were one of the most common attachment locations for \u003cem\u003eP. canaliculus\u003c/em\u003e juveniles smaller than 10 mm SL and the most common area of attachment for juveniles 10 mm SL and larger. Given the commonality in the substrate attachment preferences reported for other species of mussel, these features of macroalgal morphology are likely to have similar effects on other mussel species in other locations. Therefore, mussel reef restoration practitioners evaluating new restoration sites or provisioning existing sites with macroalgal substrates in an attempt to facilitate juvenile recruitment should look for macroalgae with three-dimensionally complex holdfasts and widespread canopy covers to improve the chances of supporting the establishment of juvenile mussels within the reef.\u003c/p\u003e \u003cp\u003eUnlike holdfast complexity and canopy cover, the majority of macroalgal features that describe aspects of branching morphology qualified as strong predictors of the presence of attached juvenile \u003cem\u003eP. canaliculus\u003c/em\u003e in the smallest size class but were poor predictors for the two larger juvenile size classes at both sites. This finding, along with the observation that juveniles smaller than 10 mm SL were attached to the branches and branch nodes of macroalgae much more often than juveniles 10 mm SL and larger, suggests that the branching features of macroalgal morphology are only drivers of the attachment of juvenile mussels in their earlier stages of development. For juveniles smaller than 10 mm SL, branch width and planes of branching qualified as strong predictors of presence among macroalgae at both sites, while degree of branching was a strong predictor only at Waipū Cove and branch spacing was only a strong predictor at Pākiri Beach. In many studies, juvenile mussels smaller than 1 mm SL settled in higher densities on macroalgae characterised primarily by thinner branch widths but also by a higher degree of branching (Buchanan and Babcock \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e1997\u003c/span\u003e; Dobretsov and Wahl \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2001\u003c/span\u003e; Alfaro et al. \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2004\u003c/span\u003e; Yang et al. \u003cspan citationid=\"CR112\" class=\"CitationRef\"\u003e2007\u003c/span\u003e). However, in many of these studies, juveniles began to attach in higher densities to macroalgae with characteristically broader and less numerous branches starting at just 1 or 2 mm SL (Hunt and Scheibling \u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e1995\u003c/span\u003e; Kelly \u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e2001\u003c/span\u003e; Alfaro and Jeffs \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2002\u003c/span\u003e; Alfaro et al. \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2004\u003c/span\u003e), and some studies showed higher settlement onto these types of macroalgae in juveniles below this size (Loucks \u003cspan citationid=\"CR66\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Toone et al. \u003cspan citationid=\"CR101\" class=\"CitationRef\"\u003e2023c\u003c/span\u003e). Therefore, while thinner, highly branched macroalgae species, such as \u003cem\u003eJ. sphaeroramosa\u003c/em\u003e, \u003cem\u003eC. ferreyrae\u003c/em\u003e, \u003cem\u003eCaulacanthus ustulatus\u003c/em\u003e, \u003cem\u003eLaurencia thyrsifera\u003c/em\u003e, and \u003cem\u003eChondria macrocarpa\u003c/em\u003e, typified many of the moderate- to high-performing attachment substrates for juvenile \u003cem\u003eP. canaliculus\u003c/em\u003e smaller than 10 mm SL in this study, it is not unusual for wider, sparsely branched macroalgae, such as \u003cem\u003eU. australis, C. maschalocarpum\u003c/em\u003e, and \u003cem\u003eS. sinclairii\u003c/em\u003e, to have also supported a higher presence of smaller juveniles. It is likely a limitation of this study that the 10 mm size classes used were potentially too large to capture any variation in preference for different branch widths or degrees of branching that juvenile mussels exhibit during the earliest stages of development.\u003c/p\u003e \u003cp\u003eHigher presence of smaller \u003cem\u003eP. canaliculus\u003c/em\u003e juveniles on wider, sparsely branched \u003cem\u003eU. australis\u003c/em\u003e and \u003cem\u003eS. sinclairii\u003c/em\u003e may have also been the result of higher scores for branch spacing and/or planes of branching due to tightly overlapping branches that twisted or folded onto multiple planes. There is less known about the influence of these two other branching features on juvenile mussel attachment to macroalgae. Studies on mussel settlement often sort macroalgae species into loosely defined functional groups that focus on a single branching characteristic, usually the branch width or degree of branching, when for many species of macroalgae, thinner branch widths, a higher degree of branching, multiple planes of branching, and shorter branch spacing are all frequently associated with each other. These commonly overlapping morphological traits make it difficult to determine which of these characteristics is primarily responsible for observed patterns in mussel attachment among macroalgae (Wu \u003cspan citationid=\"CR109\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). Moreover, studies utilising plastic analogues of macroalgae to isolate and test the influence of each branching feature on the settlement of juvenile mussels smaller than 1 mm SL revealed mixed results for all features except the number of branching planes, which remains to be tested in this manner. In these studies, settlement densities were unaffected by changing the distance between branches (Kelly \u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e2001\u003c/span\u003e; Wu \u003cspan citationid=\"CR109\" class=\"CitationRef\"\u003e2018\u003c/span\u003e), either remain unchanged or declined with increasing degree of branching (Harvey et al. \u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e1995\u003c/span\u003e; Kelly \u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e2001\u003c/span\u003e; Wu \u003cspan citationid=\"CR109\" class=\"CitationRef\"\u003e2018\u003c/span\u003e), and generally increased as branch width decreased\u0026ndash;except in one study that reported no change (Harvey et al. \u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e1995\u003c/span\u003e; Kelly \u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e2001\u003c/span\u003e; Wu \u003cspan citationid=\"CR109\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Wu and Jeffs \u003cspan citationid=\"CR110\" class=\"CitationRef\"\u003e2025\u003c/span\u003e; Wu et al. \u003cspan citationid=\"CR111\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). Despite these mixed results, artificial substrates designed to collect mussel larvae and grow them for aquaculture that employ the use of thinner filaments, a higher degree of three-dimensional complexity across multiple planes, and interwoven or looped filaments that provide small refuge spaces within the substrate each support higher densities of early juvenile mussel settlement (Walter and Liebezeit \u003cspan citationid=\"CR105\" class=\"CitationRef\"\u003e2003\u003c/span\u003e; Filgueira et al. \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e2007\u003c/span\u003e; Brenner and Buck \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2010\u003c/span\u003e; Protopopescu and Beal \u003cspan citationid=\"CR86\" class=\"CitationRef\"\u003e2015\u003c/span\u003e). These structurally complex morphologies allow juvenile mussels to increase their stability on the substrate through a greater surface area for the attachment of more byssal threads and the ability to surround themselves with the substrate, which can help protect against dislodgement and predation (Brenner and Buck \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2010\u003c/span\u003e; Filgueira et al. \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e2007\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eIn this study, juvenile \u003cem\u003eP. canaliculus\u003c/em\u003e attached to thinner, more highly branched macroalgal substrates were observed forming densely packed mussel-substrate conglomerates, in which juveniles were tightly attached to and surrounded by both neighbouring juveniles and the branches of the macroalgae. Similar observations were made for \u003cem\u003eMytilus edulis\u003c/em\u003e juveniles attached to three-dimensionally complex, filamentous artificial substrates, where the formation of these mussel-substrate conglomerates was associated with decreased dislodgement of juveniles from these substrates in areas of greater hydrodynamic exposure (Brenner and Buck \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2010\u003c/span\u003e). However, it is possible that at sites with particularly high hydrodynamic exposure, such as Pākiri Beach, these more structurally complex branching characteristics may not be beneficial enough for juveniles larger than 10 mm SL to overcome the shear stress generated across their larger surface areas by wave action (Donker et al. \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2013\u003c/span\u003e). This would help to explain why fewer larger juveniles were found in the branches of macroalgae at Pākiri Beach compared to Waipū Cove as well as why no branching features qualified as strong predictors for the presence of larger juveniles at Pākiri Beach. While all four branching features of macroalgae are likely important determinants of the attachment of juvenile mussels during early development, the influence of branching features on larger juvenile attachments to macroalgae appears to be reduced at sites with higher hydrodynamic exposure, although further testing would be needed to confirm if wave action is the primary cause of the observed patterns. Therefore, these findings suggest that restoration practitioners seeking to assess the branching morphology of macroalgal substrates for their potential to support not only the initial settlement, but also the continued attachment, of juvenile mussels should place greater value on high-scoring branching characteristics if the substrates are found at, or intended for, restoration sites that are characterised by lower levels of hydrodynamic stress. However, additional experimentation is needed to determine the level at which hydrodynamic stress begins to have a significant negative impact on juvenile attachments to the branches of macroalgae.\u003c/p\u003e \u003cp\u003eUnlike canopy cover, holdfast complexity, and the features of branching morphology, canopy height varied greatly among the high-performing macroalgal substrates for each size class of juvenile \u003cem\u003eP. canaliculus\u003c/em\u003e. While canopy height qualified as the strongest predictor of the presence of attached juvenile mussels between 10 and 20 mm SL on macroalgae at Waipū Cove, it was a poor predictor of juvenile presence for the other two size classes at this site and all juvenile size classes at Pākiri Beach. This inconsistency is likely due to a trade-off between the costs and benefits that taller macroalgal canopy heights offer juvenile mussels. Some studies suggest that taller canopies can enhance the survival of juvenile mussels attached to their branches by lifting them away from scouring sediments, predators, and competition with adult mussels (Seed and Suchanek \u003cspan citationid=\"CR93\" class=\"CitationRef\"\u003e1992\u003c/span\u003e; Westerbom et al. \u003cspan citationid=\"CR106\" class=\"CitationRef\"\u003e2008\u003c/span\u003e) and up into the area above the seabed, where higher water flow and food availability can support higher growth rates (Fr\u0026eacute;chette and MMI. However, macroalgae with taller canopies are also more susceptible to \u0026lsquo;whiplash effects\u0026rsquo; under higher wave action, in which the movement of the thallus can dislodge both epibiotic organisms attached to the canopies and understory organisms that may be attached to the holdfasts (Leonard \u003cspan citationid=\"CR62\" class=\"CitationRef\"\u003e1999\u003c/span\u003e; O\u0026rsquo;Connor et al. \u003cspan citationid=\"CR83\" class=\"CitationRef\"\u003e2006\u003c/span\u003e; Erlandsson et al. \u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e2008\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eJuvenile mussels may be able to tolerate the more dynamic movement of taller macroalgae if they also exhibit morphological characteristics that allow for multiple, firm attachment points, such as species with more three-dimensionally complex branching or holdfast morphologies, as greater structural complexity has been shown to reduce the dislodgement of juvenile mussels attached to artificial substrates (Brenner and Buck \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2010\u003c/span\u003e). Variable morphologies like these would help to explain why large but more highly branched species, such as \u003cem\u003eS. sinclairii\u003c/em\u003e and \u003cem\u003eC. maschalocarpum\u003c/em\u003e, or taller species with more complex holdfasts, such as \u003cem\u003eP. alveatus\u003c/em\u003e and \u003cem\u003eH. banksii\u003c/em\u003e, supported a higher presence of juvenile mussels than tall but less structurally complex species, such as \u003cem\u003eE. radiata\u003c/em\u003e or \u003cem\u003eDictyota kunthii\u003c/em\u003e. Additionally, greater shelter from wave action among the more physically complex rock formations at Waipū Cove may have reduced the dislodging effects of whiplash on mid-sized juveniles attached to taller macroalgae, allowing the advantages of taller canopy heights to drive greater presence of juveniles between 10 and 20 mm SL in more sheltered areas of the reef. Confirmation that the interaction of macroalgal morphology and wave action is responsible for these observed patterns in the distribution of juvenile mussels will need to come from further research. These findings once again highlight the importance of considering the hydrodynamic conditions at restoration sites when selecting suitable macroalgal substrates for the facilitation of juvenile mussel recruitment. However, while canopy height may have a stronger influence on the presence of juvenile mussels among macroalgae at some sites, supporting its inclusion in the MMI, the overall inconsistency of its impact suggests that this feature one of the least important drivers of juvenile mussel attachment.\u003c/p\u003e \u003cp\u003eClustering frequency was another morphological feature that only once qualified as a strong predictor of \u003cem\u003eP. canaliculus\u003c/em\u003e presence among macroalgae in the case of juveniles smaller than 10 mm SL at Pākiri Beach. This outcome can most likely be attributed to the large number of both high- and low-performing macroalgal substrates for each juvenile size class that received high scores for their tendency to frequently cluster in patches or continuous turfs comprised of the same species, especially at Waipū Cove. Macroalgae with high clustering frequencies are awarded high scores in the MMI because a greater number macroalgal individuals across a smaller area is more likely to produce a stronger localised chemical cue, and previous studies show that chemical cues from algal extracts and macroalgae placed in close proximity can promote the settlement and retention of juvenile mussels (Dobretsov \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e1999\u003c/span\u003e; Alfaro et al. \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2006\u003c/span\u003e; Skelton and Jeffs \u003cspan citationid=\"CR95\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). While the MMI does not score the biochemical effects of macroalgae on juvenile mussel attachment, scoring the clustering frequency has the benefit of incorporating some of the influence of these effects on mussel settlement, justifying the preservation of this feature within the MMI. However, since a higher clustering frequency was not more common in the high-performing macroalgae than in the low-performing macroalgae, this feature is only likely to be useful in the assessment of macroalgae as attachment substrates when a substrate exhibits a particularly low clustering frequency, suggesting a lower suitability for use in restoration.\u003c/p\u003e \u003cp\u003eEvaluation of the Macroalgal Morphology Index\u003c/p\u003e \u003cp\u003eAt both sites, the species of macroalgae that supported the highest presence of attached \u003cem\u003eP. canaliculus\u003c/em\u003e juveniles in each size class received moderate to high total scores according to the criteria in the MMI. This finding suggests that the MMI has a high capability of predicting the relative presence of juvenile mussels among different macroalgae species based on their morphology. Patterns in the macroalgal associations of \u003cem\u003eP. canaliculus\u003c/em\u003e juveniles on two remnant mussel reefs indicate that macroalgae with morphological characteristics that receive a total index score of at least 14 out of 24 likely have a high potential to facilitate juvenile recruitment at mussel reef restoration sites. However, the observed macroalgal association patterns only partially validate the collective accuracy of the MMI\u0026rsquo;s scoring criteria because there were a number of other macroalgae species with total index scores of 14 or higher that supported a relatively low presence of \u003cem\u003eP. canaliculus\u003c/em\u003e juveniles in each size class but especially in the two larger size classes. The MMI is limited in its ability to capture all aspects of a macroalgal substrate that can positively or negatively affect the attachment of juvenile mussels because in addition to a macroalga\u0026rsquo;s morphology, the chemical cues that it produces (Kelly \u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e2001\u003c/span\u003e; Alfaro et al. \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2006\u003c/span\u003e; Gribben et al. \u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e2011\u003c/span\u003e), the microscale texture of its surfaces (Gribben et al. \u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e2011\u003c/span\u003e; Loucks \u003cspan citationid=\"CR66\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Wu and Jeffs \u003cspan citationid=\"CR110\" class=\"CitationRef\"\u003e2025\u003c/span\u003e), and the location and environmental conditions in which it is normally found all have the potential to impact juvenile presence on that substrate (Hunt and Scheibling \u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e1995\u003c/span\u003e, \u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e1996\u003c/span\u003e, \u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e2001\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThe scores and performances of \u003cem\u003eG. caulacantheum\u003c/em\u003e and \u003cem\u003eU. australis\u003c/em\u003e provide good examples of the limitations of the MMI. The red macroalga \u003cem\u003eG. caulacantheum\u003c/em\u003e is a highly and thinly branched, turf-forming species that received a high total index score of 21, but it was a low-performing attachment substrate for juveniles in all size classes at Waipū Cove. Its total index score does not reflect the tendency for this species to grow exclusively on vertical rockfaces in the upper intertidal, where higher exposure to wave action and desiccation stress would make this substrate a much less suitable location for juvenile mussels to establish, especially as their susceptibility to these stressors increases with size. In contrast to \u003cem\u003eG. caulacantheum\u003c/em\u003e, the sparsely and broadly branched green macroalga \u003cem\u003eU. australis\u003c/em\u003e was a high-performing attachment substrate for juveniles up to 20 mm SL at Waipū Cove, but it only received a moderate total index score of 14. This score does not reflect the tendency for \u003cem\u003eU. australis\u003c/em\u003e to grow in close proximity to available freshwater outputs (Nelson \u003cspan citationid=\"CR82\" class=\"CitationRef\"\u003e2020\u003c/span\u003e), such as the stream that empties into the ocean on the north side of the mussel reef at Waipū Cove. These locations tend to experience a high input of allochthonous organic nutrients that can benefit the growth of both \u003cem\u003eU. australis\u003c/em\u003e and juvenile mussels, which may explain why \u003cem\u003eU. australis\u003c/em\u003e supported such a high presence of juvenile mussels at Waipū Cove but not at Pākiri Beach, where there are no nearby freshwater outputs. These findings underscore the need for mussel reef restoration practitioners using the MMI to interpret the total index score that a substrate receives with a reasonable amount of caution and consideration for whether that species of macroalgae is likely to grow best under environmental conditions that also benefit the growth and survival of juvenile mussels.\u003c/p\u003e \u003cp\u003eThe MMI\u0026rsquo;s criteria could potentially be improved by testing and modifying its scores based on additional observations of juvenile mussel presence among macroalgae at more sites that experience a variety of environmental conditions and support different species of macroalgae than those observed in this study. In particular, the MMI would benefit from being tested on macroalgal association data collected from subtidal mussel reefs, where the lack of desiccation stress and reduced wave action may promote different morphological features of macroalgae as the primary drivers of juvenile mussel attachment than those identified for intertidal mussel reefs. Additionally, more studies are needed to improve our understanding of the mechanisms through which each of these morphological features influences the settlement, survival, and establishment of juvenile mussels on macroalgae under a range of physical and biological stressors. Importantly, while the morphological features and scoring criteria used in the MMI were identified from literature on a wide range of mussel species, observations of the macroalgal associations of other species of juvenile mussels on remnant reefs would be valuable for confirming that the MMI is applicable for mussel species other than \u003cem\u003eP. canaliculus\u003c/em\u003e.\u003c/p\u003e"},{"header":"Conclusions","content":"\u003cp\u003eWhen totalled, the scores assigned to each species of macroalgae for each of the eight morphological features in the MMI were able to predict the presence of attached juvenile \u003cem\u003eP. canaliculus\u003c/em\u003e with a relatively high accuracy of 75% for juveniles smaller than 10 mm SL but a low accuracy of 40\u0026ndash;60% for juveniles 10 mm SL and larger. The majority of morphological features used to create the MMI were identified from studies that focus on the impact of substrate morphology on larval and early juvenile mussel settlement, while very few studies have been conducted on how substrate morphology influences the retention of juveniles throughout later development. Therefore, many of these features, in particular the features of macroalgae that describe their branching morphology, while useful in predicting a macroalga\u0026rsquo;s potential to support the initial attachment of settling juveniles, become largely irrelevant as juveniles attach less frequently to the branches of macroalgae and more frequently to the holdfasts as they grow. Above 10 mm SL, juveniles likely become more susceptible to sources of post-settlement detachment, stress, or mortality that appear to be primarily mediated by just two of the eight features in the MMI, macroalgal canopy cover and holdfast complexity. Therefore, the total index score appears to be collectively useful for predicting a macroalga\u0026rsquo;s potential to support the settlement of larvae and smaller juvenile mussels. However, evaluating the likelihood that a substrate will aid the continued attachment of larger juvenile mussels may require users of the MMI to give greater weight to scores for canopy cover and holdfast complexity as well as consider whether environmental conditions at the proposed restoration site are likely to have a positive or negative interaction with canopy height. Additionally, it is possible that if multiple macroalgae species with different sets of high-scoring morphological characteristics are transplanted into restoration sites together, such as species that exhibit thinner branch widths, a higher degree of branching, shorter branch spacing, and multiple planes of branching being paired with other species that exhibit more complex holdfasts and widespread canopy covers, these substrates may work together to secure the attachment of juvenile mussels throughout all stages of their development.\u003c/p\u003e \u003cp\u003eOverall, the results of this study suggest that the MMI can be used to accurately rank various macroalgal substrates for their potential to support the recruitment of juvenile mussels based on the morphology of the substrate. The quantitative and qualitative criteria outlined in the MMI are designed to ensure that macroalgal substrata are assessed and scored in a manner that is objective and repeatable for its users. Additionally, the MMI can begin to remove the ambiguity associated with the wide variety of terms for differing macroalgal morphology that have been used throughout the literature to describe the settlement preferences of juvenile mussels, instead condensing these descriptions into just three scored variations of eight overarching structural features of macroalgae. Ultimately, the MMI can serve as a useful tool for mussel reef restoration practitioners seeking to: 1) determine the most suitable species of macroalgae to transplant into current restoration sites, 2) assess new restoration sites with existing macroalgal habitat for their potential to support juvenile recruitment, and 3) apply the highest scoring morphological characteristics outlined in the MMI to the design of artificial substrata for particularly degraded restoration sites that are unlikely to support the growth of macroalgal substrates. While the MMI\u0026rsquo;s scoring criteria were only tested on the attachment substrates of \u003cem\u003eP. canaliculus\u003c/em\u003e juveniles, its application extends to all mussel species that are the focus of restoration efforts. The MMI and the methods used in this study to validate its accuracy provide a solid foundation for future studies evaluating the effective use of macroalgal substrates in mussel reef restoration projects across the globe.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAcknowledgements\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe thank and acknowledge the tangata whenua (indigenous people of the land) of Waipū Cove and Pākiri Beach as the traditional owners and kaitiaki (guardians) of the land, coast, and oceans within which this study was undertaken. We would also like to thank The Nature Conservancy Aotearoa New Zealand for providing the funds to make this research possible.\u003c/p\u003e\n\u003cp\u003eFunding\u003c/p\u003e\n\u003cp\u003eFunding for this research was provided by The Nature Conservancy Aotearoa New Zealand.\u003c/p\u003e\n\u003cp\u003eConflict of Interest\u003c/p\u003e\n\u003cp\u003eAll\u0026nbsp;authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.\u003c/p\u003e\n\u003cp\u003eAuthors\u0026rsquo; Contribution\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eKatherine A. Burnham\u003c/strong\u003e: Writing \u0026ndash; original draft, Visualization,Validation, Methodology, Investigation, Formal analysis, Data curation,Conceptualization. \u003cstrong\u003eJenny R. Hillman\u003c/strong\u003e: Writing \u0026ndash; review \u0026amp; editing,Validation, Supervision, Resources, Project administration, Methodology,Funding acquisition, Conceptualization. \u003cstrong\u003eAndrew G. Jeffs\u003c/strong\u003e: Writing\u0026ndash; review \u0026amp; editing, Validation, Supervision, Resources, Project administration,Methodology, Funding acquisition, Conceptualization.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eEthics Approval\u003c/p\u003e\n\u003cp\u003eThis study was conducted in compliance with ethical standards and all applicable guidelines for sampling of organisms have been followed. \u0026nbsp;\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eData Availability\u003c/p\u003e\n\u003cp\u003eThe datasets generated and analysed during the current study are available in the Mendeley data repository, [10.17632/2dy4hgn9fw.1, 10.17632/fd2fhwxfdp.1].\u0026nbsp;\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eAlder A, Jeffs A, Hillman JR (2020) Considering the use of subadult and juvenile mussels for mussel reef restoration. Restor Ecol 29(3):e13322. https://doi.org/10.1111/rec.13322\u003c/li\u003e\n\u003cli\u003eAlfaro AC, Jeffs AG, Hooker SH (2001) Reproductive behavior of the green-lipped mussel, \u003cem\u003ePerna canaliculus\u003c/em\u003e, in Northern New Zealand. Bull Mar Sci 69(3). https://www.researchgate.net/publication/233520444\u003c/li\u003e\n\u003cli\u003eAlfaro AC, Jeffs AG (2002) Small-scale mussel settlement patterns within morphologically distinct substrates at Ninety Mile Beach, northern New Zealand. Malacologia 44:1\u0026ndash;16. https://www.researchgate.net/publication/285776689 \u003c/li\u003e\n\u003cli\u003eAlfaro AC, Jeffs AG, Creese RG (2004) Bottom-drifting algal/mussel spat associations along a sandy coastal region in northern New Zealand. Aquaculture 241(1\u0026ndash;4):269\u0026ndash;290. https://doi.org/10.1016/j.aquaculture.2004.07.029\u003c/li\u003e\n\u003cli\u003eAlfaro AC, Copp BR, Appleton DR, Kelly S, Jeffs AG (2006) Chemical cues promote settlement in larvae of the green-lipped mussel, \u003cem\u003ePerna canaliculus\u003c/em\u003e. Aquacult Int 14:405\u0026ndash;412. https://doi.org/10.1007/s10499-005-9041-y\u003c/li\u003e\n\u003cli\u003eAghzar A, Talbaoui M, Benajiba MH, Presa P (2012) Influence of depth and diameter of rope collectors on settlement density of \u003cem\u003eMytilus galloprovincialis\u003c/em\u003e spat in Baie de M\u0026rsquo;diq (Alboran Sea). Mar Freshw Behav Physiol 45(1):51\u0026ndash;61. https://doi.org/10.1080/10236244.2012.678759 \u003c/li\u003e\n\u003cli\u003eBanke TL, Steinfurth RC, Lange T, Canal‐Verg\u0026eacute;s P, Svane N, Flindt MR (2024) Dislodgement and mortality challenges when restoring shallow mussel beds (\u003cem\u003eMytilus edulis\u003c/em\u003e) in a Danish estuary. Restor Ecol 32(5):e14160. https://doi.org/10.1111/rec.14160\u003c/li\u003e\n\u003cli\u003eBayne BL (1964) Primary and secondary settlement in \u003cem\u003eMytilus edulis\u003c/em\u003e L. (Mollusca). J Anim Ecol 33(3):513. https://doi.org/10.2307/2569\u003c/li\u003e\n\u003cli\u003eBeck MW, Brumbaugh RD, Airoldi L, Carranza A, Coen LD, Crawford C, Defeo O, Edgar GJ, Hancock B, Kay M, Lenihan H, Luckenbach MW, Toropova CL, Zhang G (2009) Shellfish Reefs at Risk: A Global Analysis of Problems and Solutions. The Nature Conservancy, Arlington, Virginia, United States of America. https://www.researchgate.net/publication/265238947\u003c/li\u003e\n\u003cli\u003eB\u0026eacute;gin C, Johnson LE, Himmelman JH (2004) Macroalgal canopies: distribution and diversity of associated invertebrates and effects on the recruitment and growth of mussels. Mar Ecol Prog Ser 271:121\u0026ndash;132. https://doi.org/10.3354/meps271121\u003c/li\u003e\n\u003cli\u003eBenjamin ED, Handley SJ, Hale R, Toone TA, Jeffs A, Hillman JR (2022a) Biodiversity associated with restored small-scale mussel habitats has restoration decision implications. Biodivers Conserv 31(11):2833\u0026ndash;2855. https://doi.org/10.1007/s10531-022-02462-1\u003cu\u003e \u003c/u\u003e\u003c/li\u003e\n\u003cli\u003eBenjamin ED, Hillman JR, Handley SJ, Toone TA, Jeffs A (2022b) The effectiveness of providing shell substrate for the restoration of adult mussel reefs. Sustainability 14(23):15746. https://doi.org/10.3390/su142315746\u003c/li\u003e\n\u003cli\u003eBenjamin E, Jeffs A, Handley S, Toone T, Hillman J (2023) Determining restoration potential by transplanting mussels of different size classes over a range of aerial exposures. Mar Ecol Prog Ser 713:71\u0026ndash;81. https://doi.org/10.3354/meps14337\u003c/li\u003e\n\u003cli\u003eBeukema JJ, Cad\u0026eacute;e GC (1996) Consequences of the sudden removal of nearly all mussels and cockles from the Dutch Wadden Sea. Mar Ecol 17(1-3):279\u0026ndash;289. https://doi.org/10.1111/j.1439-0485.1996.tb00508.x \u003c/li\u003e\n\u003cli\u003evan den Bogaart LA, Schotanus J, Capelle JJ, Bouma TJ (2023) Using a biodegradable substrate to increase transplantation success: effect of density and sediment on aggregation behavior of mussels. Ecol Eng 196:107096. https://doi.org/10.1016/j.ecoleng.2023.107096 \u003c/li\u003e\n\u003cli\u003eBourget E, DeGuise J, Daigle G (1994) Scales of substratum heterogeneity, structural complexity, and the early establishment of a marine epibenthic community. J Exp Mar Biol Ecol 181(1):31\u0026ndash;51. https://doi.org/10.1016/0022-0981(94)90102-3\u003c/li\u003e\n\u003cli\u003eBrenner M, Buck BH (2010) Attachment properties of blue mussel (\u003cem\u003eMytilus edulis\u003c/em\u003e L.) byssus threads on culture-based artificial collector substrates. Aquac Eng 42(3):128\u0026ndash;139. https://doi.org/10.1016/j.aquaeng.2010.02.001 \u003c/li\u003e\n\u003cli\u003eBrumbaugh RD, Beck MW, Coen LD, Craig L, Hicks P (2006) A practitioners\u0026rsquo; guide to the design and monitoring of shellfish: an ecosystem services approach. MRD Education Report, The Nature Conservancy, Arlington, Virginia, United States of America. https://repository.library.noaa.gov/view/noaa/3470\u003c/li\u003e\n\u003cli\u003eBrumbaugh RD, Coen LD (2009) Contemporary approaches for small-scale oyster reef restoration to address substrate versus recruitment limitation: a review and comments relevant for the Olympia oyster, \u003cem\u003eOstrea lurida\u003c/em\u003e Carpenter 1864. J Shellfish Res 28(1):147\u0026ndash;161. https://doi.org/10.2983/035.028.0105\u003c/li\u003e\n\u003cli\u003eBuchanan S, Babcock R (1997) Primary and secondary settlement by the green-shell mussel \u003cem\u003ePerna canaliculus\u003c/em\u003e. Oceanogr Lit Rev 12:1500. https://www.researchgate.net/publication/279937722\u003c/li\u003e\n\u003cli\u003eBulleri F, Airoldi L, Branca GM, Abbiati M (2006) Positive effects of the introduced green alga, \u003cem\u003eCodium fragile\u003c/em\u003e ssp. \u003cem\u003etomentosoides\u003c/em\u003e, on recruitment and survival of mussels. Mar Biol 148:1213\u0026ndash;1220. https://doi.org/10.1007/s00227-005-0181-4 \u003c/li\u003e\n\u003cli\u003eBurnham KA, Hillman JR, Jeffs AG (2025) Macroalgae and conspecifics serve as critical attachment substrates for juvenile green-lipped mussels (\u003cem\u003ePerna canaliculus\u003c/em\u003e) during different stages of development. J Exp Mar Biol Ecol 588:152108. https://doi.org/10.1016/j.jembe.2025.152108\u003c/li\u003e\n\u003cli\u003eCapelle JJ, Leuchter L, de Wit M, Hartog E, Bouma TJ (2019) Creating a window of opportunity for establishing ecosystem engineers by adding substratum: a case study on mussels. Ecosphere 10(4):e02688. https://doi.org/10.1002/ecs2.2688\u003c/li\u003e\n\u003cli\u003eCarl C, Poole AJ, Sexton BA, Glenn FL, Vucko MJ, Williams MR, Whalan S, de Nys R (2012a) Enhancing the settlement and attachment strength of pediveligers of \u003cem\u003eMytilus galloprovincialis\u003c/em\u003e by changing surface wettability and microtopography. Biofouling 28(2):175\u0026ndash;186. https://doi.org/10.1080/08927014.2012.662676\u003c/li\u003e\n\u003cli\u003eCarl C, Poole AJ, Williams MR, de Nys R (2012b) Where to settle \u0026ndash; settlement preferences of \u003cem\u003eMytilus galloprovincialis\u003c/em\u003e and choice of habitat at a micro spatial scale. PLoS ONE 7(12):e52358. https://doi.org/10.1371/journal.pone.0052358\u003c/li\u003e\n\u003cli\u003eCarranza A, Zu Ermgassen PS (2020) A global overview of restorative shellfish mariculture. Front Mar Sci 7:722. \u003cu\u003ehttps://doi.org/10.3389/fmars.2020.00722\u003c/u\u003e\u003c/li\u003e\n\u003cli\u003eChristensen HT, Dolmer P, Hansen BW, Holmer M, Kristensen LD, Poulsen LK, Sternberg C, Albertsen CM, St\u0026oslash;ttrup JG (2015) Aggregation and attachment responses of blue mussels, \u003cem\u003eMytilus edulis\u003c/em\u003e \u0026ndash; impact of substrate composition, time scale and source of mussel seed. Aquaculture 435:245\u0026ndash;251. https://doi.org/10.1016/j.aquaculture.2014.09.043\u003c/li\u003e\n\u003cli\u003eCommito JA, Commito AE, Platt RV, Grupe BM, Piniak WED, Gownaris NJ, Reeves KA, Vissichelli AM (2014) Recruitment facilitation and spatial pattern formation in soft‐bottom mussel beds. Ecosphere 5(12):1\u0026ndash;26. https://doi.org/10.1890/ES14-00200.1\u003c/li\u003e\n\u003cli\u003ede Nesnera KL (2016) Stress, ontogeny, and movement determine the relative importance of facilitation for juvenile mussels. Ecol 97(9):2199\u0026ndash;2205. https://doi.org/10.1002/ecy.1505\u003c/li\u003e\n\u003cli\u003ede Paoli H, van de Koppel J, van der Zee E, Kangeri A, van Belzen J, Holthuijsen S, van den Berg A, Herman P, Olff H, van der Heide T (2015) Processes limiting mussel bed restoration in the Wadden Sea. J Sea Res 103:42-49. https://doi.org/10.1016/j.seares.2015.05.008\u003c/li\u003e\n\u003cli\u003eDobretsov SV (1999) Effects of macroalgae and biofilm on settlement of blue mussel (\u003cem\u003eMytilus edulis\u003c/em\u003e L.) larvae. Biofouling 14(2):153\u0026ndash;165. https://doi.org/10.1080/08927019909378406 \u003c/li\u003e\n\u003cli\u003eDobretsov S, Wahl M (2001) Recruitment preferences of blue mussel spat (\u003cem\u003eMytilus edulis\u003c/em\u003e) for different substrates and microhabitats in the White Sea (Russia). Hydrobiologia 445:27\u0026ndash;35. https://doi.org/10.1023/A:1017502126707 \u003c/li\u003e\n\u003cli\u003eDobretsov S, Wahl M (2008) Larval recruitment of the blue mussel \u003cem\u003eMytilus edulis\u003c/em\u003e: the effect of flow and algae. J Exp Mar Biol Ecol 355(2):137\u0026ndash;144. https://doi.org/10.1016/j.jembe.2007.12.018 \u003c/li\u003e\n\u003cli\u003eDonker JJA, Van Der Vegt M, Hoekstra P (2013) Wave forcing over an intertidal mussel bed. J Sea Res 82:54\u0026ndash;66. https://doi.org/10.1016/j.seares.2012.08.010\u003c/li\u003e\n\u003cli\u003eEckman JE, Duggins DO (1991) Life and death beneath macrophyte canopies: effects of understory kelps on growth rates and survival of marine, benthic suspension feeders. Oecologia 87:473\u0026ndash;487. https://doi.org/10.1007/BF00320409 \u003c/li\u003e\n\u003cli\u003eElsevier (2025) Scopus. https://www.scopus.com\u003c/li\u003e\n\u003cli\u003eErlandsson J, McQuaid C (2004) Spatial structure of recruitment in the mussel \u003cem\u003ePerna perna \u003c/em\u003eat local scales: effects of adults, algae and recruit size. Mar Ecol Prog Ser 267:173\u0026ndash;185. https://doi.org/10.3354/meps267173\u003c/li\u003e\n\u003cli\u003eErlandsson J, McQuaid CD, Kostylev VE (2005) Contrasting spatial heterogeneity of sessile organisms within mussel (\u003cem\u003ePerna perna\u003c/em\u003e L.) beds in relation to topographic variability. J Exp Mar Biol Ecol 314(1):79\u0026ndash;97. https://doi.org/10.1016/j.jembe.2004.09.010\u003c/li\u003e\n\u003cli\u003eFilgueira R, Peteiro LG, Labarta U, Fern\u0026aacute;ndez-Reiriz MJ (2007) Assessment of spat collector ropes in Galician mussel farming. Aquac Eng 37(3):195\u0026ndash;201. https://doi.org/10.1016/j.aquaeng.2007.06.001 \u003c/li\u003e\n\u003cli\u003eEriksson BK, van der Heide T, van de Koppel J, Piersma T, van der Veer HW, Olff H (2010) Major changes in the ecology of the Wadden Sea: human impacts, ecosystem engineering and sediment dynamics. Ecosyst 13:752\u0026ndash;764. https://doi.org/10.1007/s10021-010-9352-3 \u003c/li\u003e\n\u003cli\u003eErlandsson J, Porri F, McQuaid CD (2008) Ontogenetic changes in small-scale movement by recruits of an exploited mussel: implications for the fate of larvae settling on algae. Mar Biol 153(3):365\u0026ndash;373. https://doi.org/10.1007/s00227-007-0812-z\u003c/li\u003e\n\u003cli\u003eFitzsimons J, Branigan S, Brumbaugh RD, McDonald T, Zu Ermgassen PSE (2019) Restoration Guidelines for Shellfish Reefs. The Nature Conservancy, Arlington, Virginia, United States of America. https://doi.org/10.13140/RG.2.2.22511.38563 \u003c/li\u003e\n\u003cli\u003eFitzsimons JA, Branigan S, Gillies CL, Brumbaugh RD, Cheng J, DeAngelis BM, Geselbracht L, Hancock B, Jeffs A, McDonald T, McLeod IM, Pogoda B, Theuerkauf SJ, Thomas M, Westby S, Zu Ermgassen PSE (2020) Restoring shellfish reefs: global guidelines for practitioners and scientists. Conserv Sci Pract 2(6):e198. https://doi.org/10.1111/csp2.198\u003c/li\u003e\n\u003cli\u003eFord JR, Hamer P (2016) The forgotten shellfish reefs of coastal Victoria: documenting the loss of a marine ecosystem over 200 years since European settlement. Proceedings of the Royal Society of Victoria 128(1):87. https://doi.org/10.1071/RS16008\u003c/li\u003e\n\u003cli\u003eFrandsen R, Dolmer P (2002) Effects of substrate type on growth and mortality of blue mussels (\u003cem\u003eMytilus edulis\u003c/em\u003e) exposed to the predator \u003cem\u003eCarcinus maenas\u003c/em\u003e. Mar Biol 141:253\u0026ndash;262. https://doi.org/10.1007/s00227-002-0831-8 \u003c/li\u003e\n\u003cli\u003eFr\u0026eacute;chette M, Bourget E (1985) Food-limited growth of \u003cem\u003eMytilus edulis\u003c/em\u003e L. in relation to the benthic boundary layer. Can J Fish Aquat Sci 42(6):1166\u0026ndash;1170. https://doi.org/10.1139/f85-144 \u003c/li\u003e\n\u003cli\u003eGagnon K, Hartvig C, Didderen K, Fagerli CW, Govers LL, Gräfnings MLE, Heusinkveld JHT, Kaljurand K, Lengkeek W, Martin G, Meysick L, Pajusalu L, Rinde E, van der Heide T, Boström C (2021) Incorporating facilitative interactions into small‐scale eelgrass restoration\u0026mdash;challenges and opportunities. Restor Ecol 29(5):e13398. https://doi.org/10.1111/rec.13398\u003c/li\u003e\n\u003cli\u003eGann GD, McDonald T, Walder B, Aronson J, Nelson CR, Jonson J, Hallett JG, Eisenberg C, Guariguata MR, Liu J, Hua F, Echeverr\u0026iacute;a C, Gonzales E, Shaw N, Decleer K, Dixon KW (2019) International principles and standards for the practice of ecological restoration. Restor Ecol 27(S1):S1\u0026ndash;S46. https://doi.org/10.1111/rec.13035\u003c/li\u003e\n\u003cli\u003eGreenway JPC (1969) Survey of mussels (Mollusca: Lamellibranchia) in the firth of Thames, 1961\u0026ndash;1967. N Z J Mar Freshw Res 3:304\u0026ndash;317. https://ref.coastalrestorationtrust.org.nz/site/assets/files/9479/surveys_of_mussels_mollusca_lamellibranchia_in_the_firth_of_thames_1961_67.pdf \u003c/li\u003e\n\u003cli\u003eGribben PE, Jeffs AG, de Nys R, Steinberg PD (2011) Relative importance of natural cues and substrate morphology for settlement of the New Zealand Greenshell\u003csup\u003e\u0026trade;\u003c/sup\u003e mussel, \u003cem\u003ePerna canaliculus\u003c/em\u003e. Aquaculture 319(1\u0026ndash;2):240\u0026ndash;246. https://doi.org/10.1016/j.aquaculture.2011.06.026 \u003c/li\u003e\n\u003cli\u003eHarvey M, Bourget E, Ingram RG (1995) Experimental evidence of passive accumulation of marine bivalve larvae on filamentous epibenthic structures. Limnol Oceanogr 40(1):94\u0026ndash;104. https://doi.org/10.4319/lo.1995.40.1.0094 \u003c/li\u003e\n\u003cli\u003eHauraki Gulf Forum (2023) State of our Gulf Report. https://gulfjournal.org.nz/wp- content/uploads/2023/08/SOER-sum-online.pdf \u003c/li\u003e\n\u003cli\u003eHolthuis TD, Bergstr\u0026ouml;m P, Lindegarth M, Lindegarth S (2015) Monitoring recruitment patterns of mussels and fouling tunicates in mariculture. J Shellfish Res 34(3):1007\u0026ndash;1018. https://doi.org/10.2983/035.034.0327\u003c/li\u003e\n\u003cli\u003eHunt HL, Scheibling R (1995) Structure and dynamics of mussel patches in tidepools on a rocky shore in Nova Scotia, Canada. Mar Ecol Prog Ser 124:105\u0026ndash;115. https://doi.org/10.3354/meps124105\u003c/li\u003e\n\u003cli\u003eHunt H, Scheibling R (1996) Physical and biological factors influencing mussel (\u003cem\u003eMytilus trossulus, M. edulis) \u003c/em\u003esettlement on a wave-exposed rocky shore. Mar Ecol Prog Ser 142:135\u0026ndash;145. https://doi.org/10.3354/meps142135\u003c/li\u003e\n\u003cli\u003eHunt H, Scheibling R (1998) Spatial and temporal variability of patterns of colonization by mussels (\u003cem\u003eMytilus trossulus, M. edulis\u003c/em\u003e) on a wave-exposed rocky shore. Mar Ecol Prog Ser 167:155\u0026ndash;169. https://doi.org/10.3354/meps167155\u003c/li\u003e\n\u003cli\u003eHunt HL, Scheibling RE (2001) Patch dynamics of mussels on rocky shores: integrating process to understand pattern. Ecol 82(11):3213\u0026ndash;3231. \u003cu\u003ehttps://www.jstor.org/stable/2679845\u003c/u\u003e \u003c/li\u003e\n\u003cli\u003eKelly S (2001) Improving the supply of mussel spat. Kinnears Ropes Ltd. and Kaitaia Spat Ltd. Technical Report. \u003c/li\u003e\n\u003cli\u003eKobak J (2001) Light, gravity and conspecifics as cues to site selection and attachment behaviour of juvenile and adult \u003cem\u003eDreissena polymorpha\u003c/em\u003e Pallas, 1771. J Molluscan Stud 67(2):183\u0026ndash;189. https://doi.org/10.1093/mollus/67.2.183\u003c/li\u003e\n\u003cli\u003eLanham BS, Pomeroy AW, Swearer SE, Marusic I, Javaherchian J, Morris RL (2025) Substrate-mediated alterations to hydrodynamic conditions enhances shellfish larval settlement: implications for artificial reef restoration. Ecol Eng 212:107474. https://doi.org/10.1016/j.ecoleng.2024.107474\u003c/li\u003e\n\u003cli\u003eLao NCH (2016) Rehabilitation of Red Macroalgae in the Hauraki Gulf. Master\u0026rsquo;s Thesis, University of Auckland. https://hdl.handle.net/2292/31829\u003c/li\u003e\n\u003cli\u003eLeonard GH (1999) Positive and negative effects of intertidal algal canopies on recruitment and survival of barnacles. Mar Ecol Prog Ser 178:241\u0026ndash;249. https://www.jstor.org/stable/24852065\u003c/li\u003e\n\u003cli\u003eLipcius RN, Burke RP (2018) Successful recruitment, survival and long-term persistence of eastern oyster and hooked mussel on a subtidal, artificial restoration reef system in Chesapeake Bay. PLoS ONE 13(10):e0204329. https://doi.org/10.1371/journal.pone.0204329\u003c/li\u003e\n\u003cli\u003eLiu W, Xu M, Wang Z, Yu D, Zhour X (2017) Experimental study of attachment characteristics of golden mussel larvae. Acta Ecol Sin 37(8):2779\u0026ndash;2787. https://doi.org/10.5846/stxb201601030010 \u003c/li\u003e\n\u003cli\u003eLotze HK, Lenihan HS, Bourque BJ, Bradbury RH, Cooke RG, Kay MC, Kidwell SM, Kirby MX, Peterson CH, Jackson JBC (2006) Depletion, degradation, and recovery potential of estuaries and coastal seas. Science 312(5781):1806\u0026ndash;1809. https://www.jstor.org/stable/3846557 \u003c/li\u003e\n\u003cli\u003eLoucks AJ (2023) Settlement preferences of green-lipped mussel larvae on key seaweed species from Hauraki Gulf. Honours Thesis, University of Auckland.\u003c/li\u003e\n\u003cli\u003eManighetti B, Carter L (1999) Across-shelf sediment dispersal, Hauraki Gulf, New Zealand. Mar Geol 160(3):271\u0026ndash;300. https://doi.org/10.1016/S0025-3227(99)00024-9 \u003c/li\u003e\n\u003cli\u003eMarsden JE, Lansky DM (2000) Substrate selection by settling zebra mussels, \u003cem\u003eDreissena polymorpha\u003c/em\u003e, relative to material, texture, orientation, and sunlight. Can J Zool 78(5):787\u0026ndash;793. https://doi.org/10.1139/z00-004 \u003c/li\u003e\n\u003cli\u003eMcAfee D, Connell SD (2020) Cuing oyster recruitment with shell and rock: implications for timing reef restoration. Restor Ecol 28(3):506\u0026ndash;511. https://doi.org/10.1111/rec.13134\u003c/li\u003e\n\u003cli\u003eMcAfee D, McLeod LD, Carruthers S, Connell SD (2024) Reversing functional extinction: successful restoration of eradicated oyster reefs. Restor Ecol 32(5):e14169. https://doi.org/10.1111/rec.14169 \u003c/li\u003e\n\u003cli\u003eMcCay DPF, Peterson CH, DeAlteris JT, Catena J (2003) Restoration that targets function as opposed to structure: replacing lost bivalve production and filtration. Mar Ecol Prog Ser\u003cem\u003e \u003c/em\u003e264:197\u0026ndash;212. https://www.jstor.org/stable/24867511\u003c/li\u003e\n\u003cli\u003eMcCook LJ, Chapman ARO (1991) Community succession following massive ice-scour on an exposed rocky shore: effects of \u003cem\u003eFucus\u003c/em\u003e canopy algae and of mussels during late succession. J Exp Mar Bio Ecol 154(2):137\u0026ndash;169. https://doi.org/10.1016/0022-0981(91)90161-O\u003c/li\u003e\n\u003cli\u003eMcLeod IM (2009) Green-lipped mussels, \u003cem\u003ePerna canaliculus\u003c/em\u003e, in soft-sediment systems in northeastern New Zealand. Master\u0026rsquo;s Thesis, University of Auckland. https://doi.org/10.13140/2.1.2822.3046\u003c/li\u003e\n\u003cli\u003eMcLeod IM, Parsons DM, Morrison MA, Le Port A, Taylor RB (2012) Factors affecting the recovery of soft-sediment mussel reefs in the Firth of Thames, New Zealand. Mar Freshwater Res 63(1):78. https://doi.org/10.1071/MF11083 \u003c/li\u003e\n\u003cli\u003eMcLeod IM, Parsons DM, Morrison MA, Van Dijken SG, Taylor RB (2014) Mussel reefs on soft sediments: a severely reduced but important habitat for macroinvertebrates and fishes in New Zealand. N Z J Mar Freshw Res 48(1):48\u0026ndash;59. https://doi.org/10.1080/00288330.2013.834831 \u003c/li\u003e\n\u003cli\u003eMcQuaid CD, Lindsay JR (2005) Interacting effects of wave exposure, tidal height and substrate on spatial variation in densities of mussel \u003cem\u003ePerna perna\u003c/em\u003e plantigrades. Mar Ecol Prog Ser 301:173\u0026ndash;184. https://doi.org/10.3354/meps301173 \u003c/li\u003e\n\u003cli\u003eMeadows PS, Meadows A, West FJC, Shand PS, Shaikh MA (1998) Mussels and mussel beds (\u003cem\u003eMytilus edulis\u003c/em\u003e) as stabilizers of sedimentary environments in the intertidal zone. Geological Society, London, Special Publications 139(1):331\u0026ndash;347. https://doi.org/10.1144/GSL.SP.1998.139.01.26\u003c/li\u003e\n\u003cli\u003eMenge BA (1978) Predation intensity in a rocky intertidal community. Oecologia 34:1\u0026ndash;16. https://doi.org/10.1007/BF00346237 \u003c/li\u003e\n\u003cli\u003eMoody JA, Gentry MJ, Bouboulis SA, Kreeger DA (2020) Effects of substrate (protection and type) on ribbed mussel (\u003cem\u003eGeukensia demissa\u003c/em\u003e) recruitment for living shoreline applications. J Coast Res\u003cem\u003e \u003c/em\u003e36(3):619\u0026ndash;627. https://doi.org/10.2112/JCOASTRES-D-19-00062.1 \u003c/li\u003e\n\u003cli\u003eMoreno CA (1995) Macroalgae as a refuge from predation for recruits of the mussel \u003cem\u003eChoromytilus chorus\u003c/em\u003e (Molina, 1782) in Southern Chile. J Exp Mar Bio Ecol 191(2):181\u0026ndash;193. https://doi.org/10.1016/0022-0981(95)00050-2\u003c/li\u003e\n\u003cli\u003eMorrison MA, Lowe M, Parsons D, Usmar N, McLeod I (2009) A review of land-based effects on coastal fisheries and supporting biodiversity in New Zealand. Wellington, New Zealand: Ministry of Fisheries. https://ref.coastalrestorationtrust.org.nz/documents/a-review-of-land-based-effects-on-coastal-fisheries-and-supporting-biodiversity-in-new-zealand-1/ \u003c/li\u003e\n\u003cli\u003eNelson WA (2020) New Zealand seaweeds: an illustrated guide. Te Papa Press. \u003c/li\u003e\n\u003cli\u003eO\u0026rsquo;Connor NE, Crowe TP, McGrath D (2006) Effects of epibiotic algae on the survival, biomass and recruitment of mussels, \u003cem\u003eMytilus\u003c/em\u003e L. (Bivalvia: Mollusca). J Exp Mar Bio Ecol 328(2):265\u0026ndash;276. https://doi.org/10.1016/j.jembe.2005.07.013\u003c/li\u003e\n\u003cli\u003eOverton K, Dempster T, Swearer SE, Morris RL, Barrett LT (2024) Predictors of outplanted marine bivalve survival in restoration: a review and synthesis. J Appl Ecol 61(12):2884\u0026ndash;2896. https://doi.org/10.1111/1365-2664.14795 \u003c/li\u003e\n\u003cli\u003ePaul LJ (2012) A history of the Firth of Thames dredge fishery for mussels: use and abuse of a coastal resource. New Zealand Aquatic Environment and Biodiversity Report No. 94. NIWA. \u003cu\u003ehttps://www.mpi.govt.nz/dmsdocument/4016/direct\u003c/u\u003e\u003c/li\u003e\n\u003cli\u003eProtopopescu GC, Beal BF (2015) Settlement response to various rope substrates in blue mussels (\u003cem\u003eMytilus edulis\u003c/em\u003e Linnaeus) in a hatchery setting. J Shellfish Res 34(2):383\u0026ndash;391. https://doi.org/10.2983/035.034.0221 \u003c/li\u003e\n\u003cli\u003eR Core Team (2024) R: a language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. https://www.R-project.org/\u003c/li\u003e\n\u003cli\u003eReid B (1969) Mussel survey Hauraki Gulf and Firth of Thames. 1958 Fisheries Technical Report No. 34, New Zealand Marine Department, Wellington, New Zealand. \u003c/li\u003e\n\u003cli\u003eReusch TB, Williams SL (1999) Macrophyte canopy structure and the success of an invasive marine bivalve. Oikos 398\u0026ndash;416. https://www.jstor.org/stable/3546420?seq=1andcid=pdf- \u003c/li\u003e\n\u003cli\u003eRowden AA, Berkenbusch K, Brewin PE, Dalen J, Neill KF, Nelson WA, Oliver MD, Probert PK, Schwarz A, Sui PH, Sutherland D (2012) A review of the marine soft-sediment assemblages of New Zealand. Wellington, New Zealand: Ministry for Primary Industries. https://www.researchgate.net/publication/306394493 \u003c/li\u003e\n\u003cli\u003eSchotanus J, Capelle JJ, Paree E, Fivash GS, Van De Koppel J, Bouma TJ (2020) Restoring mussel beds in highly dynamic environments by lowering environmental stressors. Restor Ecol 28(5):1124\u0026ndash;1134. https://doi.org/10.1111/rec.13168\u003c/li\u003e\n\u003cli\u003eSea MA, Hillman JR, Thrush SF (2022) Enhancing multiple scales of seafloor biodiversity with mussel restoration. Sci Rep 12(1):5027. https://doi.org/10.1038/s41598-022-09132-w\u003c/li\u003e\n\u003cli\u003eSeed R, Suchanek TH (1992) Population and community ecology of \u003cem\u003eMytilus. \u003c/em\u003eIn E. Gosling (Ed.), The mussel \u003cem\u003eMytilus\u003c/em\u003e: ecology, physiology, genetics and culture (pp. 87\u0026ndash;169). New York, NY: Elsevier. https://www.scirp.org/reference/referencespapers?referenceid= 1139224\u003c/li\u003e\n\u003cli\u003eSkelton BM, Jeffs AG (2020) The importance of physical characteristics of settlement substrate to the retention and fine-scale movements of \u003cem\u003ePerna canaliculus\u003c/em\u003e spat in suspended longline aquaculture. Aquaculture 521:735054. https://doi.org/10.1016/j.aquaculture.2020.735054 \u003c/li\u003e\n\u003cli\u003eSkelton BM, Jeffs AG (2021) An assessment of the use of macroalgae to improve the retention of Greenshell\u003csup\u003e\u0026trade;\u003c/sup\u003e mussel (\u003cem\u003ePerna canaliculus\u003c/em\u003e) spat in longline culture. Aquac Int 29(4):1683\u0026ndash;1695. https://doi.org/10.1007/s10499-021-00710-9 \u003c/li\u003e\n\u003cli\u003eSuplicy FM (2020) A review of the multiple benefits of mussel farming. Rev Aquacult 12(1):204\u0026ndash;223. https://doi.org/10.1111/raq.12313\u003c/li\u003e\n\u003cli\u003eTemmink RJM, Angelini C, Fivash GS, Swart L, Nouta R, Teunis M, Lengkeek W, Didderen K, Lamers LPM, Bouma TJ, Van Der Heide T (2021) Life cycle informed restoration: engineering settlement substrate material characteristics and structural complexity for reef formation. J Appl Ecol 58(10):2158\u0026ndash;2170. https://doi.org/10.1111/1365-2664.13968\u003c/li\u003e\n\u003cli\u003eTemmink RJM, Fivash GS, Govers LL, Nauta J, Marin-Diaz B, Cruijsen PMJM, Didderen K, Penning E, Olff H, Heusinkveld JHT, Lamers LPM, Lengkeek W, Christianen MJA, Reijers VC, Bouma TJ, Van Der Heide T (2022) Initiating and upscaling mussel reef establishment with life cycle informed restoration: successes and future challenges. Ecol Eng 175:106496. https://doi.org/10.1016/j.ecoleng.2021.106496\u003c/li\u003e\n\u003cli\u003eToone TA, Benjamin ED, Hillman JR, Handley S, Jeffs A (2023a) Multidisciplinary baselines quantify a drastic decline of mussel reefs and reveal an absence of natural recovery. Ecosphere 14(3):e4390. https://doi.org/10.1002/ecs2.4390\u003c/li\u003e\n\u003cli\u003eToone TA, Hillman JR, Benjamin ED, Handley S, Jeffs AG (2023b) Out of their depth: the successful use of cultured subtidal mussels for intertidal restoration. Conserv Sci Prac 5(4):e12914. https://doi.org/10.1111/csp2.12914\u003c/li\u003e\n\u003cli\u003eToone TA, Hillman JR, South PM, Benjamin ED, Handley S, Jeffs AG (2023c) Bottlenecks and barriers: patterns of abundance in early mussel life stages reveal a potential obstacle to reef recovery. Aquat Conserv 33(8):810\u0026ndash;821. https://doi.org/10.1002/aqc.3979\u003c/li\u003e\n\u003cli\u003eToone TA, Hillman JR, Benjamin ED, Handley S, Jeffs AG (2023d) Provision of early mussel life stages via macroalgae enhances recruitment and uncovers a novel restoration technique. J Exp Mar Bio Ecol 566:151919. https://doi.org/10.1016/j.jembe.2023.151919\u003c/li\u003e\n\u003cli\u003evan der Heide T, Tielens E, van der Zee EM, Weerman EJ, Holthuijsen S, Eriksson BK, Piersma T, van de Koppel J, Olff H (2014) Predation and habitat modification synergistically interact to control bivalve recruitment on intertidal mudflats. Biol Conserv 172:163\u0026ndash;169. https://doi.org/10.1016/j.biocon.2014.02.036\u003c/li\u003e\n\u003cli\u003evan der Schatte Olivier A, Jones L, Vay LL, Christie M, Wilson J, Malham SK (2020) A global review of the ecosystem services provided by bivalve aquaculture. Rev Aquacult 12(1):3\u0026ndash;25. https://doi.org/10.1111/raq.12301\u003c/li\u003e\n\u003cli\u003eWalter U, Liebezeit G (2003) Efficiency of blue mussel (\u003cem\u003eMytilus edulis\u003c/em\u003e) spat collectors in highly dynamic tidal environments of the Lower Saxonian coast (southern North Sea). Biomol Eng 20(4\u0026ndash;6):407\u0026ndash;411. https://doi.org/10.1016/S1389-0344(03)00064-9 \u003c/li\u003e\n\u003cli\u003eWesterbom M, Mustonen O, Kilpi M (2008) Distribution of a marginal population of \u003cem\u003eMytilus edulis\u003c/em\u003e: responses to biotic and abiotic processes at different spatial scales. Mar Biol 153(6):1153\u0026ndash;1164. https://doi.org/10.1007/s00227-007-0886-7\u003c/li\u003e\n\u003cli\u003eWilcox M, Jeffs A (2017) Is attachment substrate a prerequisite for mussels to establish on soft-sediment substrate? J Exp Mar Bio Ecol 495:83\u0026ndash;88. https://doi.org/10.1016/j.jembe.2017.07.004\u003c/li\u003e\n\u003cli\u003eWilcox M, Kelly S, Jeffs A (2018) Ecological restoration of mussel beds onto soft-sediment using transplanted adults. Restor Ecol 26(3):581\u0026ndash;590. https://doi.org/10.1111/rec.12607\u003c/li\u003e\n\u003cli\u003eWu W (2018) The role of physical structure in the attachment of juvenile green-lipped mussels.\u003cem\u003e \u003c/em\u003eMaster\u0026rsquo;s Thesis, University of Auckland. https://hdl.handle.net/2292/45020\u003c/li\u003e\n\u003cli\u003eWu W, Jeffs AG (2025) Influence of microstructure of substrate surface on the attachment of juvenile mussels. Fishes 10(3):135. https://doi.org/10.3390/fishes10030135 \u003c/li\u003e\n\u003cli\u003eWu W, Anderson I, Jeffs AG (2025) The role of substrates width and millimeter scale surface micro-structure in the attachment of juvenile mussels. Aquaculture 595:741593. https://doi.org/10.1016/j.aquaculture.2024.741593 \u003c/li\u003e\n\u003cli\u003eYang JL, Satuito CG, Bao WY, Kitamura H (2007) Larval settlement and metamorphosis of the mussel \u003cem\u003eMytilus galloprovincialis\u003c/em\u003e on different macroalgae. Mar Biol 152:1121\u0026ndash;1132. https://doi.org/10.1007/s00227-007-0759-0 \u003c/li\u003e\n\u003cli\u003eZhao S, Zhang B, Yang J, Zhou J, Xu Y (2024) Linear discriminant analysis. Nature Reviews Methods Primers 4(1):70. \u003cu\u003ehttps://doi.org/10.1038/s43586-024-00346-y\u003c/u\u003e\u003c/li\u003e\n\u003c/ol\u003e"},{"header":"Tables","content":"\u003cp\u003eTables 1 - 4 are available in the Supplementary Files section.\u003c/p\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-6745848/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6745848/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eUnderstanding and addressing substrate availability has been shown to be critical for facilitating recruitment in shellfish reef restoration initiatives globally. For many mussel species, macroalgae is vital for the settlement and establishment of juveniles on reefs, but the factors which determine a macroalga\u0026rsquo;s effectiveness as an attachment substrate are poorly understood. This study aimed to; 1) develop an index that can be used by restoration practitioners to score the morphological features of macroalgae for their potential to support juvenile mussel recruitment and 2) test the accuracy of this index on the macroalgal substrate attachments of different juvenile size classes of the green-lipped mussel \u003cem\u003ePerna canaliculus\u003c/em\u003e on two remnant mussel reefs in northeast New Zealand. Eight morphological features of macroalgae, identified from published studies, were used to create the Macroalgal Morphology Index (MMI). The index scoring criteria for each of these features were able to predict the likelihood of \u003cem\u003eP. canaliculus\u003c/em\u003e attachments to macroalgae for juveniles\u0026thinsp;\u0026lt;\u0026thinsp;10 mm in shell length with 75% accuracy but with only 40\u0026ndash;60% accuracy for juveniles 10 \u0026ndash; \u0026lt;30 mm. Holdfast complexity and canopy cover were the two most useful features of macroalgae for predicting the attachment of all sizes of juvenile mussels. Meanwhile, the four features of macroalgae that describe their branching morphology were only strong predictors of attachment for juveniles\u0026thinsp;\u0026lt;\u0026thinsp;10 mm. Overall, these findings suggest that the MMI can aid restoration practitioners in the selection of suitable macroalgal substrates for facilitating juvenile recruitment at mussel reef restoration sites in New Zealand and potentially elsewhere.\u003c/p\u003e","manuscriptTitle":"The morphology of macroalgal substrates can help predict the attachment of juvenile mussels","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-06-09 09:02:25","doi":"10.21203/rs.3.rs-6745848/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"98af38d5-c3f6-4691-9125-7e1450febc80","owner":[],"postedDate":"June 9th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2025-06-28T20:56:19+00:00","versionOfRecord":[],"versionCreatedAt":"2025-06-09 09:02:25","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-6745848","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6745848","identity":"rs-6745848","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}