Correlation of the Structural Characteristics of an Artificial Oyster Reef with Its Wake Region

Abstract

Oyster reefs are currently at risk of severe decline due to dangerous human interference and its aftermath; hence, artificial oyster reefs (AORs) have been utilized for their restoration. AORs with high vertical reliefs interact with the surrounding flow, constitute a reverse flow, and create a wake region in which concentrated nutrients and food organisms exist. However, the correlations of the structural characteristics of an AOR with its wake regions have not been studied. Thus, we established 96 AOR models, carried out flow analyses, and obtained their wake volumes, considering shell orientation, composition, penetration depth, and growth stage. We found that the growth stage is the most critical parameter for establishing a normalized wake volume. This implies that the number of oyster shells (N) is the most critical factor in securing a normalized wake volume, in which their correlation was linear and significant (R2=0.89). We also found that the correlations of the normalized wake volume with blocking and surface complexity indices were linearly significant, respectively. Additionally, wake volume efficiency increased with the number of oyster shells; specifically, the criterion for wake volume efficiency of EI (efficiency index) ≥ 2.0 was satisfied when N≥50 per 900 cm2.

Full text

Citation: Kim, M.; Jung, S.; Chau,
T.V.; Na, W.-B. Correlation of the
Structural Characteristics of an
Artificial Oyster Reef with Its Wake
Region. J. Mar. Sci. Eng. 2023,11, 775.
https://doi.org/10.3390/
jmse11040775
Academic Editor: Tom Spencer
Received: 5 March 2023
Revised: 26 March 2023
Accepted: 29 March 2023
Published: 3 April 2023
Copyright: © 2023 by the authors.
Licensee MDPI, Basel, Switzerland.
This article is an open access article
distributed under the terms and
conditions of the Creative Commons
Attribution (CC BY) license (https://
creativecommons.org/licenses/by/
4.0/).
Journal of
Marine Science
and Engineering
Article
Correlation of the Structural Characteristics of an Artificial
Oyster Reef with Its Wake Region
Minju Kim, Somi Jung , Than Van Chau and Won-Bae Na *
Department of Ocean Engineering, Pukyong National University, Busan 48513, Republic of Korea;
202155161@pukyong.ac.kr (M.K.); smjung@pknu.ac.kr (S.J.); 202156806@pukyong.ac.kr (T.V.C.)
*Correspondence: wna@pknu.ac.kr; Tel.: +82-51-629-6588
Abstract:
Oyster reefs are currently at risk of severe decline due to dangerous human interference
and its aftermath; hence, artificial oyster reefs (AORs) have been utilized for their restoration. AORs
with high vertical reliefs interact with the surrounding flow, constitute a reverse flow, and create a
wake region in which concentrated nutrients and food organisms exist. However, the correlations
of the structural characteristics of an AOR with its wake regions have not been studied. Thus, we
established 96 AOR models, carried out flow analyses, and obtained their wake volumes, considering
shell orientation, composition, penetration depth, and growth stage. We found that the growth stage
is the most critical parameter for establishing a normalized wake volume. This implies that the
number of oyster shells (
N
) is the most critical factor in securing a normalized wake volume, in which
their correlation was linear and significant (
R2=
0.89). We also found that the correlations of the
normalized wake volume with blocking and surface complexity indices were linearly significant,
respectively. Additionally, wake volume efficiency increased with the number of oyster shells;
specifically, the criterion for wake volume efficiency of EI (efficiency index)
≥
2.0 was satisfied when
N≥50 per 900 cm2.
Keywords: flow analysis; oyster reefs; structural characteristics; wake region; wake volume
1. Introduction
Oysters form a colony attached to a hard substrate (e.g., rock); such an oyster colony is
often called an oyster reef (OR) [
1
]. ORs provide various ecosystem services. These services
include: a space with crevices and substrates for fish, crustaceans, shellfish, seagrass, and
seaweeds; an excellent attachment space for the survival of marine life; and a shelter and
spawning ground for various marine organisms to rest [
2
–
5
]. ORs also filter seawater to
stabilize water quality and reduce wave energy to protect coastlines [6,7].
Such a three-dimensional physical structure formed in the coastal region interacts
with currents and waves and generates turbulence in the surrounding flow region [
8
,
9
].
ORs therefore promote the movement of nutrients with surrounding sediments and have
significant effects on marine organisms’ settlement, food utilization, and survival [
9
–
11
].
Such a positive impact of ORs is also explained by marine habitat complexity (or substrate
diversity) [
12
]. The habitat heterogeneity hypothesis posits that structurally complex
habitats provide many niches and multiple ways to utilize environmental resources and
can increase species diversity; this hypothesis has been validated on numerous occasions
(e.g., [13]).
Oyster populations have declined rapidly over the centuries. This is due to dangerous
human-initiated interference and its aftermath, such as overfishing, seawater pollution,
habitat destruction, and global warming [
3
]. Over the past few years, oyster habitats have
declined by ~63% in the United States [
14
] and ~85% of ORs have been lost globally [
15
].
To address this population decline, oyster habitat restoration projects using artificial oyster
reefs (AORs) are ongoing in Australia, Europe, and the United States [
3
,
16
–
21
]. These
J. Mar. Sci. Eng. 2023,11, 775. https://doi.org/10.3390/jmse11040775 https://www.mdpi.com/journal/jmse

J. Mar. Sci. Eng. 2023,11, 775 2 of 16
projects are based on scientific observations that artificially installed ORs can successfully
grow and perform the functions of natural ORs [
2
,
3
,
7
,
15
]. AORs are known to exhibit the
flow conditions and characteristics of natural ORs within 1 year after installation [22,23].
The substrates for AORs can be classified into two types [
24
–
26
]. The first type utilizes
natural materials on which oyster reefs accrete, e.g., recycled, fossilized, and dredged native
oyster shells. However, these materials are often limited due to the increased demand
for OR restoration projects [
27
], and their use may be unavailable or undesirable due to
the biosecurity regulation on dissolution of calcium carbonate shell bases in acidifying
oceans [
15
]. Accordingly, the second type includes crushed limestone or rock, other bivalve
shells, standard concrete, concrete with various additives, and biodegradable products. To
secure the structural stability of the AORs made from these materials, various fastening
methods, such as mesh nets, are used [
7
]. AORs installed with a relatively flat relief do
not dramatically change the surrounding flow. On the other hand, AORs installed with
a relatively high relief interact with the surrounding flow to constitute a reverse flow for
creating a wake region. Here, the wake region refers to the recirculating flow area where
concentrated nutrients and food organisms exist; as such, this is also home to a variety
of organisms [
28
,
29
]. Therefore, when installing AORs with a relatively high relief, it
is necessary to understand how the structural characteristics of the AORs affects their
wake regions, as well as how oyster reef growth (or recruitment) affects the wake region.
However, little research is available on these aspects of AORs.
In this research, we propose the hypothesis that “a unique wake region is formed
around an AOR due to its structural characteristics”. Detailed research questions to test
the hypothesis are as follows. Do the structural characteristics (e.g., composition) of oyster
shells used to construct an AOR have an effect on its wake region? What is the correlation
between the structural characteristics of an AOR and its wake region? To answer these
questions, we constructed a total of 96 AOR models, including 48 for the initial stage and
48 for the growth stages. Then, we used computational fluid dynamics to obtain the wake
region characteristics and utilized wake volumes for their quantification. We also utilized
evaluation indices such as the efficiency index (EI) [
28
], an improved blocking index, and
a surface complexity index to further correlate the structural characteristics to the wake
region characteristics.
2. Materials and Methods
2.1. Artificial Oyster Reef Models
The oyster shell model considered in this study was adopted from the oyster species
called Crassostrea virginica (C. virginica), which inhabits the east coasts of North and
South America [
30
]. This selection was made due to the habitat restoration projects for
the oyster species in these areas. The maximum size of the species reaches ~14.7 cm
(length) ×6.8 cm (width) ×
4.3 cm (thickness), and the shell has a sharp layered surface
with irregularities [
30
]. We idealized the shell as shown in Figure 1by selecting a medium
size of 5 cm (length)
×
4 cm (width)
×
0.1 cm (thickness) with a smooth surface and
combining two ellipses.
J. Mar. Sci. Eng. 2023, 11, x FOR PEER REVIEW 3 of 16
Figure 1. Schematic diagram of oyster shells considered in this study.
Figure 2. Schematic diagrams of the initial artificial oyster reef (AOR) model (not to scale).
Table 1. The initial artificial oyster reef (AOR) model. A total of 48 initial AOR models were con-
structed by considering four shell orientations, six compositions, and two shell penetration depths.
Shell Orientation
Shell Composition
Number of Oysters Penperdicu-
lar to Flow Direction (Interval)
Number of Oysters in Flow Di-
rection (Interval)
Convex
Concave
Mixed 1
Mixed 2
3 (12 cm) 1 (–)
3 (12 cm) 3 (12 cm)
3 (12 cm) 5 (6 cm)
5 (6 cm) 1 (–)
5 (6 cm) 3 (12 cm)
5 (6 cm) 5 (6 cm)
Figure 1. Schematic diagram of oyster shells considered in this study.

J. Mar. Sci. Eng. 2023,11, 775 3 of 16
Oysters generally grow faster when living on muddy bottoms, but they are fragile
and easily destroyed [
30
]. We therefore first constructed an initial AOR model by fixing
the shells to a flat substrate of 0.3 m (length)
×
0.3 m (width)
×
0.05 m (height) (Figure 2).
We then considered four shell orientations such as convex, concave, mixed 1, and mixed 2
and six compositions such as 3
×
1, 3
×
3, 3
×
5, 5
×
1, 5
×
3, and 5
×
5 (Figure 2); here,
the first and second numbers indicate the number of oyster shells perpendicular to and
parallel to the inlet flow direction, respectively (Table 1). Moreover, we considered two
shell penetration depths of 10% and 50%; hence, the 48 initial AOR models were designed
as illustrated in Figure 2.
J. Mar. Sci. Eng. 2023, 11, x FOR PEER REVIEW 3 of 16
Figure 1. Schematic diagram of oyster shells considered in this study.
Figure 2. Schematic diagrams of the initial artificial oyster reef (AOR) model (not to scale).
Table 1. The initial artificial oyster reef (AOR) model. A total of 48 initial AOR models were con-
structed by considering four shell orientations, six compositions, and two shell penetration depths.
Shell Orientation
Shell Composition
Number of Oysters Penperdicu-
lar to Flow Direction (Interval)
Number of Oysters in Flow Di-
rection (Interval)
Convex
Concave
Mixed 1
Mixed 2
3 (12 cm) 1 (–)
3 (12 cm) 3 (12 cm)
3 (12 cm) 5 (6 cm)
5 (6 cm) 1 (–)
5 (6 cm) 3 (12 cm)
5 (6 cm) 5 (6 cm)
Figure 2. Schematic diagrams of the initial artificial oyster reef (AOR) model (not to scale).
Oysters grow in a dense ecological space and agglomerate as an oyster reef. Such char-
acteristics create a large number of crevices inside an oyster reef, increasing the substrate
surface on which the larvae can settle [
31
]. Moreover, the flow velocity in such a crevice
is small compared to the external flow space, providing a survival space for living organ-
isms [
11
]. However, it can be difficult to account for these geometric characteristics in an
AOR model because ORs grow without any apparent regularity. Therefore, we considered
16 growth stages (GS in figures) of AORs by controlling the number of oyster shells (
N
)
from 50 to 200 (Figure 3). Some of the shells were randomly inserted into the substrate to
partly reflect the initial models; notably, their penetration depths did not exceed 50%. For
each growth (or recruitment) stage, three representative models were made by randomly
arranging the oysters; hence, a total of 48 growth AOR models were constructed (Figure 3,
Table 2).

J. Mar. Sci. Eng. 2023,11, 775 4 of 16
Table 1.
The initial artificial oyster reef (AOR) model. A total of 48 initial AOR models were
constructed by considering four shell orientations, six compositions, and two shell penetration
depths.
Shell Orientation
Shell Composition
Number of Oysters Penperdicular to Flow Direction (Interval) Number of Oysters in Flow Direction (Interval)
Convex
Concave
Mixed 1
Mixed 2
3 (12 cm) 1 (–)
3 (12 cm) 3 (12 cm)
3 (12 cm) 5 (6 cm)
5 (6 cm) 1 (–)
5 (6 cm) 3 (12 cm)
5 (6 cm) 5 (6 cm)
J. Mar. Sci. Eng. 2023, 11, x FOR PEER REVIEW 4 of 16
Oysters grow in a dense ecological space and agglomerate as an oyster reef. Such
characteristics create a large number of crevices inside an oyster reef, increasing the sub-
strate surface on which the larvae can sele [31]. Moreover, the flow velocity in such a
crevice is small compared to the external flow space, providing a survival space for living
organisms [11]. However, it can be difficult to account for these geometric characteristics
in an AOR model because ORs grow without any apparent regularity. Therefore, we con-
sidered 16 growth stages (GS in figures) of AORs by controlling the number of oyster
shells () from 50 to 200 (Figure 3). Some of the shells were randomly inserted into the
substrate to partly reflect the initial models; notably, their penetration depths did not ex-
ceed 50%. For each growth (or recruitment) stage, three representative models were made
by randomly arranging the oysters; hence, a total of 48 growth AOR models were con-
structed (Figure 3, Table 2).
Figure 3. Schematic diagrams of the 48 growth AOR models from growth stage 1 (GS 1) to 16 (GS
16) (not to scale).
Figure 3.
Schematic diagrams of the 48 growth AOR models from growth stage 1 (GS 1) to 16 (GS 16)
(not to scale).

J. Mar. Sci. Eng. 2023,11, 775 5 of 16
Table 2. Growth stages of the growth AOR models (to be continued).
Growth Stage
GS
Number of
Shells N
Width
W(cm)
Length L
(cm)
Height H
(cm)
Volume
(cm3)
Total
Surface
Area (cm2)
Projection
Area (cm2)
1
Rep. 1 50 30.66 30.55 9.97 4623.37 4906.79 279.60
Rep. 2 50 30.35 30.00 9.80 4626.75 4978.35 281.46
Rep. 3 50 30.00 30.23 8.87 4590.10 4206.49 241.82
2
Rep. 1 60 30.66 30.55 13.45 4649.25 5225.37 355.33
Rep. 2 60 30.35 30.00 14.50 4652.93 5511.11 358.63
Rep. 3 60 30.40 30.42 12.10 4613.75 4678.59 276.62
3
Rep. 1 70 30.66 30.78 13.45 4674.13 5926.94 360.21
Rep. 2 70 30.69 30.00 14.50 4678.69 6035.90 383.25
Rep. 3 70 30.40 30.42 12.12 4646.65 5156.80 315.48
4
Rep. 1 80 30.66 30.78 16.40 4694.84 6344.33 389.63
Rep. 2 80 31.51 30.00 17.50 4704.36 6553.99 425.60
Rep. 3 80 30.63 30.42 12.12 4671.32 5646.30 339.27
5
Rep. 1 90 30.94 31.24 16.80 4720.66 6862.75 417.85
Rep. 2 90 31.51 31.84 17.69 4730.41 7075.31 436.84
Rep. 3 90 32.51 30.42 12.83 4697.92 6181.34 362.12
6
Rep. 1 100 30.94 33.37 18.42 4744.21 7339.73 443.11
Rep. 2 100 31.51 31.84 17.71 4756.25 7594.81 463.90
Rep. 3 100 32.51 31.47 12.83 4725.18 6672.37 364.56
7
Rep. 1 110 32.42 33.37 19.29 4770.46 7863.62 482.67
Rep. 2 110 31.51 32.34 20.05 4781.96 8113.50 503.89
Rep. 3 110 32.51 31.90 16.01 4751.47 7201.67 386.06
8
Rep. 1 120 32.42 33.37 21.81 4796.22 8383.68 529.98
Rep. 2 120 31.51 32.34 22.07 4808.13 8644.26 517.64
Rep. 3 120 32.73 34.51 16.01 4778.98 7706.13 397.70
9
Rep. 1 130 32.42 33.37 21.81 4820.80 8891.89 530.86
Rep. 2 130 31.51 32.34 22.07 4833.87 9156.26 545.31
Rep. 3 130 32.73 34.51 16.06 4805.34 8234.50 419.21
10
Rep. 1 140 33.87 33.37 21.81 4846.33 9401.96 538.54
Rep. 2 140 32.15 34.55 22.07 4859.06 9666.91 560.63
Rep. 3 140 32.73 34.51 16.06 4835.33 8680.43 440.96
11
Rep. 1 150 33.87 34.52 21.81 4870.18 9875.24 552.31
Rep. 2 150 32.15 35.80 22.07 4882.69 10,134.06 572.38
Rep. 3 150 32.73 34.76 19.40 4862.30 9217.09 471.87
12
Rep. 1 160 33.87 34.52 21.81 4896.19 10,397.32 561.01
Rep. 2 160 32.43 35.80 22.07 4908.55 10,655.81 581.27
Rep. 3 160 32.73 34.76 19.40 4892.84 9702.75 502.82
13
Rep. 1 170 33.87 35.42 21.81 4922.08 10,910.30 573.46
Rep. 2 170 32.43 35.80 22.07 4934.32 11,171.07 593.24
Rep. 3 170 33.61 36.36 19.67 4919.47 10,241.49 530.95

J. Mar. Sci. Eng. 2023,11, 775 6 of 16
Table 2. Cont.
Growth Stage
GS
Number of
Shells N
Width
W(cm)
Length L
(cm)
Height H
(cm)
Volume
(cm3)
Total
Surface
Area (cm2)
Projection
Area (cm2)
14
Rep. 1 180 33.87 35.70 21.81 4947.57 11,420.00 592.16
Rep. 2 180 32.43 35.80 23.78 4959.01 11,666.11 613.22
Rep. 3 180 33.61 37.47 22.52 4950.13 10,738.68 564.82
15
Rep. 1 190 33.99 35.70 22.69 4972.73 11,923.42 613.67
Rep. 2 190 33.94 35.80 26.25 4984.80 12,188.37 658.90
Rep. 3 190 33.97 37.47 22.52 4976.82 11,272.94 589.31
16
Rep. 1 200 35.07 35.70 22.69 4998.45 12,440.46 643.39
Rep. 2 200 33.94 35.80 26.25 5010.37 12,705.07 666.88
Rep. 3 200 33.97 37.47 24.59 5006.37 11,768.37 620.63
It should be noted that the current 48 growth AOR models have certain limitations.
First, the oyster shells modeled had the same size (Figure 1), although their orientation, com-
position, and penetration depth were randomly selected for the 48 growth models (
Figure 3
).
In other words, we modeled the growth stages not by increasing the shell size but by adding
oyster shells. Accordingly, a positive reef accretion rate was not fully implemented in this
study, although there are some relevant studies (e.g.,
7.0–16.9 mm year−1[10]
). Second, the
plate substrate remained the same size (0.3 m (length)
×
0.3 m (width)
×
0.05 m (height)) for
all models. This design was intentionally made to allow for an investigation of the effect of
the number of shells on the normalized wake volume and the associated evaluation indices.
Third, live oysters were not considered in the models. Oysters with the potential to settle
and grow on the AORs were modeled as single shell segments rather than closed bivalves.
2.2. Flow Analysis
Using the 96 AOR models, flow analyses were performed using ANSYS-CFX [
32
],
a software package for flow analysis utilizing the element-based finite volume method
(EbFVM). The numerical method has been used to obtain wake volumes of artificial reefs
on several occasions [
28
,
29
,
33
], mainly because it has the advantage of relatively easy
mesh generation [
28
], taking the merits of both the finite element method and the finite
volume method. For example, the EbFVM has been used to calculate wake volumes of
various artificial reefs and to subsequently evaluate their efficiency, tranquility, and stability
indices [
28
], efficient placement models [
29
], and flatly distributed placement models [
33
].
The governing equation is the Reynolds Averaged Navier Stokes (RANS) equation, and the
turbulence equation uses the Shear Stress Transport (SST) turbulence model to improve the
predictability of the flow separation on smooth surfaces and prevent excessive prediction
of eddy viscosity [34,35].
The size of the flow field was determined to be 3 m (length)
×
3 m (width)
×
0.5 m
(height), a sufficient space to prevent the pressure gradient from occurring at the boundary
(Figure 4a). The front surface was set as an inlet with a velocity of 1 m s
−1
, the rear surface
was set as an outlet so that all the flows could escape, the bottom surface and the surface of
the AORs were set as a no-slip wall, and all of the other surfaces were set to be symmetric
so as not to affect the analysis (Figure 4b).

J. Mar. Sci. Eng. 2023,11, 775 7 of 16
J. Mar. Sci. Eng. 2023, 11, x FOR PEER REVIEW 7 of 16
Figure 4. Flow domain (a) and boundary conditions (b) showing inlet (b1), outlet (b2), boom (b3),
and symmetry (b4).
Figure 5. Mesh independence: (a) drag coefficient and (b) wake volume.
Figure 4.
Flow domain (
a
) and boundary conditions (
b
) showing inlet (
b1
), outlet (
b2
), bottom (
b3
),
and symmetry (b4).
Each AOR model had curved shapes, sharp edges, and irregular surfaces. To reflect
these characteristics, all meshes of the flow field were created as tetrahedrons. This is
because the tetrahedron meshes can be easily generated, even in a structure having a
complex shape, and can yield improved mesh quality. To increase the accuracy of the flow
analyses, the mesh quality was checked for all of the models. As a result, the initial AOR
models all satisfied the orthogonal quality (minimum 0.15; i.e., more than acceptable) and
skewness quality (maximum 0.94; i.e., more than acceptable) [
32
]. However, due to various
crevices, it was difficult for all of the growth AOR models to satisfy the orthogonal and
skewness qualities. Thus, we improved the mesh quality of all the growth models until
the percentage of meshes satisfying the mesh quality reached 99.995%. We also considered
the number of elements in the flow field by adjusting several mesh sizes and examined the
mesh independence of the plate model using drag coefficients and wake volumes (Figure 5).
The drag coefficients converged with an average of 0.79 for
≥
20 million elements, and the
wake volumes converged to ~1472 cm
3
for
≥
40 million elements. As our key focus was
to identify the wake volumes of the AOR models, all of the analyses were performed to
satisfy the number of elements of 40 million or more.
J. Mar. Sci. Eng. 2023, 11, x FOR PEER REVIEW 7 of 16
Figure 4. Flow domain (a) and boundary conditions (b) showing inlet (b1), outlet (b2), boom (b3),
and symmetry (b4).
Figure 5. Mesh independence: (a) drag coefficient and (b) wake volume.
Figure 5. Mesh independence: (a) drag coefficient and (b) wake volume.

J. Mar. Sci. Eng. 2023,11, 775 8 of 16
2.3. Wake Volume and Evaluation Indices
The wake region, an important flow characteristic in artificial reefs research, is a space
in a region downstream of the artificial reef where recirculating flow occurs (Figure 6a).
This region has a significantly reduced flow velocity compared to the external flow velocity
field and contains a large amount of nutrients from the seabed due to the relatively stagnant
recirculation flow. For this reason, many marine organisms utilize the wake region as a
resting, spawning, and/or feeding ground [
36
–
39
]. Various measures such as wake area
and wake length have been used to quantify the wake region [
40
]. However, wake area
and length cannot capture the three-dimensional characteristics of the wake region as these
values depend on the selection of a reference plane. To solve this problem, we used the
wake volume concept proposed by Kim et al. [
41
], utilizing the EbFVM, as illustrated
Figure 6b. The wake volume of each model was obtained as follows. First, each element
was regarded as a finite wake volume if water flows at the nodes were all recirculating
water flows. Second, the total wake volume was obtained by collecting the finite wake
volumes (or elements). Once the wake volumes (
WV AOR
) of the 96 AOR models were
obtained, we normalized these volumes with respect to the wake volume of the plate-only
model (WV plate), such that WV =WVAOR /WV plate.
J. Mar. Sci. Eng. 2023, 11, x FOR PEER REVIEW 8 of 16
2.3. Wake Volume and Evaluation Indices
The wake region, an important flow characteristic in artificial reefs research, is a space
in a region downstream of the artificial reef where recirculating flow occurs (Figure 6a).
This region has a significantly reduced flow velocity compared to the external flow veloc-
ity field and contains a large amount of nutrients from the seabed due to the relatively
stagnant recirculation flow. For this reason, many marine organisms utilize the wake re-
gion as a resting, spawning, and/or feeding ground [36–39]. Various measures such as
wake area and wake length have been used to quantify the wake region [40]. However,
wake area and length cannot capture the three-dimensional characteristics of the wake
region as these values depend on the selection of a reference plane. To solve this problem,
we used the wake volume concept proposed by Kim et al. [41], utilizing the EbFVM, as
illustrated Figure 6b. The wake volume of each model was obtained as follows. First, each
element was regarded as a finite wake volume if water flows at the nodes were all recir-
culating water flows. Second, the total wake volume was obtained by collecting the finite
wake volumes (or elements). Once the wake volumes (
) of the 96 AOR models were
obtained, we normalized these volumes with respect to the wake volume of the plate-only
model (
), such that 





=
 

⁄.
Figure 6. Illustration of the wake region of an AOR model (the number of shells =50): (a) velocity
vectors showing recirculating flow and (b) views of the wake volume (orange color).
The structural characteristics of submerged structures influence the surrounding
flow characteristics [28,42]. Consequently, we established evaluation indices to correlate
the structural characteristics with the flow characteristics of the AOR models. First, we
used the EI proposed by Kim et al. [28] to quantify the efficiency of the wake volume
(
) of the 96 AOR models with respect to their structural volumes (
), i.e., EI =

 

⁄. According to Kim et al. [28], the wake volume efficiency is regarded to be
excellent if EI ≥2, indicating that the wake volume is more than twice the structural vol-
ume. Second, we proposed the improved blocking index (hereinafter, BI), reflecting the
blocking index conceptually stated by Kim et al. [43] and clearly proposed by Jung et al.
[44]. The original blocking index () is defined by the ratio of the actual front surface area
() to the blocking front surface area ( ), of which the front part of the structure is
completely blocked, i.e., =
⁄. Here, the front surface is the one perpendicular to
the inlet flow. This index has been used to quantify the water blocking effect of the front
surface area on the wake volume [44]. However, considering the structural characteristics
Figure 6.
Illustration of the wake region of an AOR model (the number of shells
N=
50): (
a
) velocity
vectors showing recirculating flow and (b) views of the wake volume (orange color).
The structural characteristics of submerged structures influence the surrounding flow
characteristics [
28
,
42
]. Consequently, we established evaluation indices to correlate the
structural characteristics with the flow characteristics of the AOR models. First, we used the
EI proposed by Kim et al. [
28
] to quantify the efficiency of the wake volume (
WV AOR
) of the
96 AOR models with respect to their structural volumes (
VAOR
), i.e.,
EI =WV AOR /VAOR
.
According to Kim et al. [
28
], the wake volume efficiency is regarded to be excellent if
EI ≥
2, indicating that the wake volume is more than twice the structural volume. Second,
we proposed the improved blocking index (hereinafter, BI), reflecting the blocking index
conceptually stated by Kim et al. [
43
] and clearly proposed by Jung et al. [
44
]. The original
blocking index (
φb
) is defined by the ratio of the actual front surface area (
Af
) to the
blocking front surface area (
Af b
), of which the front part of the structure is completely
blocked, i.e.,
φb=Af/Af b
. Here, the front surface is the one perpendicular to the inlet
flow. This index has been used to quantify the water blocking effect of the front surface
area on the wake volume [
44
]. However, considering the structural characteristics of the
AOR models (e.g., irregular aggregations and crevices), the blocking index is not easy to

J. Mar. Sci. Eng. 2023,11, 775 9 of 16
simulate. Thus, we devised an improved
BI
defined by the ratio of the projection area
(
Ap
) of each AOR in the inlet flow direction to the front curtain area (
Ac
), i.e.,
BI =Ap/Ac
.
Here, the front curtain area is defined by the product of the reef width (
W
) and the water
depth (
D
), i.e.,
Ac=W×D
. Third, we used the surface complexity index (hereinafter, SI),
which is the ratio of the total surface area (
At
) of each AOR model to its bottom projection
area (
Ab
), i.e.,
φs=At/Ab
. The bottom projection area is defined by the product of the
reef length (
L
) and the reef width (
W
), i.e.,
Ab=L×W
. Accordingly, the surface complex
index is a surface area-based complexity measure and is similar to the one proposed by
Jung et al. [44].
3. Results and Discussion
3.1. Initial AOR Model
Figure 7a shows the wake volumes of the 48 initial AOR models when the shell penetra-
tion depth was 10%. The wake volumes of the 24 models (i.e.,
4 orientations ×6 compositions
)
ranged from 2664–9943 cm
3
with an average of 5310 cm
3
. These wake volumes were
1.83–6.84 times
higher than that of the plate model (1454 cm
3
). Figure 7b shows the 24 wake
volumes when the shell penetration depth was 50%. The range of the wake volumes was
2036–5204 cm
3
(average: 3431 cm
3
), 1.40–3.58 times higher than that of the plate model.
Our results showed that the high-relief reef models (i.e., 10% penetration depth) resulted
in larger wake volumes than the low-relief reef models (i.e., 50% penetration depth). This
observation becomes clearer when we compare the wake volumes of the 48 initial AOR
models with that of the plate-only model. Considering the shell orientations, the concave
orientation resulted in larger wake volumes, particularly when the shell compositions
were 5
×
5 and 5
×
3 for both shell penetration depths. Accordingly, the two largest wake
volumes of 9943 and 9297 cm
3
were generated when the shell penetration depth was 10%, the
shell orientation was concave, and the shell compositions were 5 ×5 and 5 ×3, respectively.
J. Mar. Sci. Eng. 2023, 11, x FOR PEER REVIEW 9 of 16
of the AOR models (e.g., irregular aggregations and crevices), the blocking index is not
easy to simulate. Thus, we devised an improved BI defined by the ratio of the projection
area () of each AOR in the inlet flow direction to the front curtain area (), i.e., BI =

⁄. Here, the front curtain area is defined by the product of the reef width () and the
water depth (), i.e., =×. Third, we used the surface complexity index (hereinaf-
ter, SI), which is the ratio of the total surface area () of each AOR model to its boom
projection area (), i.e., =
⁄. The boom projection area is defined by the product
of the reef length () and the reef width (), i.e., =×. Accordingly, the surface
complex index is a surface area-based complexity measure and is similar to the one pro-
posed by Jung et al. [44].
3. Results and Discussion
3.1. Initial AOR Model
Figure 7a shows the wake volumes of the 48 initial AOR models when the shell pen-
etration depth was 10%. The wake volumes of the 24 models (i.e., 4 orientations × 6 com-
positions) ranged from 2664–9943 cm3 with an average of 5310 cm3. These wake volumes
were 1.83–6.84 times higher than that of the plate model (1454 cm3). Figure 7b shows the
24 wake volumes when the shell penetration depth was 50%. The range of the wake vol-
umes was 2036–5204 cm3 (average: 3431 cm3), 1.40–3.58 times higher than that of the plate
model. Our results showed that the high-relief reef models (i.e., 10% penetration depth)
resulted in larger wake volumes than the low-relief reef models (i.e., 50% penetration
depth). This observation becomes clearer when we compare the wake volumes of the 48
initial AOR models with that of the plate-only model. Considering the shell orientations,
the concave orientation resulted in larger wake volumes, particularly when the shell com-
positions were 5 × 5 and 5 × 3 for both shell penetration depths. Accordingly, the two
largest wake volumes of 9943 and 9297 cm3 were generated when the shell penetration
depth was 10%, the shell orientation was concave, and the shell compositions were 5 × 5
and 5 × 3, respectively.
Figure 7. Wake volumes of the initial AOR models: (a) shell penetration depth of 10%; (b) shell pen-
etration depth of 50%.
Figure 7.
Wake volumes of the initial AOR models: (
a
) shell penetration depth of 10%; (
b
) shell
penetration depth of 50%.

J. Mar. Sci. Eng. 2023,11, 775 10 of 16
3.2. Growth AOR Models
Figure 8a shows the correlation between the growth stages of the AOR models and
their wake volumes. The wake volume increased as the growth stage increased, and the
correlation coefficient was
R2=
0.98 when considering the average values. For example,
the average wake volumes were 9063 and 32,226 cm
3
for the 1st growth stage (50 oyster
shells) and the 16th growth stage (200 oyster shells), respectively, indicating an increase
of ~3.56 times; these two wake volumes (GS = 1 and GS = 16) are shown in Figure 8b.
J. Mar. Sci. Eng. 2023, 11, x FOR PEER REVIEW 10 of 16
3.2. Growth AOR Models
Figure 8a shows the correlation between the growth stages of the AOR models and
their wake volumes. The wake volume increased as the growth stage increased, and the
correlation coefficient was =0.98 when considering the average values. For example,
the average wake volumes were 9063 and 32,226 cm3 for the 1st growth stage (50 oyster
shells) and the 16th growth stage (200 oyster shells), respectively, indicating an increase
of ~3.56 times; these two wake volumes (GS = 1 and GS = 16) are shown in Figure 8b.
Figure 9a shows the normalized wake volumes (





) of the AOR models and their
correlation with the numbers of oyster shells () introduced to the models. According to
the correlation coefficient of =0.89, the positive correlation was linearly significant,
indicating that the number of oyster shells is critical in increasing 





. Figure 9b also
shows a positive linear correlation (=0.91) between 





and the structural volumes
of the AOR models. Figure 9 therefore indicates that the number of oyster shells is a key
factor for increasing the structural volumes of the AOR models and their wake volumes.
Figure 8. Wake volumes of the growth AOR models: (a) correlation between the wake volumes and
the growth stages (GSs) and (b) plane views of wake volumes when GS = 1 and GS = 16, respectively.
Figure 10a shows 





of the AOR models and their correlations with the improved
BIs. The correlation of the BIs with 





was signi ficant (=0.94 with a linear trend
line); hence, it is shown that the blocking area of an AOR model can help create 





. A
similar observation was made when Jung et al. [44] obtained the BIs of artificial reefs to
quantify their wake volumes. Moreover, fluid flow activity around an oyster reef on the
seabed causes sediments to accumulate inside over time. This phenomenon is expected to
increase the BI and, consequently, the wake volume. Figure 10b shows the correlation be-
tween 





of the AOR models with the surface complexity indices (SIs). The correlation
of the SIs with 





was significant (=0.92 with a linear trend line); this demonstrates
that the surface complexity of an AOR model can also help create its wake volume. Con-
sidering the ranges of SIs (2.77 ≤SI ≤4.11 and 4.64 ≤SI ≤10.46 for the initial and
growth AOR models, respectively), each growth AOR model increased its surface com-
plexity and 





.
Figure 8.
Wake volumes of the growth AOR models: (
a
) correlation between the wake volumes and
the growth stages (GSs) and (
b
) plane views of wake volumes when GS = 1 and GS = 16, respectively.
Figure 9a shows the normalized wake volumes (
WV
) of the AOR models and their
correlation with the numbers of oyster shells (
N
) introduced to the models. According to
the correlation coefficient of
R2=
0.89, the positive correlation was linearly significant,
indicating that the number of oyster shells is critical in increasing
WV
. Figure 9b also
shows a positive linear correlation (
R2=
0.91) between
WV
and the structural volumes
of the AOR models. Figure 9therefore indicates that the number of oyster shells is a key
factor for increasing the structural volumes of the AOR models and their wake volumes.
Figure 10a shows
WV
of the AOR models and their correlations with the improved
BIs. The correlation of the BIs with
WV
was significant (
R2=
0.94 with a linear trend line);
hence, it is shown that the blocking area of an AOR model can help create
WV
. A similar
observation was made when Jung et al. [
44
] obtained the BIs of artificial reefs to quantify
their wake volumes. Moreover, fluid flow activity around an oyster reef on the seabed
causes sediments to accumulate inside over time. This phenomenon is expected to increase
the BI and, consequently, the wake volume. Figure 10b shows the correlation between
WV
of the AOR models with the surface complexity indices (SIs). The correlation of the SIs
with
WV
was significant (
R2=
0.92 with a linear trend line); this demonstrates that the
surface complexity of an AOR model can also help create its wake volume. Considering
the ranges of SIs
(2.77 ≤SI ≤4.11
and 4.64
≤SI ≤
10.46 for the initial and growth AOR
models, respectively), each growth AOR model increased its surface complexity and WV.

J. Mar. Sci. Eng. 2023,11, 775 11 of 16
J. Mar. Sci. Eng. 2023, 11, x FOR PEER REVIEW 11 of 16
Figure 9. Normalized wake volume (





) according to (a) the number of oyster shells () and (b)
the structural volume of AOR (
).
Figure 9.
Normalized wake volume (
WV
) according to (
a
) the number of oyster shells (
N
) and
(b) the structural volume of AOR (VAOR ).
Figure 11 shows the EIs of the 96 AOR models and their correlation with the number
of oyster shells (
N
). According to their significant linear correlation (i.e.,
R2=
0.88 with
a linear trend line), the EI increased linearly with the number of oyster shells. The EIs
ranged from 0.45–7.21, and 89.6% of the growth AOR models (i.e., 43 models among 48)
had an
EI ≥
2.0 (Figure 11). Reflecting that only 26% of the 34 representative artificial reefs
in South Korea satisfy the efficiency range of
EI ≥
2.0 [
28
], growth AOR models would
be required to establish the necessary wake volume, as opposed to conventional artificial

J. Mar. Sci. Eng. 2023,11, 775 12 of 16
reef structures. This is because the growth AOR models have surface irregularities (or
roughness) due to the oyster shell accumulations. Here, the criterion
EI ≥
2.0 indicates that
the wake volume generated is more than twice that of the reef volume; hence, the criterion
has been recommended as a design condition of an artificial reef [
28
]. In this respect, such a
growth AOR model is ideal if it has a certain number of oyster shells (e.g.,
N≥
50 per the
plate area of 900 cm2) for the given composition.
J. Mar. Sci. Eng. 2023, 11, x FOR PEER REVIEW 12 of 16
Figure 10. Normalized wake volume (





) according to (a) the improved blocking index (BI) and
(b) the surface complexity index (SI).
Figure 11 shows the EIs of the 96 AOR models and their correlation with the number
of oyster shells (). According to their significant linear correlation (i.e., =0.88 with
a linear trend line), the EI increased linearly with the number of oyster shells. The EIs
ranged from 0.45–7.21, and 89.6% of the growth AOR models (i.e., 43 models among 48)
had an EI ≥2.0 (Figure 11). Reflecting that only 26% of the 34 representative artificial
reefs in South Korea satisfy the efficiency range of EI ≥2.0 [28], growth AOR models
would be required to establish the necessary wake volume, as opposed to conventional
artificial reef structures. This is because the growth AOR models have surface irregulari-
ties (or roughness) due to the oyster shell accumulations. Here, the criterion EI ≥2.0 in-
dicates that the wake volume generated is more than twice that of the reef volume; hence,
the criterion has been recommended as a design condition of an artificial reef [28]. In this
Figure 10.
Normalized wake volume (
WV
) according to (
a
) the improved blocking index (BI) and
(b) the surface complexity index (SI).

J. Mar. Sci. Eng. 2023,11, 775 13 of 16
J. Mar. Sci. Eng. 2023, 11, x FOR PEER REVIEW 13 of 16
respect, such a growth AOR model is ideal if it has a certain number of oyster shells (e.g.,
≥50 per the plate area of 900 cm2) for the given composition.
Figure 11. Correlation of the efficiency indices with number of oyster shells ().
So far, it has been shown that 





increased with the number of oyster shells (), BI,
and SI. Moreover, the EI increased with the number of oyster shells (). Therefore, it is
necessary to clarify the correlations between the three evaluation indices: EI, BI, and SI.
Figure 12 shows a strong linear correlation between each pair of the three indices (i.e.,
=0.94 for EI vs. BI; =0.92 for EI vs. SI; and =0.98 for SI vs. BI). This indicates
that the number of oyster shells simultaneously contributes to EI, BI, and SI, although
there are some slight deviations depending on the shell orientation, composition, and pen-
etration depth. According to Jung et al. [44], conventional artificial reefs in South Korea
generally do not simultaneously have a certain degree of rugosity (i.e., surface complex-
ity) and a water blocking effect. For example, a tunnel or arch-type artificial reef has a
certain degree of rugosity (e.g., 2.0), whereas a cube or box-type artificial reef has a certain
degree of water blocking effect (e.g., 0.5). This comparison illustrates that the AOR models
(i.e., the oyster shells accumulated on the plate substrate) have relatively good structural
characteristics. In other words, the growth stage (or the number of oyster shells) can be
regarded as a design variable as it is a necessary component for all three evaluation indices
as well as for 





.
Figure 12. Correlations of the three evaluation indices (EI, BI, and SI): (a) EI vs. BI, (b) EI vs. SI, and
(c) SI vs. BI. EI: efficiency index; BI: improved blocking index; SI: surface complexity index.
Despite the limitations of the shell models (i.e., equal shell size, equal plate substrate,
and single shell segment), the current AOR models provide insight into the construction
Figure 11. Correlation of the efficiency indices with number of oyster shells (N).
So far, it has been shown that
WV
increased with the number of oyster shells (
N
),
BI, and SI. Moreover, the EI increased with the number of oyster shells (
N
). Therefore, it
is necessary to clarify the correlations between the three evaluation indices: EI, BI, and
SI. Figure 12 shows a strong linear correlation between each pair of the three indices (i.e.,
R2=
0.94 for EI vs. BI;
R2=
0.92 for EI vs. SI; and
R2=
0.98 for SI vs. BI). This indicates
that the number of oyster shells simultaneously contributes to EI, BI, and SI, although there
are some slight deviations depending on the shell orientation, composition, and penetration
depth. According to Jung et al. [
44
], conventional artificial reefs in South Korea generally
do not simultaneously have a certain degree of rugosity (i.e., surface complexity) and a
water blocking effect. For example, a tunnel or arch-type artificial reef has a certain degree
of rugosity (e.g., 2.0), whereas a cube or box-type artificial reef has a certain degree of water
blocking effect (e.g., 0.5). This comparison illustrates that the AOR models (i.e., the oyster
shells accumulated on the plate substrate) have relatively good structural characteristics. In
other words, the growth stage (or the number of oyster shells) can be regarded as a design
variable as it is a necessary component for all three evaluation indices as well as for WV.
J. Mar. Sci. Eng. 2023, 11, x FOR PEER REVIEW 13 of 16
respect, such a growth AOR model is ideal if it has a certain number of oyster shells (e.g.,
≥50 per the plate area of 900 cm2) for the given composition.
Figure 11. Correlation of the efficiency indices with number of oyster shells ().
So far, it has been shown that 





increased with the number of oyster shells (), BI,
and SI. Moreover, the EI increased with the number of oyster shells (). Therefore, it is
necessary to clarify the correlations between the three evaluation indices: EI, BI, and SI.
Figure 12 shows a strong linear correlation between each pair of the three indices (i.e.,
=0.94 for EI vs. BI; =0.92 for EI vs. SI; and =0.98 for SI vs. BI). This indicates
that the number of oyster shells simultaneously contributes to EI, BI, and SI, although
there are some slight deviations depending on the shell orientation, composition, and pen-
etration depth. According to Jung et al. [44], conventional artificial reefs in South Korea
generally do not simultaneously have a certain degree of rugosity (i.e., surface complex-
ity) and a water blocking effect. For example, a tunnel or arch-type artificial reef has a
certain degree of rugosity (e.g., 2.0), whereas a cube or box-type artificial reef has a certain
degree of water blocking effect (e.g., 0.5). This comparison illustrates that the AOR models
(i.e., the oyster shells accumulated on the plate substrate) have relatively good structural
characteristics. In other words, the growth stage (or the number of oyster shells) can be
regarded as a design variable as it is a necessary component for all three evaluation indices
as well as for 





.
Figure 12. Correlations of the three evaluation indices (EI, BI, and SI): (a) EI vs. BI, (b) EI vs. SI, and
(c) SI vs. BI. EI: efficiency index; BI: improved blocking index; SI: surface complexity index.
Despite the limitations of the shell models (i.e., equal shell size, equal plate substrate,
and single shell segment), the current AOR models provide insight into the construction
Figure 12.
Correlations of the three evaluation indices (EI, BI, and SI): (
a
) EI vs. BI, (
b
) EI vs. SI, and
(c) SI vs. BI. EI: efficiency index; BI: improved blocking index; SI: surface complexity index.
Despite the limitations of the shell models (i.e., equal shell size, equal plate substrate,
and single shell segment), the current AOR models provide insight into the construction of
AORs in coastal waters. We can control the normalized wake volume simply by adjusting
the number of oyster shells on the plate substrate without pinpointing the shell orientation,
composition, or penetration depth. Moreover, the criterion for wake volume efficiency (i.e.,
EI ≥2.0) can be obtained when the number of shells reaches N≥50 per 900 cm2.

J. Mar. Sci. Eng. 2023,11, 775 14 of 16
4. Conclusions
This study proposed the hypothesis that “a unique wake region is formed around
an AOR due to its structural characteristics”. To test the hypothesis, we asked whether
the structural characteristics of an AOR have an effect on its wake region and what their
correlation is. To answer these questions, we established 96 AOR models, carried out flow
analyses, and obtained their wake volumes and the related evaluation indices. Considering
shell orientation, composition, penetration depth, and growth stage, we found that the
growth stage is most critical to the normalized wake volume. This implies that the number
of oyster shells is the most critical factor in securing a normalized wake volume considering
the establishment of the growth stages in the AOR models. Their correlation was linear
and significant (
R2=
0.89); thus, we can control the normalized wake volume by simply
adjusting the number of oyster shells on the plate substrate without pinpointing the shell
orientation, composition, or penetration depth. The correlations of the normalized wake
volume with the improved BI and SI were linear and significant (
R2=
0.94 and
R2=
0.92,
respectively). This indicates that the number of shells is closely connected to the indices;
hence, their strong linear correlations were also verified. The EI also increased with the
number of oyster shells, and the linear correlation shows that the criterion for wake volume
efficiency (i.e., EI
≥
2.0) can be obtained when the number of shells reaches
N≥
50 per
900 cm
2
. Therefore, the growth stage (or number of oyster shells) can be regarded as a
design variable when establishing an AOR in coastal waters as it responds sensitively to
both the three evaluation indices (BI, SI, and EI) and the normalized wake volume.
Author Contributions:
Conceptualization, M.K. and W.-B.N.; methodology, M.K., S.J. and W.-B.N.;
software, M.K., S.J. and T.V.C.; validation, S.J., T.V.C. and W.-B.N.; formal analysis, M.K., S.J. and
W.-B.N.; investigation, M.K., S.J., T.V.C. and W.-B.N.; resources, M.K., S.J., T.V.C. and W.-B.N.; data
curation, M.K., S.J. and W.-B.N.; writing—original draft preparation, M.K.; writing—review and
editing, W.-B.N.; visualization, M.K., S.J. and W.-B.N.; supervision, W.-B.N.; project administration,
W.-B.N.; funding acquisition, W.-B.N. All authors have read and agreed to the published version of
the manuscript.
Funding:
This research was supported by Korea Institute of Marine Science & Technology Promotion
(KIMST) and was funded by the Ministry of Ocean and Fisheries (20220252).
Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.
Data Availability Statement: Not applicable.
Conflicts of Interest: The authors declare no conflict of interest.
References
1. Yamaguchi, K. Shell structure and behaviour related to cementation in oysters. Mar. Biol. 1994,118, 89–100. [CrossRef]
2.
Reeves, S.E.; Renzi, J.J.; Fobert, E.K.; Silliman, B.R.; Hancock, B.; Gillies, C.L. Facilitating better outcomes: How positive species
interactions can improve oyster reef restoration. Front. Mar. Sci. 2020,7, 656. [CrossRef]
3.
Gillies, C.L.; Castine, S.A.; Alleway, H.K.; Crawford, C.; Fitzsimons, J.A.; Hancock, B.; Koch, P.; McAfee, D.; McLeod, I.M.; zu
Ermgassen, P.S.E. Conservation status of the Oyster Reef Ecosystem of Southern and Eastern Australia. Glob. Ecol. Conserv.
2020
,
22, e00988. [CrossRef]
4.
Sharma, S.; Goff, J.; Moody, R.M.; Byron, D.; Heck, K.L., Jr.; Powers, S.P.; Ferraro, C.; Cebrian, J. Do restored oyster reefs benefit
seagrasses? An experimental study in the Northern Gulf of Mexico. Restor. Ecol. 2016,24, 306–313. [CrossRef]
5.
Glenn, M.; Mathieson, A.; Grizzle, R.; Burdick, D. Seaweed communities in four subtidal habitats within the Great Bay estuary,
New Hampshire: Oyster farm gear, oyster reef, eelgrass bed, and mudflat. J. Exp. Mar. Biol. Ecol. 2020,524, 151307. [CrossRef]
6.
Barbier, E.B.; Hacker, S.D.; Kennedy, C.; Koch, E.W.; Stier, A.C.; Silliman, B.R. The value of estuarine and coastal ecosystem
services. Ecol. Monogr. 2011,81, 169–193. [CrossRef]
7.
Weaver, R.J.; Stehno, A.; Bonanno, N.; Sager, A.; Kenny, A.; Zehnder, J.A.; Glasser, C.; Allen, A. Scale model design of oyster reef
structures as part of a showcase living shoreline project. Shore Beach 2017,85, 41–54.
8. Styles, R. Flow and turbulence over an oyster reef. J. Coast. Res. 2015,31, 978–985. [CrossRef]
9.
Kitsikoudis, V.; Kibler, K.M.; Walters, L.J. In-situ measurements of turbulent flow over intertidal natural and degraded oyster
reefs in an estuarine lagoon. Ecol. Eng. 2020,143, 105688. [CrossRef]

J. Mar. Sci. Eng. 2023,11, 775 15 of 16
10.
Hitzegrad, J.; Brohmann, L.; Pfennings, K.; Hoffmann, T.K.; Eilrich, A.K.; Paul, M.; Welzel, M.; Schlurmann, T.; Aberle, J.;
Wehrmann, A.; et al. Oyster reef surfaces in the central Wadden Sea: Intra-reef classification and comprehensive statistical
description. Front. Mar. Sci. 2022,9, 808018. [CrossRef]
11.
Humphries, A.T.; La Peyre, M.K.; Decossas, G.A. The effect of structural complexity, prey density, and “predator-free space” on
prey survivorship at created oyster reef mesocosms. PLoS ONE 2011,6, e28339. [CrossRef] [PubMed]
12.
Walles, B.; Mann, R.; Ysebaert, T.; Troost, K.; Herman, P.M.J.; Smaal, A.C. Demography of the ecosystem engineer Crassostrea
gigas, related to vertical reef accretion and reef persistence. Estuar. Coast. Shelf Sci. 2015,154, 224–233. [CrossRef]
13. Loke, L.H.L.; Chisholm, R.A. Measuring habitat complexity and spatial heterogeneity in ecology. Ecol. Lett. 2022,25, 2269–2288.
[CrossRef] [PubMed]
14.
Windle, A.E.; Puckett, B.; Huebert, K.B.; Knorek, Z.; Johnston, D.W.; Ridge, J.T. Estimation of intertidal oyster reef density using
spectral and structural characteristics derived from unoccupied aircraft systems and structure from motion photogrammetry.
Remote Sens. 2022,14, 2163. [CrossRef]
15.
Howie, A.H.; Bishop, M.J. Contemporary oyster reef restoration: Responding to a changing world. Front. Ecol. Evol.
2021
,9,
689915. [CrossRef]
16.
O’Beirn, F.X.; Luckenbach, M.W.; Nestlerode, J.A.; Coates, G.M. Toward design criteria in constructed oyster reefs: Oyster
recruitment as a function of substrate type and tidal height. J. Shellfish Res. 2000,19, 387–395.
17.
Soniat, T.M.; Finelli, C.M.; Ruiz, J.T. Vertical structure and predator refuge mediate oyster reef development and community
dynamics. J. Exp. Mar. Biol. Ecol. 2004,310, 163–182. [CrossRef]
18.
McAfee, D.; McLeod, I.M.; Boström-Einarsson, L.; Gillies, C.L. The value and opportunity of restoring Australia’s lost rock oyster
reefs. Restor. Ecol. 2020,28, 304–314. [CrossRef]
19.
Christianen, M.J.A.; Lengkeek, W.; Bergsma, J.H.; Coolen, J.W.P.; Didderen, K.; Dorenbosch, M.; Driessen, F.M.F.; Kamermans,
P.; Reuchlin-Hugenholtz, E.; Sas, H.; et al. Return of the native facilitated by the invasive? Population composition, substrate
preferences and epibenthic species richness of a recently discovered shellfish reef with native European flat oysters (Ostrea edulis)
in the North Sea. Mar. Biol. Res. 2018,14, 590–597. [CrossRef]
20.
La Peyre, M.K.; Humphries, A.T.; Casas, S.M.; La Peyre, J.F. Temporal variation in development of ecosystem service from oyster
reef restoration. Ecol. Eng. 2014,63, 34–44. [CrossRef]
21.
Fivash, G.S.; Stüben, D.; Bachmann, M.; Walles, B.; van Belzen, J.; Didderen, K.; Temmink, R.J.M.; Lengkeek, W.; van der Heide, T.;
Bouma, T.J. Can we enhance ecosystem-based coastal defense by connecting oysters to marsh edges? Analyzing the limits of
oyster reef establishment. Ecol. Eng. 2021,165, 106221. [CrossRef]
22.
Cannon, D.; Kibler, K.; Walters, L.; Chambers, L. Hydrodynamic and biogeochemical evolution of a restored intertidal oyster
(Crassostrea virginica) reef. Sci. Total Environ. 2022,831, 154879. [CrossRef] [PubMed]
23.
Cannon, D.; Kibler, K.M.; Kitsikoudis, V.; Medeiros, S.C.; Walters, L.J. Variation of mean flow and turbulence characteristics
within canopies of restored intertidal oyster reefs as a function of restoration age. Ecol. Eng. 2022,180, 106678. [CrossRef]
24.
Goelz, T.; Vogt, B.; Hartley, T. Alternative substrates used for oyster reef restoration: A review. J. Shellfish Res.
2020
,39, 1–12.
[CrossRef]
25.
Walles, B.; Troost, K.; van den Ende, D.; Nieuwhof, S.; Smaal, A.C.; Ysebaert, T. From artificial structures to self-sustaining oyster
reefs. J. Sea Res. 2016,108, 1–9. [CrossRef]
26.
Kuykendall, K.M.; Moreno, P.; Powell, E.N.; Soniat, T.M.; Colley, S.; Mann, R.; Munroe, D.M. The exposed surface area to volume
ratio: Is shell more efficient than limestone in promoting oyster recruitment? J. Shellfish Res. 2015,34, 217–225. [CrossRef]
27.
La Peyre, M.; Furlong, J.; Brown, L.A.; Piazza, B.P.; Brown, K. Oyster reef restoration in the northern Gulf of Mexico: Extent,
methods and outcomes. Ocean Cost. Manag. 2014,89, 20–28. [CrossRef]
28.
Kim, D.; Woo, J.; Yoon, H.S.; Na, W.B. Efficiency, tranquillity and stability indices to evaluate performance in the artificial reef
wake region. Ocean Eng. 2016,122, 253–261. [CrossRef]
29.
Woo, J.; Kim, D.; Yoon, H.S.; Na, W.B. Efficient placement models of labyrinth-type artificial concrete reefs according to wake
volume estimation to support natural submerged aquatic vegetation. Bull. Mar. Sci. 2018,94, 1259–1272. [CrossRef]
30.
Wheaton, F. Review of the properties of Eastern oysters, Crassostrea virginica: Part I. Physical properties. Aquac. Eng.
2007
,37,
3–13. [CrossRef]
31.
Whitman, E.R.; Reidenbach, M.A. Benthic flow environments affect recruitment of Crassostrea virginica larvae to an intertidal
oyster reef. Mar. Ecol. Prog. Ser. 2012,463, 177–191. [CrossRef]
32. ANSYS Inc. ANSYS CFX-Solver Theory Guide Release 2020-R1; ANSYS Inc.: Canonsburg, PA, USA, 2020.
33.
Kim, D.; Jung, S.; Kim, J.; Na, W.B. Efficiency and unit propagation indices to characterize wake volumes of marine forest artificial
reefs established by flatly distributed placement models. Ocean Eng. 2019,175, 138–148. [CrossRef]
34.
Menter, F.R. Review of the shear-stress transport turbulence model experience from an industrial perspective. Int. J. Comput.
Fluid Dyn. 2009,23, 305–316. [CrossRef]
35.
Kim, D.; Jung, S.; Na, W.B. Evaluation of turbulence models for estimating the wake region of artificial reefs using particle image
velocimetry and computational fluid dynamics. Ocean Eng. 2021,223, 108673. [CrossRef]
36.
Ono, M.; Deguchi, I. Relation between aggregating effect of artificial fish reef and flow pattern around reef. In Coastal Structures
2003; American Society of Civil Engineers: Portland, OR, USA, 2003; pp. 1164–1175. [CrossRef]

J. Mar. Sci. Eng. 2023,11, 775 16 of 16
37.
Miller, D.C.; Norkko, A.; Pilditch, C.A. Influence of diet on dispersal of horse mussel Atrina zelandica biodeposits. Mar. Ecol. Prog.
Ser. 2002,242, 153–167. [CrossRef]
38.
Yoon, H.S.; Kim, D.; Na, W.B. Estimation of effective usable and burial volumes of artificial reefs and the prediction of cost-effective
management. Ocean Coast. Manag. 2016,120, 135–147. [CrossRef]
39.
Rouse, S.; Porter, J.S.; Wilding, T.A. Artificial reef design affects benthic secondary productivity and provision of functional
habitat. Ecol. Evol. 2020,10, 2122–2130. [CrossRef] [PubMed]
40.
Kim, D.; Woo, J.; Yoon, H.S.; Na, W.B. Wake lengths and structural responses of Korean general artificial reefs. Ocean Eng.
2014
,
92, 83–91. [CrossRef]
41.
Kim, D.; Woo, J.; Na, W.B.; Yoon, H.S. Flow and structural response characteristics of a box-type artificial reef. J. Korean Soc. Coast.
Ocean Eng. 2014,26, 113–119. [CrossRef]
42.
Lee, I.C.; Kim, D.; Jung, S.; Na, W.B. Prediction of primary physical measures for cost-effective management of artificial seaweed
reefs. Mar. Technol. Soc. J. 2020,54, 25–43. [CrossRef]
43.
Kim, D.; Jung, S.; Na, W.B. Two perspectives for increasing of artificial wake region. J. Fish. Mar. Sci. Educ.
2020
,32, 1623–1631.
[CrossRef]
44. Jung, S.; Na, W.B.; Kim, D. Rugosity and blocking indices of artificial reefs and their correlations with wake volume. Ocean Eng.
2022,261, 112204. [CrossRef]
Disclaimer/Publisher’s Note:
The statements, opinions and data contained in all publications are solely those of the individual
author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to
people or property resulting from any ideas, methods, instructions or products referred to in the content.