Experimental and computational analysis of a novel …rsfs.royalsocietypublishing.org/content/royfocus/5/1/...rsfs.royalsocietypublishing.org Research Cite this article: Dimartino

on May 23, 2018http://rsfs.royalsocietypublishing.org/Downloaded from

rsfs.royalsocietypublishing.org

ResearchCite this article: Dimartino S, Mather AV,

Alestra T, Nawada S, Haber M. 2015

Experimental and computational analysis of a

novel flow channel to assess the adhesion

strength of sessile marine organisms. Interface

Focus 5: 20140059.

http://dx.doi.org/10.1098/rsfs.2014.0059

One contribution of 15 to a theme issue

‘Biological adhesives: from biology to

biomimetics’.

Subject Areas:biomechanics, biomimetics, chemical

engineering

Keywords:flow channel, bioadhesion, computational fluid

dynamics, adhesion strength, macroalgae,

Hormosira banksii

Author for correspondence:Simone Dimartino

e-mail: [email protected]

†These authors contributed equally to this

study.

& 2014 The Author(s) Published by the Royal Society. All rights reserved.

Experimental and computational analysisof a novel flow channel to assess theadhesion strength of sessile marineorganisms

Simone Dimartino1,2,†, Anton V. Mather3,†, Tommaso Alestra3, Suhas Nawada2

and Meir Haber4

1Department of Chemical and Process Engineering, 2Biomolecular Interaction Centre, and 3Department ofBiological Sciences, University of Canterbury, Christchurch, New Zealand4Biota Ltd, PO Box 220, Or Akiva 30600, Israel

SD, 0000-0002-9695-1278

Bioadhesives produced by marine macroalgae represent a potential source

of inspiration for the development of water-resistant adhesives. Assessing

their adhesion strength, however, remains difficult owing to low volumes of

adhesive material produced, low solubility and rapid curing time. These diffi-

culties can be circumvented by testing the adhesion strength of macroalgae

propagules attached to a substrate. In this paper, we present a simple, novel

flow channel used to test the adhesion strength of the germlings of the fucalean

alga Hormosira banksii to four substrates of biomedical relevance (PMMA, agar,

gelatin and gelatin þ lipid). The adhesion strength of H. banksii germlings was

found to increase in a time-dependent manner, with minimal adhesion success

after a settlement period of 6 h and maximum adhesion strength achieved 24 h

after initial settlement. Adhesion success increased most dramatically between

6 and 12 h settlement time, while no additional increase in adhesion strength

was recorded for settlement times over 24 h. No significant difference in

adhesion strength to the various substrates was observed. Computational

fluid dynamics (CFD) was used to estimate the influence of fluid velocity

and germling density on drag force acting on the settled organisms. CFD mod-

elling showed that, on average, the drag force decreased with increasing

germling number, suggesting that germlings would benefit from gregarious

settlement behaviour. Collectively, our results contribute to a better under-

standing of the mechanisms allowing benthic marine organisms to thrive in

hydrodynamically stressful environments and provide useful insights for

further investigations.

1. IntroductionAdhesives produced by sessile marine organisms, including macroalgae, exhibit

strong adhesion to substrates with different surface properties [1,2]. Adhesion in

the marine environment occurs despite heavily biofouled surfaces [3,4] and the

strong drag forces that tidal and wave currents generate [5]. Greater understand-

ing of the mechanisms of adhesion present in these marine organisms could lead

to the development of bioinspired adhesives with applications in the underwater

engineering and biomedical arenas.

Brown algae (class Phaeophyta) include large species such as the kelps

Macrocystis pyrifera [6] and Nereocystis luetkaena, and the fucoid Durvillaea antarc-tica [7]. Adult plants of these species can be up to 10 m long, weigh more than

50 kg and are able to withstand severe hydrodynamic drag in wave-exposed

areas. These features make brown algae particularly interesting in the bioadhe-

sion arena. The adhesive secretions from brown algae are mostly composed of a

network of cross-linked polyphenolic compounds and negatively charged poly-

saccharides [8–12]. Mechanical testing has been performed on adult specimens of

rsfs.royalsocietypublishing.orgInterface

Focus5:20140059

2

on May 23, 2018http://rsfs.royalsocietypublishing.org/Downloaded from

D. antarctica [13] and Hormosira banksii [14], revealing the ability

to withstand pull-out forces up to 300 N. Although experiments

on adult plants set the grounds for further research, adhesion of

young germlings is of ecological and engineering relevance.

Dispersal and settlement of seaweed propagules is, in fact,

crucial for the persistence of adult stands [15], while the per-

formance of the initial adhesive (i.e. before curing) is of

obvious interest in biomedical applications. Among the brown

algae, H. banksii is particularly interesting because its dioecious

nature permits the separate collection of male and female

gametes. This allows us to control fertilization, which in turn

triggers germination and the production of the adhesive

materials [16]. Chemical characterization of the adhesive secre-

tions produced by H. banksii is currently under investigation by

our group, and only a microscopic analysis of the adhesive pad

is available so far [17]. A rigorous mechanical study of the

adhesion force has never been undertaken.

Assessing the adhesion strength is key to characterize

the mechanical properties of the bioadhesives produced

by marine organisms. Conventional assays such as tensile,

shear and peel tests require the extraction of significant quan-

tities of adhesives, often an unrealistic task for marine

bioadhesives owing to their low solubility, fast curing time

and small quantities secreted [10]. Ideally, a more practical

and low-cost testing method would not require the extraction

of adhesive material while being rapid and repeatable. The

use of a water-generated, controllable hydrodynamic force

is useful to test the adhesion of sessile marine organisms

while minimizing collateral stresses generated by the exper-

imental conditions. Methods employing hydrodynamically

generated forces for assessing the adhesion strength of sessile

marine organisms have been used in the past, ranging from

exposure to naturally generated waves [18] to apparatuses

such as radial, annular and straight flow channel designs,

water jets [19,20] and turbulence tanks [21]. Among these

methods, flow channels have been common, ranging from micro-

fluidic devices using laminar flows to test the adhesion of

bacteria and microalgae [22,23] to larger devices used to assess

diatom attachment across multiple microscope slides using tur-

bulent flow regimes [24,25]. Flow channel-based adhesion

assessment is also common in the study of biofouling-resistant

and fouling-release coatings [26–29]. However, these studies

do not investigate the actual drag force acting on the settled

organisms and propose the wall shear stress generated in

the channel as an estimate for the adhesion strength.

Conversion between these two forces is not obvious and is

likely to be complicated by a number of factors such as the

geometry of the flow channel and the relative dimension of

the species considered.

Computational fluid dynamics (CFD) represents a useful

tool to estimate the drag forces acting on underwater organ-

isms. CFD methods are relatively easy to implement and have

been mostly applied to determine the flow field and hydro-

dynamic drag acting on isolated cells adhering to flat

surfaces [30–32], or to walls of parallel plate flow chambers.

From a mechanistic standpoint, the forces estimated using a

computational approach are more compelling than the

ambiguous value of the wall shear stress. In fact, Lu et al.[33] remarked that the actual shear stress felt by the settled

organisms could be significantly different from the wall

shear stress because of a non-negligible distortion of the

flow by the cell. However, the model of an isolated cell is not

appropriate for the case of kelp and fucoid propagules, as

these often colonize surfaces gregariously, forming random

arrays of settled germlings. The presence of neighbouring

germlings further distorts the flow field, creating wakes that

protect the germlings settled further down the flow channel.

Brooks & Tozeren [34] have considered the effect of uniformly

distributed arrays of cells on the drag force, while a random

distribution has not been taken into account so far.

In this work, we introduce a novel and simple parallel

plate flow channel designed to test the adhesion strength of

different marine organisms on different surfaces. The flow

channel was tested using H. banksii germlings settled on a

variety of substrates of biomedical relevance. CFD modelling

was used to determine the drag forces acting on H. banksiigermlings and to quantify the actual drag force experienced

by random arrays of settled germlings.

2. Material and methods2.1. Preparation of the substratesSeawater was collected from Sumner, Christchurch, New Zealand

(43857015.3100 S, 172876056.7000 E), and filtered through a 0.2 mm

filter. Agar (bacteriological agar; Oxoid Ltd, Basingstoke, UK)

was purchased from Thermo-Fisher Scientific. Absolute ethanol

was obtained from Nuplex Specialties (Auckland, New Zealand).

The ingredients used to prepare the skin model substrates were

used as received without further purification: gelatin (type A,

275 bloom porcine skin; Gelita, Sergeant Bluff, IA, USA), lipid (Pro-

lipid 141; International Specialty Products, Wayne, NJ, USA) and a

cross-linker, microbial transglutaminase (Activa TG; Ajinomoto,

Tokyo, Japan). Poly(methyl-methacrylate) (PMMA) flat sheet

with a thickness of 2 mm (average roughness 63.5+16.8 nm)

was purchased from Dotmar Engineering Plastics (Christchurch,

New Zealand).

Carbohydrate-based substrates (agar) were prepared from a

1.5% w/v solution of agar dissolved in purified water (Milli-Q,

Millipore) at 708C. A measure of 100 ml aliquots of agar solution

were transferred to clean PMMA slides (25 � 75 mm) and uni-

formly spread across the slide surface on a horizontal table.

The gelatin hydrogels (XL-Gel) were prepared using a 5% w/v

gelatin solution in filtered seawater at a temperature of 508C.

Cross-linking of the protein fraction was induced by adding a

5% w/v stock solution of the cross-linker in RO water to give a

final concentration of 0.2% of the cross-linker in the gelatin sol-

ution. A measure of 100 ml aliquots of the cross-linker and

gelatin solution were transferred to clean PMMA slides before

the cross-linking reaction completed. The gelatin hydrogels

with added lipid (XL-Gel-Lip) were prepared following the

same protocol, with the addition of a 20% w/v lipid solution

in ethanol to a final concentration of 0.8% lipid prior to the

addition of the cross-linker. The thickness of the hydrogel

layers was 50+3 mm.

For each substrate, 16 replicate slides were prepared (64

slides in total). These were kept in separate plastic containers

(80 � 27 � 22 mm) and covered with 30 ml of filtered seawater

to equilibrate the hydrogels. Two additional slides of each sub-

strate were also prepared: one was seeded with H. banksiizygotes and used to monitor germling development while the

other was used to measure contact angles.

In the remainder of this paper, the four surface types will be

referred to as PMMA, Agar (agarose gel), XL-Gel (cross-linked

gelatin) and XL-Gel-Lip (cross-linked gelatin with lipid component).

2.2. Contact angle measurementContact angle measurements were performed depositing 2 ml dro-

plets of deionized water on each substrate at 228C. Advancing

(a) (b)flow inlet f 2 mm flow outlet f 2 mm

65 mm

100 mm

4 mm

50 mm

Figure 1. (a) Lateral and (b) frontal cross sections of the flow channel. Hatching denotes the substrate (slide) tested, while the silicon gasket is shown in black.

rsfs.royalsocietypublishing.orgInterface

Focus5:20140059

3

on May 23, 2018http://rsfs.royalsocietypublishing.org/Downloaded from

contact angles (uAW) were measured using a goniometer (CAM

200; KSV Instruments Ltd, Helsinki, Finland). Ten images of the

water droplet were taken over the course of 1 min and the average

contact angle was calculated using software (KSV CAM Software

v. 4.01) fitting the droplet shape to the Young–Laplace equation.

Measurements were repeated four times on each substrate.

Figure 2. Flow apparatus during operation placed under the Nikon SMZ-1Bdissecting microscope with ring light. (Online version in colour.)

2.3. Sample collection, gamete release and fertilization,and settlement of germlings

The protocol used to obtain the germling suspensions is similar to

the one previously described by Taylor et al. [18,21]. The H. banksiiplants were collected at low tide from Shag Point, along the east

coast of the South Island of New Zealand (45827035.5000 S,

170848047.2600 E), in April of 2014. All plants were thoroughly

washed with filtered seawater to remove larger contaminants

and stored in a dark environment at 68C overnight. Gamete release

was triggered by placing the plants under halogen lights (two

lights, 200 W each) for 30–60 min at a temperature of 258C. Egg

and sperm solutions were prepared by washing female and male

plants in separate seawater baths. Egg and sperm solutions were

then filtered through 105 mm and 25 mm filters, respectively. The

egg suspension was further clarified by allowing the eggs to

deposit on the bottom of the container and replacing the super-

natant seawater three times. Sperm and egg suspensions were

then mixed to allow fertilization to occur. Sperm activity, egg

viability and fertilization were checked using a compound micro-

scope (Nikon model SE, Tokyo, Japan) equipped with a 1.3

megapixel USB CMOS camera (ODCM0130C; ProSciTech, Towns-

ville, Australia) controlled using ToupView (x64 v3.7.1691). The

volume of the suspension was adjusted with seawater to obtain a

concentration of approximately 20 000 zygotes per millilitre. Here-

after, we will refer to fertilized eggs in suspension as zygotes, while

zygotes attached to a substrate will be referred to as germlings.

Each replicate of the four substrates was seeded with 2 ml

zygote solution. Additional slides of each substrate (four) were

also seeded to monitor germling development over time. The

germlings were cultured in a temperature-controlled room at

158C under a pair of fluorescent bulbs (light intensity ¼

40 mmol photons m22 s21 PAR) on a 12 L : 12 D cycle. Seawater

in the trays was replaced every 24 h to prevent nutrient depletion.

Germling adhesion on the different substrates was assessed at

different post-settlement stages: after 6, 12, 24 and 96 h following

the seeding with the zygotes. Four replicates of each substrate

were randomly selected to be tested with the flow channel at the

different post-settlement times. Germling development was also

recorded after 6, 12, 24 and 96 h by measuring the diameter of

the germlings settled on the additional slides.

2.4. Flow channel design and testing of germlingadhesion strength

A diagram of the flow channel is shown in figure 1, while a photo-

graph of the actual set-up is presented in figure 2. The internal

dimensions of the flow channel were 65 mm L � 4 mm W �0.5 mm H, with inlet and outlet ports of 2 mm inner diameter. Sub-

strates to be tested were clamped between a 22 mm thick Perspex

top and a 10 mm thick stainless-steel base. A 0.7 mm thick silicone

gasket maintained water-tightness and defined the channel walls.

The gasket was held in place in a 0.6 mm deep groove surround-

ing the channel in the Perspex top. Seawater used in the flow

experiments was filtered at 0.2 mm and kept in a temperature-

controlled environment at 158C to minimize the stress caused to

the germlings because of temperature variations. Seawater was

pumped through the channel using the feed pump block (P-984;

four pump heads) of an AKTAcrossflow (GE Healthcare, Uppsala,

Sweden) controlled by UNICORN software.

Sample slides were clamped into the flow channel and

exposed to a stepwise flow rate, starting at 50 ml min21 and

increasing up to a maximum of 300 ml min21 with 50 ml min21

steps, corresponding to average fluid velocities between 0.42

and 2.5 m s21 with 0.42 m s21 step size. Each step lasted 15 s.

Germling attachment to the substrate was monitored in a

single field of view using the USB camera mentioned before

Table 1. Contact angles of the substrates investigated (+s.e., n ¼ 4).

substrate contact angle, uAW

PMMA 728+ 38

Agar 318+ 38

XL-Gel 568+ 28

XL-Gel-Lip 658+ 18

rsfs.royalsocietypublishing.orgInterface

Focus5:20140059

4

on May 23, 2018http://rsfs.royalsocietypublishing.org/Downloaded from

and a Nikon SMZ-1B dissecting microscope (Nikon Corporation,

Tokyo, Japan) capable of up to 3.5� magnification. The dissect-

ing microscope’s focal length easily accommodated the flow

channel. Viewing at 3� magnification allowed the differentiation

of cells as small as 30 mm within a viewable area of 9 mm2. The

ToupView software was used to control the camera and capture

video of the experiments. Selected frames were extracted from

the flow channel footage using VLC Media Player (v. 2.1.3).

For each replicate, an initial frame was extracted before the

germlings were exposed to the flow to calculate their initial den-

sity. In the following, density and surface density will be

interchangeably used to denote the number of germlings per

unit area. Additional frames were extracted at the end of every

15 s flow step. Germling densities were measured using the auto-

matic particle counting feature of the software IMAGEJ (v. 1.47).

Percentage survival after each flow step, Sv, was expressed as

the ratio between the density of remaining germlings, Dv, and

the initial density, D0,

Sv ¼Dv

D0� 100: (2:1)

Factorial analysis of variance (ANOVA) was used to test the

influence of different settlement times and substrates on the

initial abundance of the germlings (i.e. before being exposed to

the flow) and on their survival rates recorded at the conclusion

of the flow experiment (i.e. following the exposure to all flow

intensities). Statistical significance was judged setting a at 0.025

since two repeated, non-independent measures were taken

from each replicate slide. The initial abundance of the germlings

was analysed using a two-way ANOVA with the fixed factors:

settlement time (four levels: 6, 12, 24 or 96 h) and material

(four levels: PMMA, Agar, XL-Gel or XL-Gel-Lip). Data of germ-

ling percentage survival recorded at the conclusion of the flow

experiment were analysed using a two-way ANOVA with the

fixed factors: settlement time (three levels: 12, 24 or 96 h) and

material (four levels: PMMA, Agar, XL-Gel or XL-Gel-Lip). The

slides cultured for 6 h were excluded from this second analysis

because survival rates could not be calculated due to the low start-

ing germling densities compared with the other combinations of

treatments (see Results and discussion). Before all analyses, var-

iance heterogeneity was tested with the Cochran’s C-tests and

removed with appropriate transformations when required. When

homogeneity of variances could not be achieved by transform-

ation, data were analysed nonetheless by judging significance

more conservatively (a ¼ 0.01). Student–Newman–Keuls (SNK)

tests were performed for a posteriori comparisons of the means [35].

2.5. Computational fluid dynamicsThe software COMSOL MULTIPHYSICS 4.2a (COMSOL Inc., Stockholm,

Sweden) was used to reproduce the experimental flow channel

and generate CFD models. Minor geometrical imperfections or

surface roughness in the flow channel were not considered in the

CFD simulations.

Hormosira banksii germlings were modelled as smooth spheri-

cal bodies randomly placed and rigidly connected to the bottom

surface of the flow cell. Average values of germling diameter and

settlement density were obtained from the adhesion experiment

(i.e. 70 mm and 20 germlings mm22, respectively; see Results and

discussion). CFD simulations were repeated on five different

random arrays of spheres to account for the variability in the vel-

ocity profile and drag force acting on different random

configurations of settled germlings. Flow rates used in the CFD

models were based on those used in the adhesion experiment, cor-

responding to average velocities in the channel between 0.42 and

2.5 m s21. Viscosity and density of seawater at 158C were used

in the CFD simulations, i.e. 1.08 � 1023 Pa s and 1026 kg m23,

respectively [36].

Additional CFD simulations were also generated for (i) the

empty flow channel under the same flow conditions experimen-

tally investigated, (ii) a single isolated cell under the same flow

conditions experimentally investigated, and (iii) random arrays

of germlings settled at seven intermediate cell densities, ranging

from 1 to 22 germlings per mm2, at the maximum experimental

average velocity of 2.5 m s21.

The finite-elements method was used to solve the Navier–

Stokes equations of fluid motion for a time-independent velocity

profile across the domain. A flat velocity profile was assumed at

the inlet orifice and a zero pressure was set at the outlet tube of

the flow channel. No-slip conditions were considered along both

the channel walls and the surface of the germlings. The mesh was

refined at a minimum element size of 7.41 mm at the sphere per-

iphery (with an individual sphere consisting of 5438 elements)

and was gradually coarsened up to 110.24 mm at the peripheries

of the flow channel. This arrangement provided the optimal

mesh size as a trade-off between a grid-independent solution

and computational time. The overall drag force was calculated

as the sum of the elementary forces acting on each mesh element

of the periphery of the spherical bodies. To further reduce com-

putational time, mostly associated with the spheres and the fine

mesh at their periphery, the array was built only on a limited

portion of the flow channel where a fully developed velocity

field was established. A preliminary study was conducted con-

sidering different areas populated with random spheres at

densities of 5 and 20 spheres per mm2. This study revealed

that a 1 � 1 mm2 section provided invariant results while mini-

mizing computational time. Accordingly, all simulations were

performed considering random arrays of spherical bodies con-

fined in a 1 � 1 mm2 area.

3. Results and discussion3.1. Substrate selection and contact angle measurementFour different substrates were considered to assess the

adhesion strength of the biological glue secreted by H. banksiizygotes: untreated PMMA as a reference solid substrate, and

three hydrogels based on proteinaceous and carbohydrate

components. The substrates were chosen on the basis of

three considerations: (i) composition of the biofilm covering

the substrate in the natural environment, (ii) relevance in

biomedical applications, and (iii) diversity of chemical

compositions.

The PMMA is a material widely used in biomedical appli-

cations, with particular relevance in dentistry [37], orthopaedic

surgery and in contact and intraocular lenses [38]. Human skin,

the organ with the largest exposed surface, is mainly composed

of collagen and lipids [39]. Cross-linked gelatin is widely

applied in the medical field as absorbable haemostats, tissue

adhesives and sealants, and scaffolds for tissue engineering

[39–41]. Agarose hydrogel approximates the polysaccharide

component of extracellular polymeric substances (EPSs)

(a)

100 mm

10× (b)

100 mm

10×

(c)

100 mm

10× (d )

100 mm

10×

Figure 3. Photographs of H. banksii germlings settled on PMMA slides at different times after fertilization. (a) 6 hours. A thick cell wall is visible on the spherical-shaped germlings, surrounded by residual sperm. (b) 12 hours. (c) 24 hours. Polarization of the germlings and first cell division (marked by arrows) is visible.(d) 96 hours. Numerous cell divisions and rhizoid formation visible. Arrows indicate the accumulation of adhesive-containing physodes at the rhizoidal tip[12]. (Online version in colour.)

Table 2. Average zygote/germling diameter across all materials and at different stages of development (+s.e., n ¼ 10). Zygote diameter was measuredimmediately after fertilization and before settlement.

developmental stage

germling diameter (mm)

PMMA Agar XL-Gel XL-Gel-Lip

zygotes 65.9+ 0.9 mm

6 h germling 67.0+ 1.3 67.1+ 1.0 66.9+ 0.7 65.4+ 1.2

12 h germling 68.5+ 0.3 68.3+ 0.6 70.0+ 0.9 68.3+ 0.6

24 h germling 70.7+ 0.6 71.4+ 0.9 71.1+ 0.9 70.8+ 1.0

96 h germling 75.6+ 1.4 75.7+ 1.3 75.6+ 1.3 76.3+ 1.5

rsfs.royalsocietypublishing.orgInterface

Focus5:20140059

5

on May 23, 2018http://rsfs.royalsocietypublishing.org/Downloaded from

produced by bacterial biofilms [42] common on substrates

available to marine macroalgae [43].

Surface wettability can have a significant impact on settle-

ment rate and adhesion strength of marine organisms [44].

Contact angles were measured to assess the relative hydro-

phobic/hydrophilic character of the different substrates

(table 1). All studied substrates displayed hydrophilic charac-

ter (uAW , 908). The agar substrate had the most hydrophilic

characteristics (uAW ¼ 318) associated with the high content of

hydroxyl groups in the carbohydrate backbone [27]. PMMA

is inherently hydrophobic, but hydroxyl groups are often

introduced in the polymeric structure to decrease its hydro-

phobicity in biomedical applications [45]. The PMMA

material used in our study had a moderate hydrophilic char-

acter with a contact angle of 728. The XL-Gel substrate

displayed intermediate hydrophobicity, while the addition

of the lipid component into the gelatin hydrogel, even if in

relatively small amount, shifted the contact angle from 568to 658.

3.2. Germling developmentHormosira banksii germling growth was monitored at 6, 12, 24

and 96 h following settlement as illustrated in figure 3, while

table 2 details the diameters of germlings cultured on the

different substrates at the different developmental stages.

Initial zygote diameter immediately post-fertilization was

65.9 mm, while an average diameter of 75.8 mm was measured

at 96 h post-settlement, indicating a size increase around 15%,

with negligible differences among substrates. Germling size

increased slightly during the first hours after fertilization, but

polarization, cell duplication and formation of a pro-rhizoid

only occurred after around 24 h. Between 48 and 72 h the

germlings shed the primary cell wall that formed immediately

following fertilization (pictures not shown), with the rhizoid

tip becoming the main point of attachment. While these obser-

vations do not necessarily imply that the specific processes

involved in the production and secretion of adhesives are iden-

tical regardless of substrate, they do indicate that progression

06 12

settlement time (h)24 96

Agar

PMMA

XL-Gel

XL-Gel-Lip

5

10

15D

0 (g

erm

lings

mm

–2)

20

25

30

Figure 4. Initial density (þs.e., n ¼ 4) of H. banksii germlings settled on the various substrates measured at different post-settlement times.

00 0.42 0.83 1.25

v (m s–1)

Dv

(ger

mlin

gs m

m–2

)

1.67

Agar

PMMA

XL-Gel

XL-Gel-Lip

2.08 2.50

5

10

15

20

25

Figure 5. Surface density of H. banksii germlings settled for 6 hours and exposed to increasing linear velocities (þs.e., n ¼ 4).

rsfs.royalsocietypublishing.orgInterface

Focus5:20140059

6

on May 23, 2018http://rsfs.royalsocietypublishing.org/Downloaded from

through the developmental stages is consistent; therefore, pri-

mary and secondary adhesion events might be similar across

all substrates.

The dimensions of the germlings were always one order

of magnitude higher than the roughness of the PMMA

slides and the hydrogel coatings. Accordingly, the influence

of surface roughness on germling attachment was neglected.

3.3. Hormosira banksii zygote settlementInitial germling densities, measured at different post-

settlement stages before the flow experiment, varied with

settlement time (figure 4). Substrates sampled after 6 h had

significantly fewer germlings than those sampled after

12, 24 and 96 h, which did not differ among each other

(F3,48 ¼ 28.71, p , 0.001). There were no significant differ-

ences among materials (F3,48 ¼ 2.45, p ¼ 0.07) and no

interactive effects of material and settlement time (F9,48 ¼

2.08, p ¼ 0.05). The average germling density 12, 24 and

96 h after the settlement was around 20 germlings mm22.

Reduced densities at 6 h were probably due to the detach-

ment of the germlings as the slides were transferred from

their trays to the flow channel. This observation suggests a

poor attachment of the germlings in the early post-fertiliza-

tion stages, probably due to the fact that the adhesion

curing process requires longer than 6 h to achieve significant

strength. However, there is a suggested, albeit not statistically

significant, elevated initial density on the Agar substrate after

6 h settlement. Bearing in mind that this may be the result of

uncontrolled variability in forces generated when transferring

substrates to the flow channel, it is also worth noting the

possibility that this is the result of rapid interactions between

the polyphenolic adhesive secretion, the polyanionic moieties

of the substrate and positive divalent ions present in sea-

water, namely Ca2þ and Mg2þ [3]. Alternatively, increased

settlement on Agar may lie in haloperoxidase-mediated

cross-linking reactions involving polyphenolic adhesive

compounds and the agarose chains of the substrate [8,9,46],

though it is unlikely that these processes will have taken

place by 6 h settlement. At the other end of the spectrum,

initial settlement density after 6 h is lowest on the XL-

Gel-Lip surface. Hydrophobic lipid chains embedded in the

collagen network may have an influence in preventing

the adsorption of the adhesive to the substrate, possibly by

limiting the electrostatic forces involved in adsorption and/

or by sterically hindering adhesive access to receptive

substrate functional groups. It is apparent, though, that

exposure to even the lowest flow velocities removes the rela-

tive variations in settlement density (figure 5), supporting the

idea that any adhesion achieved by the germlings after 6 h is

defeated at shear stresses� 5 Pa regardless of substrate type.

This is consistent with the observations reported by Taylor &

Schiel [18], who concluded that H. banksii zygotes show

feeble adhesion between 1 and 6 h post-settlement. In their

experiments, cultures of H. banksii zygotes were settled on

fibre-based cement plates and challenged with artificially

100

80

60

12 h

24 h

96 h

40

20

00 0.5 1.0

v (m s–1)

PMMAS v

(%)

tw (Pa)

1.5 2.0 2.5

0 6 13 19 26 32

0 0.5 1.0v (m s–1)

Agar

tw (Pa)

1.5 2.0 2.5

0 6 13 19 26 32

100

80

60

40

20

00 0.5 1.0

v (m s–1)

XL-Gel

S v (%

)

tw (Pa)

1.5 2.0 2.5

0 6 13 19 26 32

0 0.5 1.0v (m s–1)

XL-Gel-Lip

tw (Pa)

1.5 2.0 2.5

0 6 13 19 26 32

12 h

24 h

96 h

12 h

24 h

96 h

12 h

24 h

96 h

Figure 6. Germling survival (+s.e., n ¼ 4) versus linear velocity and wall shear stress for the different material surfaces and settlement times.

rsfs.royalsocietypublishing.orgInterface

Focus5:20140059

7

on May 23, 2018http://rsfs.royalsocietypublishing.org/Downloaded from

generated and natural waves, with reported survival rates in

the order of 50–60% at 6 h post-settlement.

Average germling settlement density for 12, 24 and 96 h

settlement was 20 germlings mm22. The fact that the same

initial germling density was observed across all substrates

and settlement times after 12 h indicates that the adhesion

strength achieved by 12 h is sufficiently high so that any vari-

ation in forces generated when transferring the substrates to

the flow channel become negligible.

3.4. Test of germling adhesion strengthFlow rates used in the flow experiments, Q, ranged between

50 and 300 ml min21, corresponding to mean fluid velocities,

v, between 0.4 and 2.5 m s21. The corresponding Reynolds

numbers were calculated using the formula

Re ¼ rvDH

m, (3:1)

where DH is the hydraulic diameter of the rectangular profile

of the flow channel

DH ¼4hw

2(hþ w), (3:2)

where h and w are the channel’s height and width, respect-

ively. Resulting Reynolds numbers ranged between 352

and 2111, indicative of laminar Poiseuille regime in the

flow channel across all flow rates tested. The theoretical

wall shear stress for laminar flow of Newtonian fluid, tw,

ranged between 5 and 32 Pa as determined using the

following relationship:

tw ¼6Qm

h2w: (3:3)

Despite the different geometry, our flow channel is compar-

able to other existing flow channels in terms of wall shear

stresses that can be achieved. For example, the turbulent flow

apparatus developed by Hodson et al. [25] is able to create wall

shear stresses up to 45 Pa, while the flow channel used by Schultz

et al. [24] generates wall stresses between 0.9 and 30 Pa. We

remark, once more, that thewall shear stress is not ideal for quan-

tifying the adhesion strength of sessile organisms, mainly

because it ignores the size and shape of the adhered organisms

as well as the actual flow field present. Nevertheless, we will

use shear stress to facilitate the comparison in adhesion strength

between H. banksii germlings and other benthic species.

As the substrates sampled after 6 h presented lower germ-

ling densities than the other settlement times, it was not

possible to compare the survival rates recorded in the 6 h treat-

ment with those observed after 12, 24 and 96 h. Nonetheless,

variations in germling densities observed during the flow

experiment on the substrates tested after 6 h are shown in

figure 5 to provide an indication of germling abundance

trends across increasing linear velocities. Despite the large vari-

ations in initial density at v ¼ 0, germling density decreases

rapidly once exposed to fluid flow, confirming very low

adhesion strength after a settlement period of only 6 h.

Figure 6 presents the experimental survival data of germl-

ings settled on different materials for different settlement

rsfs.royalsocietypublishing.orgInterface

Focus5:20140059

8

on May 23, 2018http://rsfs.royalsocietypublishing.org/Downloaded from

times (12, 24 and 96 h) and exposed to increasing flow vel-

ocities. Similar germling survival trends were observed on

all substrates. Germling survival decreased more rapidly

after a settlement period of 12 h than after the 24 and 96 h

treatments. Survival at the conclusion of the flow experiment,

i.e. after exposure to all flow velocities, was significantly

lower on the substrates tested 12 h after the settlement of

the germlings, compared with those examined after 24 and

96 h (F2,36 ¼ 32.23, p , 0.001). Survival rates did not differ

among materials (F3,36 ¼ 1.14, p ¼ 0.35) and were not affected

by the interaction between materials and settlement time

(F6,36 ¼ 0.89, p ¼ 0.51).

Relative adhesion strength is often established in rela-

tion to the wall shear stress required to remove 50% of

the germlings initially present, i.e. at 50% survival. Even

though the shear stresses at the wall generated by our

flow channel are of the same order of magnitude as other

flow apparatuses currently available, the flow conditions

were not sufficient to displace 50% of the germlings from

any of the different substrates after a post-settlement

period of 12 h or longer.

The lack of variation in survival across the various substrate

types suggests the adhesion mechanism used by H. banksiipropagules is promiscuous enough to engage with a range

of substrate types. While all of the surfaces were techni-

cally hydrophilic, there was no obvious impact on adhesion

strength caused by the relative differences in hydrophobicity

of the substrates. Similarities in adhesion strength could be

due to the presence of a range of functional groups present in

the adhesive material such as glycoproteins and anionic

polysaccharides [9,10], allowing interaction with a variety of

substrate types. Additionally, the functional groups presented

on the surface of the carbohydrate- and protein-based hydro-

gels, and the polar moieties of PMMA all provide a wealth of

locations for adsorption interactions. The large number of

hydroxyl groups present in the Agar substrate could result in

the formation of metal ion coordination complexes with anio-

nic adhesive components as well as haloperoxidase-mediated

cross-linked networks [9]. The high proportion of glycine and

arginine residues in gelatin [47] may facilitate the formation

of hydrogen bonds between a polar adhesive compound and

the substrate. This observation highlights the more consistent

adhesion strength of H. banksii germlings to different substrates

than that achieved by other marine species. For example, the

shear stress required to remove 50% of the cells of the diatoms

Navicula perminuta, Amphora coffeaeformis and Craspedostaurosaustralis were 25, 10 and 3 Pa, respectively, on glass slides,

but more than 53, 24 and 17 Pa, respectively, on poly-dimethyl-

siloxane elastomer substrate when given 2 h to settle [28].

Strong adhesion in the marine environment is also achieved

by mussels, which have been found to bind equally well to a

range of inorganic and organic surfaces [48]. The capacity to

adhere effectively to a variety of surface types is important to

exploit the large range of substrate chemistries presented by

the marine environment, including a range of inorganic

(e.g. rock) chemistries, and the organic EPSs ubiquitously

present due to bacterial biofilms.

3.5. Computational fluid dynamicsThe calculated Reynolds numbers suggest laminar flow across

all experimental conditions. However, both the side walls and

the inlet and outlet ports may generate specific disturbances

that may cause a mismatch with the ideal fully developed

flow in a rectangular channel. Preliminary CFD simulations

conducted using an empty channel evidenced that the entrance

length, Le, required to achieve fully developed parabolic flow

profile was, in the worst case, equal to 23 mm. This result is

in line with a calculated entrance length for rectangular ducts

of 19 mm, as determined using the empirical relationship

proposed by Han [49],

Le ¼ 0:01(DHRe): (3:4)

The CFD modelling also indicates exit effects were con-

fined to within 4 mm of the outlet orifice and, owing to the

small height to width ratio of the channel (h/w ¼ 0.125),

the influence of the two sides on the flow field was limited

to 0.3 mm. These results clearly highlight that fully devel-

oped Poiseuille flow was always present over the region

experimentally monitored.

In general, laminar flow over an object in contact with an

infinite plane produces three different external actions on

the object: a drag force oriented in the main direction of the

flow, FD, a lift force normal to the plane, FL, and a torque

with axis parallel to the plane and orthogonal to the main

direction of the flow, T [50,51]. In the ideal case of laminar

infinite linear shear over a single spherical body, the drag,

lift and torque exerted can be computed using the following

expressions [52,53]:

FD ¼ 32:0twr2, (3:5)

T ¼ 11:9twr3 (3:6)

and FL ¼ 9:257twr2 Rew (3:7)

where r is the radius of the sphere and Rew is the wall Rey-

nolds number, defined as

Rew ¼r _gwr2

m, (3:8)

where _gw ¼ tw=m is the wall shear rate. It is important to note

that equations (3.5)–(3.7) are valid only when the inertial

forces are negligible compared with the viscous forces, i.e.

for wall Reynolds numbers lower than 1. This condition

was nearly fulfilled in the experimental system investigated,

where Rew ranged between 0.9 and 5.4.

The local stresses present within the adhesive material and

at the pad–substrate interface highly depend on the mor-

phology and structure of the adhesive itself. Different

approaches have been proposed to model the stress distri-

bution at the interface, with the uniform stress model and the

peeling model frequently considered [54–56]. Despite its sim-

plicity, the uniform model, in which all the bonds within the

contact area are equally stressed, is able to provide a good esti-

mate of the order of magnitude of the interfacial stresses.

According to this model, the mechanical stresses present in

the attachment pad serve to counterbalance the acting forces.

In particular, the drag force will result in a uniform shear

stress, tD, the lift force will create a uniform normal stress,

sL, while the torque will produce a linear normal stress, sT,

zero at the contact point between the sphere and the surface

and maximum at the distal parts of the adhesive pad in

the direction of the flow [57]. If the adhesive pad has a circular

profile with the same dimensions as the sphere (note that

the latter assumption is not strictly required and a circular

adhesive pad with any dimension could be alternatively

used; however, this assumption is used to exemplify the

0

0.200.400.600.801.001.201.401.601.80

0.5 1.0velocity (m s–1)

single cell

multiple cells

F D (

N×

106 )

1.5 2.0 2.5

Figure 7. Comparison of the estimated drag force acting on isolated germl-ings and on random arrays (density of 20 germlings mm22, +s.e., n ¼ 5)at the different flow velocities investigated experimentally.

rsfs.royalsocietypublishing.orgInterface

Focus5:20140059

9

on May 23, 2018http://rsfs.royalsocietypublishing.org/Downloaded from

following relationships) then the stresses present in the

substrate–glue interface will be

tD ¼FD

pr2¼ 10:2tw, (3:9)

smaxT ¼ T

W¼ 15:0tw (3:10)

and sL ¼FL

pr2¼ 2:95twRew, (3:11)

where W ¼ pr3=4 is the elastic section modulus of the

circular profile.

Even though equations (3.9)–(3.11) are strictly valid for a

single sphere under infinite laminar flow, they can provide

a useful estimate of the order of magnitude of the stresses

acting on an isolated H. banksii germling present in the parallel

plate flow chamber. In particular, it is apparent that, at same

flow conditions (i.e. same wall shear stress), and under the

experimental wall Reynolds number, all stresses present in

the adhesive pad of the settled germlings have the same order

of magnitude. Hence, for an isotropic adhesive, calculation of

one of the stresses offers a simplified approach for the esti-

mation of the adhesion strength of the adhesive. Accordingly,

the drag force has been considered as a reference value in this

work and computed in the CFD simulations. A comparable

reduction of all stresses is assumed when the germlings are

part of an array. This assumption is useful as it allows the use

of the drag force as estimator to compare the adhesive strength

of germling populations settled at different surface densities.

The CFD estimations of the drag force for the isolated

germling and an array of settled germlings at the experimen-

tal surface density are reported in figure 7 for the different

wall shear stresses tested. It is apparent that the drag force

acting on an isolated cell is proportional to the average vel-

ocity in the flow channel. This behaviour is, for some

aspects, similar to the Stokes law for a falling sphere, where

the drag force is linear with the relative stream velocity

under creeping flow [51]. Experimental wall Reynolds num-

bers are consistent with creeping flow regime near the

channel walls. According to the uniform stress model and

for a contact area equal to the germling dimension, the

CFD computed stress ranges from 7 to 43 Pa in the case of

the isolated germling. These results are around one order of

magnitude lower than the shear stress that can be evaluated

using equation (3.9). The difference primarily lies in the

diversity of the geometries considered, semi-infinite space

in one case and Poiseuille flow in a rectangular conduct in

the other, which in turn will be characterized by different

velocity distributions and hence different resultant forces.

This investigation highlights the necessity of a generalized

model that can reliably predict the forces/stresses acting on

an organism within a parallel wall flow apparatus, i.e. at

different specifications of the channel width, average flow

velocity and organism dimensions, especially in the light of

its wide use in bioadhesion and biofouling [26–29].

Compared with the isolated sphere, the average drag force

exerted on the germlings within the random array had a con-

cave downward trend with increasing average velocity. At

higher velocities there is a notable reduction in drag force of

approximately 40%. The main reason for the lower average

drag force is identified as a consequence of the formation of

wakes, shielding the germlings settled downstream in the

flow channel (figure 8a). The protective action of the wakes

reduces the resultant of both the viscous and pressure forces.

Viscous forces, i.e. the forces tangential to the germling

surface, are associated with the velocity profile, with higher

forces in correspondence with sharper velocity gradients.

The highest gradients are observed on the two sides of the

frontal germlings, where the velocity transits from 0 in corre-

spondence with the germling periphery to its highest values

(coded by the colour red in figure 8a). Germlings in the wake

experience a much gentler velocity gradient, indicating a

smaller viscous force. On the other hand, pressure forces

are directed orthogonally to the spherical surface and are

related to the pressure profile. The drag associated with the

normal forces is mainly due to the pressure difference

between the front and the tail of a sphere. Figure 8c clearly

shows that this difference is less pronounced when the germ-

lings are shielded inside the wake.

Given the reduced average drag force affecting the germ-

lings within a colony of conspecifics, it is interesting to

quantify the average drag force acting on random arrays of

germlings at different surface densities. Figure 8 reported the

velocity and pressure profiles for random arrays of 5 and

20 germlings per mm2, while figure 9 shows the effect of differ-

ent surface densities on average drag force. In general, a lower

proportion of the settled germlings are shielded by wakes

when lower germling densities are present on the substrate.

Consequently, a clear decrease in the drag force is observed as

the germling density increases, corresponding to lower hydro-

dynamic stress acting on the settled germlings on both the xand the z directions. This observation indicates a strong

degree of germling–germling interdependence in adhesion suc-

cess, offering interesting ecological insights on the importance

of germling density in the face of hydrodynamic forces. The

role of density-dependent interactions among seaweed juvenile

individuals is highly debated and controversial, with both posi-

tive and negative effects being reported [15]. Our results suggest

that intraspecific density-dependent facilitation may be key for

surface colonization by H. banksii in natural contexts.

We recall that the CFD simulations were performed by

approximating the settled germlings with spherical bodies

in direct contact with the substrate. This assumption is most

accurate for approximating drag forces acting on the younger

germlings, before the morphology changes associated with

polarization and cell division occur. In addition, adhesion to

the substrate progressively shifted from a broad connec-

tion underneath the germling thallus to a more localized

attachment at the rhizoid tip, where the associated elongation of

the rhizoid allowed some germlings to shift into an upright

(a) (b)3.528

3.5

3.0

2.5

2.0

1.5

1.0

0.5

00

4858.74745.5

3689.8

2634

1578.2

522.43

–533.35

–1589.1

–2663.8–2663.8

3758.53676.5

2911.3

2146.2

1381.1

615.93

–149.21

–914.35

–1693.1–1693.1

3.334

3.0

2.5

2.0

1.5

1.0

0.5

00

(c) (d)

Figure 8. Velocity profiles (a,b) and pressure fields (c,d) with a random array of 20 germlings (a,c) and five germlings (b,d). The profiles are shown in a horizontalsection parallel to the flow direction and passing through the centre of the spheres. Average velocity ¼ 2.5 m s21.

0

0.200.400.600.801.001.201.401.601.80

5 10germling density (germlings mm–2)

FD

(N×

106 )

15 20 25

Figure 9. Effects of cell density on drag force at v ¼ 2.5 m s21 (+s.e., n ¼ 5).

rsfs.royalsocietypublishing.orgInterface

Focus5:20140059

10

on May 23, 2018http://rsfs.royalsocietypublishing.org/Downloaded from

position. These complications were not taken into account in

the current CFD work and represent the object of future studies.

4. ConclusionIn this paper, a novel flow channel to test the adhesion strength of

sessile marine organisms was presented. The adhesion strength

of the germlings of the brown alga H. banksii was assessed on var-

ious materials of environmental and biomedical relevance,

namely PMMA and protein- and carbohydrate-based hydrogels.

Germling development was similar on all the surfaces, with

polarization and rhizoid development beginning after 24 h

following their settlement. No adverse effects on development

were recorded on any of the substrates. Survival rates observed

using the flow channel were comparable across the different sub-

strates investigated, with an increase in adhesion strength with

settlement time. Survival increased greatly between 12 and

24 h, but did not vary between 24 and 96 h post-settlement.

This indicates that, under the experimental conditions investi-

gated in the present study, maximum adhesion strength was

achieved approximately 24 h following fertilization.

The CFD simulations were also generated to model the drag

forces acting on germlings, both as isolated organisms and as

part of colonies of varying surface densities. The simulations

showed that, under equivalent flow regimes, germlings within

an array experienced, on average, lower drag forces than the

drag force acting on a single germling. This suggests that

the presence of neighbouring individuals may help the germ-

lings to withstand challenging hydrodynamic conditions.

This same behaviour is expected to continue as the germlings

develop, where the interactions between nascent holdfasts

could also benefit survival. However, as a proportion of

the population gets washed out over time, the net result

of increasing adhesive strength and decreasing hydrodynamic

sheltering become more complicated. Similarly, substrates

with topographical features the same order of magnitude as

the settled germlings will help reduce the drag force acting on

them, even in the case of isolated germlings. The extent of

such reduction will ultimately depend on the morphology

and surface density of the asperities.

rsfs.royalsocietypublishing.org

11

on May 23, 2018http://rsfs.royalsocietypublishing.org/Downloaded from

Future developments of this research will investigate how the

chemical, physical and topographical properties of the settlement

surface can influence the adhesion strength of marine organisms.

The flow channel will be used to study the attachment of H. bank-sii and D. antarctica on different inorganic materials with varying

nano-topographies. In an effort to compare the relative strength

of underwater adhesives produced by different marine organ-

isms, other algal species will be considered as well as other

marine sessile organisms. These studies can contribute to the

characterization of the adhesive strength of the biological glue

produced by marine organisms and the assessment of the

adhesion force of natural, biomimetic and synthetic water-resist-

ant adhesives. The flow channel and the CFD model presented in

this work can also be used to assess the characteristics of antifoul-

ing surfaces developed to prevent attachment and colonization of

immersed man-made surfaces.

Acknowledgements. The authors thank Tim Huber of the BiomolecularInteraction Centre, University of Canterbury, New Zealand, for hisassistance with contact angle measurements.

Interfac

eFocus

5:20140059

1. Callow ME, Callow JA. 2002 Marine biofouling: asticky problem. Biologist 49, 10 – 14.

2. Morrison L, Feely M, Stengel DB, Blamey N, DockeryP, Sherlock A, Timmins A. 2009 Seaweedattachment to bedrock: biophysical evidencefor a new geophycology paradigm. Geobiology7, 477 – 487. (doi:10.1111/j.1472-4669.2009.00206.x)

3. Stewart RJ, Ransom TC, Hlady V. 2011Natural underwater adhesives. J. Polym. Sci. BPolym. Phys. 49, 757 – 771. (doi:10.1002/polb.22256)

4. Tarakhovskaya ER. 2014 Mechanisms of bioadhesionof macrophytic algae. Russ. J. Plant Physiol. 61,19 – 25. (doi:10.1134/S1021443714010154)

5. Gaylord B. 1999 Detailing agents of physicaldisturbance: wave-induced velocities andaccelerations on a rocky shore. J. Exp. Mar. Biol. Ecol.239, 85 – 124. (doi:10.1016/S0022-0981(99)00031-3)

6. Millar A, Kraft G. 1994 Catalogue of marine brownalgae (Phaeophyta) of New South Wales, includingLord Howe Island, south-western Pacific. Aust. Syst.Bot. 7, 1 – 47. (doi:10.1071/SB9940001)

7. Stevens CL, Hurd CL, Smith MJ, Harder D. 2001Durvillaea antarctica, the strongest kelp in theworld. Water Atmosphere. 9, 8 – 9.

8. Bitton R, Ben-Yehuda M, Davidovich M,Balazs Y, Potin P, Delage L, Colin C, Bianco-Peled H.2006 Structure of algal-born phenolic polymericadhesives. Macromol. Biosci. 6, 737 – 746.(doi:10.1002/mabi.200600073)

9. Vreeland V, Waite JH, Epstein L. 1998 Minireview—polyphenols and oxidases in substratum adhesionby marine algae and mussels. J. Phycol. 34, 1 – 8.(doi:10.1046/j.1529-8817.1998.340001.x)

10. Petrone L, Easingwood R, Barker MF,McQuillan AJ. 2011 In situ ATR-IR spectroscopicand electron microscopic analyses of settlementsecretions of Undaria pinnatifida kelp spores. J. R. Soc.Interface 8, 410 – 422. (doi:10.1098/rsif.2010.0316)

11. Percival E. 1979 The polysaccharides of green, redand brown seaweeds: their basic structure,biosynthesis and function. Br. Phycol. J. 14,103 – 117. (doi:10.1080/00071617900650121)

12. Schoenwaelder MEA, Clayton MN. 1998 Secretionof phenolic substances into the zygote wall andcell plate in embryos of hormosira andacrocarpia (Fucales, Phaeophyceae). J. Phycol.

34, 969 – 980. (doi:10.1046/j.1529-8817.1998.340969.x)

13. Stevens CL, Hurd CL, Smith MJ. 2002 Fieldmeasurement of the dynamics of the bull kelpDurvillaea antarctica (Chamisso) Heriot. J. Exp. Mar.Biol. Ecol. 269, 147 – 171. (doi:10.1016/S0022-0981(02)00007-2)

14. McKenzie PF, Bellgrove A. 2009 Dislodgment andattachment strength of the intertidal macroalgaHormosira banksii (Fucales, Phaeophyceae).Phycologia 48, 335 – 343. (doi:10.2216/08-96.1)

15. Schiel DR, Foster MS. 2006 The population biologyof large brown seaweeds: ecological consequencesof multiphase life histories in dynamic coastalenvironments. Annu. Rev. Ecol. Evol. Syst. 37,343 – 372. (doi:10.1146/annurev.ecolsys.37.091305.110251)

16. Osborn JE. 1948 The structure and life history ofHormosira banksii (Turner) Decaisne. Trans. R. Soc.NZ 77, 47 – 71.

17. Forbes MA, Hallam ND. 1979 Embryogenesis andsubstratum adhesion in the brown alga Hormosirabanksii (Turner) Decaisne. Br. Phycol. J. 14, 69 – 81.(doi:10.1080/00071617900650101)

18. Taylor DI, Schiel DR. 2003 Wave-related mortality inzygotes of habitat-forming algae from differentexposures in southern New Zealand: the importanceof ‘stickability’. J. Exp. Mar. Biol. Ecol.290, 229 – 245. (doi:10.1016/S0022-0981(03)00094-7)

19. Zhou J, Lin CG, Duan DX. 2011 Progress on measurementof adhesion strength of marine microfouling organisms.Adv. Mater. Res. 299– 300, 883 – 886. (doi:10.4028/www.scientific.net/AMR.299-300.883)

20. Finlay JA, Callow ME, Schultz MP, Swain GW, CallowJA. 2002 Adhesion strength of settled spores of thegreen alga enteromorpha. Biofouling 18, 251 – 256.(doi:10.1080/08927010290029010)

21. Taylor D, Delaux S, Stevens C, Nokes R, Schiel D.2010 Settlement rates of macroalgal propagules:cross-species comparisons in a turbulentenvironment. Limnol. Oceanogr. 55, 66 – 76.(doi:10.4319/lo.2010.55.1.0066)

22. Arpa-Sancet MP, Christophis C, Rosenhahn A. 2012Microfluidic assay to quantify the adhesion ofmarine bacteria. Biointerphases 7, 26.

23. Barberousse H, Brayner R, Do Rego AMB, CastaingJ-C, Beurdeley-Saudou P, Colombet JF. 2007

Adhesion of facade coating colonisers, as mediatedby physico-chemical properties. Biofouling 23,15 – 24. (doi:10.1080/08927010601093026)

24. Schultz MP, Finlay JA, Callow ME, Callow JA. 2000A turbulent channel flow apparatus for thedetermination of the adhesion strength ofmicrofouling organisms. Biofouling 15, 243 – 251.(doi:10.1080/08927010009386315)

25. Hodson OM, Monty JP, Molino PJ, Wetherbee R.2012 Novel whole cell adhesion assays of threeisolates of the fouling diatom Amphoracoffeaeformis reveal diverse responses to surfaces ofdifferent wettability. Biofouling 28, 381 – 393.(doi:10.1080/08927014.2012.680020)

26. Cao X et al. 2010 Interaction of zoospores of thegreen alga ulva with bioinspired micro- andnanostructured surfaces prepared by polyelectrolytelayer-by-layer self-assembly. Adv. Funct. Mater. 20,1984 – 1993. (doi:10.1002/adfm.201000242)

27. Finlay JA et al. 2010 Barnacle settlement and theadhesion of protein and diatom microfouling toxerogel films with varying surface energy and waterwettability. Biofouling 26, 657 – 666. (doi:10.1080/08927014.2010.506242)

28. Holland R, Dugdale TM, Wetherbee R, Brennan AB,Finlay JA, Callow JA. 2004 Adhesion and motility offouling diatoms on a silicone elastomer. Biofouling20, 323 – 329. (doi:10.1080/08927010400029031)

29. Martinelli E, Suffredini M, Galli G, Glisenti A, PettittME, Callow ME, Williams D, Lyall G. 2011Amphiphilic block copolymer/poly(dimethylsiloxane)(PDMS) blends and nanocomposites for improvedfouling-release. Biofouling 27, 529 – 541. (doi:10.1080/08927014.2011.584972)

30. Olivier LA, Truskey GA. 1993 A numerical analysis offorces exerted by laminar flow on spreading cells ina parallel plate flow chamber assay. Biotechnol.Bioeng. 42, 963 – 973. (doi:10.1002/bit.260420807)

31. Bai B, Luo Z, Lu T, Xu F. 2012 Numerical simulationof cell adhesion and detachment in microfluidics.J. Mech. Med. Biol. 13, 1350002. (doi:10.1142/S0219519413500024)

32. Gaver DP, Kute SM. 1998 A theoretical model studyof the influence of fluid stresses on a cell adheringto a microchannel wall. Biophys. J. 75, 721 – 733.(doi:10.1016/S0006-3495(98)77562-9)

33. Lu H, Koo LY, Wang WM, Lauffenburger DA, GriffithLG, Jensen KF. 2004 Microfluidic shear devices for

rsfs.royalsocietypublishing.orgInterface

Focus5:20140059

12

on May 23, 2018http://rsfs.royalsocietypublishing.org/Downloaded from

quantitative analysis of cell adhesion. Anal. Chem.76, 5257 – 5264. (doi:10.1021/ac049837t)

34. Brooks SB, Tozeren A. 1996 Flow past an array ofcells that are adherent to the bottom plate of aflow channel. Comput. Fluids 25, 741 – 757. (doi:10.1016/S0045-7930(96)00024-2)

35. Underwood AJ. 1996 Experiments in ecology:their logical design and interpretation using analysisof variance. New York, NY: Cambridge UniversityPress.

36. Cox RA, McCartney MJ, Culkin F. 1970 The specificgravity/salinity/temperature relationship in naturalsea water. Deep Sea Res. Oceanogr. Abstr. 17,679 – 689. (doi:10.1016/0011-7471(70)90034-3)

37. Leigh JA. 1975 Use of PMMA in expansion dentalimplants. J. Biomed. Mater. Res. 9, 233 – 242.(doi:10.1002/jbm.820090426)

38. Frazer RQ, Byron RT, Osborne PB, West KP. 2005PMMA: an essential material in medicine anddentistry. J. Long Term Eff. Med. Implants. 15,629 – 639. (doi:10.1615/JLongTermEffMedImplants.v15.i6.60)

39. Lir I, Haber M, Dodiuk-Kenig H. 2007 Skin surfacemodel material as a substrate for adhesion-to-skintesting. J. Adhesion Sci. Technol. 21, 1497 – 1512.(doi:10.1163/156856107782844783)

40. Yoshizawa K, Taguchi T. 2014 Enhanced bondingstrength of hydrophobically modified gelatin filmson wet blood vessels. Int. J. Mol. Sci. 15,2142 – 2156. (doi:10.3390/ijms15022142)

41. Bertoni F, Barbani N, Giusti P, Ciardelli G. 2006Transglutaminase reactivity with gelatine:perspective applications in tissue engineering.Biotechnol. Lett. 28, 697 – 702. (doi:10.1007/s10529-006-9046-2)

42. Durr S, Thomason J. 2009 Biofouling. Hoboken, NJ:Wiley.

43. Fletcher RL, Callow ME. 1992 The settlement,attachment and establishment of marine algalspores. Br. Phycol. J. 27, 303 – 329. (doi:10.1080/00071619200650281)

44. Callow ME, Callow JA, Ista LK, Coleman SE, NolascoAC, Lopez GP. 2000 Use of self-assembledmonolayers of different wettabilities to studysurface selection and primary adhesion processes ofgreen algal (Enteromorpha) zoospores. Appl.Environ. Microbiol. 66, 3249 – 3254. (doi:10.1128/AEM.66.8.3249-3254.2000)

45. Kim K, Park SW, Yang SS. 2010 The optimization ofPDMS-PMMA bonding process using silane primer.BioChip J. 4, 148 – 154. (doi:10.1007/s13206-010-4210-0)

46. Bitton R, Berglin M, Elwing H, Colin C, Delage L,Potin P, Bianco-Peled H. 2007 The influence ofhalide-mediated oxidation on algae-born adhesives.Macromol. Biosci. 7, 1280 – 1289. (doi:10.1002/mabi.200700099)

47. Eastoe JE. 1955 The amino acid composition ofmammalian collagen and gelatin. Biochem. J.61, 589.

48. Lee H, Scherer NF, Messersmith PB. 2006 Single-molecule mechanics of mussel adhesion. Proc. NatlAcad. Sci. USA 103, 12 999 – 13 003. (doi:10.1073/pnas.0605552103)

49. Han LS. 1960 Hydrodynamic entrance lengths forincompressible laminar flow in rectangular ducts.J. Appl. Mech. 27, 403 – 409. (doi:10.1115/1.3644015)

50. Boulbene B, Morchain J, Bonin MM, Janel S,Lafont F, Schmitz P. 2012 A combined

computational fluid dynamics (CFD) andexperimental approach to quantify the adhesionforce of bacterial cells attached to a planesurface. AIChE J. 58, 3614 – 3624. (doi:10.1002/aic.13747)

51. Bird RB, Stewart WE, Lightfoot EN. 2007 Transportphenomena. New York, NY: John Wiley & Sons.

52. O’Neill ME. 1968 A sphere in contact with aplane wall in a slow linear shear flow. Chem. Eng.Sci. 23, 1293 – 1298. (doi:10.1016/0009-2509(68)89039-6)

53. Krishnan GP, Leighton Jr DTL. 1995 Inertial lift on amoving sphere in contact with a plane wall in ashear flow. Phys. Fluids 7, 2538 – 2545. (doi:10.1063/1.868755)

54. Gallant ND, Garcıa AJ. 2007 Model of integrin-mediated cell adhesion strengthening. J. Biomech.40, 1301 – 1309. (doi:10.1016/j.jbiomech.2006.05.018)

55. Hodges SR, Jensen OE. 2002 Spreading and peelingdynamics in a model of cell adhesion. J. Fluid Mech.460, 381 – 409. (doi:10.1017/S0022112002008340)

56. Chan BP, Bhat VD, Yegnasubramanian S, ReichertWM, Truskey GA. 1999 An equilibrium model ofendothelial cell adhesion via integrin-dependentand integrin-independent ligands. Biomaterials 20,2395 – 2403. (doi:10.1016/S0142-9612(99)00167-2)

57. Beer F, Johnston J, Russell E, DeWolf J, Mazurek D.2011 Mechanics of materials, 6th edn. New York,NY: McGraw-Hill.

58. Yasuda H, Sharma AK, Yasuda T. 1981 Effect oforientation and mobility of polymer molecules atsurfaces on contact angle and its hysteresis.J. Polym. Sci. Polym. Phys. Ed. 19, 1285 – 1291.(doi:10.1002/pol.1981.180190901)