Diffusion or advection? Mass transfer and complex boundary layer landscapes of the brown alga Fucus vesiculosus

Abstract

The role of hyaline hairs on the thallus of brown algae in the genus Fucus is long debated and several functions have been proposed. We used a novel motorized set-up for two-dimensional and three-dimensional mapping with O₂ microsensors to investigate the spatial heterogeneity of the diffusive boundary layer (DBL) and O₂ flux around single and multiple tufts of hyaline hairs on the thallus of Fucus vesiculosus. Flow was a major determinant of DBL thickness, where higher flow decreased DBL thickness and increased O₂ flux between the algal thallus and the surrounding seawater. However, the topography of the DBL varied and did not directly follow the contour of the underlying thallus. Areas around single tufts of hyaline hairs exhibited a more complex mass-transfer boundary layer, showing both increased and decreased thickness when compared with areas over smooth thallus surfaces. Over thallus areas with several hyaline hair tufts, the overall effect was an apparent increase in the boundary layer thickness. We also found indications for advective O₂ transport driven by pressure gradients or vortex shedding downstream from dense tufts of hyaline hairs that could alleviate local mass-transfer resistances. Mass-transfer dynamics around hyaline hair tufts are thus more complex than hitherto assumed and may have important implications for algal physiology and plant–microbe interactions.

1. Introduction

Compared with terrestrial plants, aquatic macrophytes experience approximately 10⁴ slower diffusion and a much lower solubility of gases in water than in air [1,2]. The efficient exchange of nutrients and gases is further exacerbated by the diffusive boundary layer (DBL) surrounding all submersed surfaces [3]. The thickness and topography of DBLs around submerged impermeable objects is affected by flow velocity and surface topography [3]. Higher flow velocities decrease the DBL thickness by exerting a higher shear stress on the viscous sublayer near the surface. The effect of surface roughness on DBL thickness is variable, where angled planes facing the flow generally will have decreased boundary layers, while the DBL downstream of protruding structures will have increased boundary layers [4]. At the same time, the effect of surface roughness on mass transfer across a boundary layer is dichotomous, where a thicker DBL will decrease mass transfer, while surface roughness tends to increase the overall surface area, thus increasing mass transfer [3]. In microsensor-based studies of the DBL, it is important to note that the presence of the microsensor tip in itself can affect the local DBL thickness [5], where flow acceleration around the microsensor shaft will compress the local DBL thickness, leading to locally enhanced O₂ fluxes of the order of 10%. However, this effect is only significant when investigated on smooth surfaces, while a clear effect is apparently undetectable when measuring over tufts in, for example, a cyanobacterial mat [6].

It has been estimated that the DBL accounts for up to 90% of the resistance to carbon fixation in freshwater plants [7], and both structural and biochemical regulations to alleviate such mass-transfer resistance have evolved across lineages. Some aquatic macrophytes have, for example, developed (i) thinner leaves and a reduced cuticle, decreasing the diffusion path length to chloroplasts, (ii) carbon-concentrating mechanisms that increase internal CO₂ concentration, and (iii) the ability to utilize , which constitutes the largest fraction of dissolved inorganic carbon at ocean pH ([8,9] and references therein). In photosymbiotic corals, it has also been proposed that vortical ciliary flow can actively enhance mass transfer between the coral tissue and the surrounding water in stagnant or very-low-flow regimes with concomitant thick DBLs [10].

Hyaline hairs, that is, colourless, filamentous multicellular structures, are often present as whitish tufts on the thallus of brown macroalgae in the genus Fucus. The hairs originate in so-called cryptostomata, that is, cavities on the apical and mid-regions of the thallus [11]. It is recognized that hyaline hairs aid in the uptake of nutrients [12,13], for example during springtime, when photosynthetic potential is higher due to increased light levels, and the need for nutrients apparently triggers growth of hyaline hairs [11].

How hyaline hairs affect solute exchange and nutrient acquisition in Fucus is still debated, and different functional roles have been suggested in the literature: (i) the hairs might increase the algal surface area available for nutrient uptake, albeit this is probably not the major limitation on nutrient uptake [12]; (ii) the hyaline hairs might decrease the DBL due to turbulence created by the hairs as water flows across them, decreasing the mass-transfer resistance imposed by the DBL [12]; (iii) the thin cell walls of the hairs relative to the thallus could have less resistance to the passage of ions [14,15]; (iv) the hyaline hairs increase DBL thickness, thereby retaining the products of thallus surface-active enzymes such as extracellular phosphatases and ensuring more efficient nutrient uptake [15].

There can be no doubt, however, that the DBL has great importance for macroalgal growth rates. The mass-transfer resistance imposed by the DBL has been correlated with nutrient limitation in the giant kelp Macrocystis pyrifera [16], and a considerable spatial variation of the DBL over the thallus and cryptostomata of Fucus vesiculosus has been observed [17]. However, current knowledge of the DBL characteristics of aquatic plants is largely based on point measurements with O₂ microsensors [15], while it is known from boundary layer studies in biofilms [4], corals [18–20] and sediments [4,21–23] that the DBL exhibits a spatio-temporal heterogeneity that is modulated by both flow velocity and surface topography. Similar studies of DBL topography are very limited in aquatic plant science [24], and the aim of this study was to explore how the DBL thickness and the local O₂ flux varied spatially over the thallus of F. vesiculosus with and without tufts of hyaline hairs. The exploration of the three-dimensional (3D) boundary layer topography was carried out with O₂ microsensors mounted in a fully automated motorized micromanipulator system that allowed measurements of O₂ concentration gradients in transects and grids over the thallus of F. vesiculosus (electronic supplementary material, figure S1). Our results reveal a complex boundary layer landscape over the algal thallus, where mass transfer across the DBL apparently can be supplemented by advective processes due to the presence of hyaline hairs.

2. Material and methods

2.1. Sampling and experimental set-up

Specimens of F. vesiculosus and seawater used in the experimental set-up were sampled on the day of usage at <1 m depth at Kronborg, Helsingør, Denmark, from May to August. When studying the influence of a single tuft of hyaline hairs on the mass-transfer boundary layer, excess hair tufts were carefully shaved off with a scalpel during observation under a dissection microscope to avoid thallus damage. Previous studies showed that mechanical removal of hyaline hairs does not induce defence mechanisms in Fucus [25]. When the influence of multiple tufts was analysed, the thallus was left intact. Prior to measurements, a piece of F. vesiculosus thallus with hyaline hairs was fixed on a slab of agar (approx. 1.5% w/w in seawater) in a small flow chamber, creating a defined unidirectional flow of seawater across the thallus surface [17]. The flow chamber was connected via tubing to a submersible water pump in a continuously aerated and thermostated seawater reservoir tank underneath the flow chamber. Flow velocity was adjusted by restricting water flow to the flow chamber with a needle valve. By collecting water from the flow chamber outlet for 1 min and dividing the sampled volume per time by the cross-sectional area of the flow chamber we could estimate the mean free-flow velocity. All measurements were carried out with mean free-flow velocities in the flow chamber of either 1.65 cm s⁻¹ or 4.88 cm s⁻¹.

The sample was illuminated from above with light from a halogen lamp (Schott KL-2500LCD), equipped with a collimating lens, that produces a bell-shaped wavelength distribution in the PAR region (see the electronic supplementary material in [26] for the spectrum). The spectral light composition was constant in all treatments and yielded a downwelling photon irradiance (400–700 nm) of approximately 350 µmol photons m⁻² s⁻¹, as measured with a quantum irradiance meter (LI-250; LiCor Inc., USA).

2.2. Microsensor measurements

Measurements of O₂ concentration above the thallus of F. vesiculosus were carried out with Clark-type O₂ microelectrodes (tip diameter 10 μm; OX10; Unisense A/S, Denmark; [27]) with a response time of less than 1–3 s and low stirring sensitivity (less than 2%). The microelectrode was connected to a pA meter (PA2000; Unisense A/S, Denmark), and sensor signals from the pA meter were acquired on a PC via a parallel port-connected A/D converter (ADC-101; Pico Technologies Ltd, UK). The O₂ microsensor was mounted in a custom-built micromanipulator set-up enabling motorized positioning at defined x, y and z coordinates at approximately 1 µm resolution by use of three interconnected motorized positioners (VT-80; Micos GmbH, Germany) and controllers (MoCo DC; Micos GmbH, Germany). Data acquisition and positioning were controlled by a custom-built software (Volfix) programmed in LabView (National Instruments, Japan).

The O₂ microelectrode signal was linearly calibrated at experimental temperature (approx. 17°C) and salinity (S = 16) from measurements in air-saturated seawater and in seawater made anoxic by the addition of sodium dithionite. The O₂ concentration in air-saturated seawater, C₀, and the molecular diffusion coefficient of O₂, D₀, in seawater at experimental temperature and salinity were taken from tabulated values (Unisense A/S, Denmark) as C₀ = 274 μmol O₂ l⁻¹ and D₀ = 1.87 10⁻⁵ cm² s⁻¹. No significant sensor drift was observed during measurements as judged from the signal stability in the constantly aerated thermostated water above the DBL between profile measurements.

2.3. Mapping of boundary layers

The boundary layer around tufts of hyaline hairs anchored in the cryptostomata of F. vesiculosus was mapped by two-dimensional (2D) transect and 3D grid measurements of O₂ concentration profiles towards the thallus surface. The Volfix software enabled us to specify a measuring grid/transect with any number of sampling points in the x-, y- and z-directions. In this study, the y-direction corresponds to the direction of flow (where negative values indicate the distance behind a single tuft of hyaline hairs), the x-direction corresponds to the width of the flow chamber, and the z-direction corresponds to the height above the thallus surface (electronic supplementary material, figure S1). The approximate height and radius of the hyaline hairs were estimated by manual manipulation of the microelectrode tip relative to the structures as observed under a stereomicroscope (SV6; Zeiss, Germany). Thallus samples were placed in the flow chamber with the length of the thallus oriented along the direction of flow, that is, in the y-direction.

2.4. Two-dimensional transect measurements

For 2D transect measurements, the O₂ microelectrode tip was positioned manually as close to the centre of the selected cryptostomata as possible using a dissection microscope for observation; this position was set to y = 0 in the Volfix measuring software. The transect measurements started 2 mm upstream (y = 2 mm) from the cryptostomata and 1 mm above the hyaline hairs (point A in the electronic supplementary material, figure S1D), and ended 4 mm downstream (y = −4 mm). Transects of O₂ concentration profiles were measured at a lateral resolution of 0.5 mm in the y-direction, with vertical O₂ concentration profiles measured at each transect point in steps of 0.1 mm in the z-direction. All profile measurements started at the same z-position and ended in the upper thallus layer, where a characteristic jump in the O₂ concentration, due to the physical impact of the O₂ microsensor and the solid thallus surface, enabled precise determination of the thallus surface.

2.5. Three-dimensional grid measurements

For measuring 3D grids of O₂ concentration profiles over thallus areas with only one tuft of hyaline hairs, nine transects covering a 24 mm² sampling grid area were measured around a central tuft of hairs (electronic supplementary material, figure S1E) with a lateral resolution of 0.5 mm (x- and y-directions) and a vertical resolution of 0.1 mm (z-direction). Measurements started approximately 1 mm above the hyaline hairs (electronic supplementary material, figure S1E).

For measurements of 3D grids of O₂ concentration profiles over larger thallus areas with multiple tufts of hyaline hairs, a 12 × 2 mm measuring grid was set up and measurements were performed at a lateral resolution of 0.5 mm (x- and y-directions) and a vertical resolution of 0.2 mm (z-direction). Tufts of hyaline hairs were scattered across the thallus, and the starting point for the grid measurement was set randomly, but with the same starting point for measurements at a flow velocity of 1.65 cm s⁻¹ and 4.88 cm s⁻¹.

Owing to the length of measurements, different fresh thallus samples were used for individual 2D and 3D experiments.

2.6. Diffusive boundary layer thickness and calculations

There are different ways of determining the effective thickness of the DBL from O₂ microsensor measurements [4]. The DBL thickness is often found by extrapolating the linear O₂ gradient in the DBL to the bulk concentration of the free-flow region. The distance from the surface to the intersection of the extrapolated linear gradient and the bulk concentration denoted the effective DBL thickness, Z_δ [3]. However, analysing large numbers of O₂ profiles in this manner is very time consuming, and a somewhat faster determination can be carried out by defining Z_δ as the distance between the surface and the vertical position above the surface where the O₂ concentration has changed by 10% relative to the O₂ concentration in the bulk water. Estimations of Z_δ via this method were found to differ by less than 10% from more precise determinations [4].

The diffusive flux of O₂ across the DBL, J (in units of nmol O₂ cm⁻² s⁻¹), was calculated from steady-state O₂ concentration profiles using Fick's first law,

2.1

where C_∞ is the O₂ concentration in the free-flow region (µmol O₂ l⁻¹ = nmol O₂ cm⁻³), C₀ is the O₂ concentration at the thallus surface (µmol O₂ l⁻¹), Z_δ is the effective DBL thickness (cm) and D₀ is the molecular diffusion coefficient of O₂ in seawater (cm² s⁻¹).

We note that the fluxes calculated with Fick's first law are only valid in a flow regime with diffusive mass transport and negligible advection close to the thallus surface. A thicker DBL will be present under laminar flow conditions, but even under turbulent flow a thin diffusive sublayer will be present lining the thallus surface [15]. We also note that our DBL thickness determination method (see above) can yield overestimates in flow regimes exhibiting a broad transition layer between the DBL and the fully mixed turbulent water phase. That is, the region where the measured O₂ concentration profiles show a curvature towards the constant O₂ concentration in the mixed water, and where there is a gradual transition from mass transfer dominated by molecular diffusion towards a predominance of eddy diffusivity. Such conditions can potentially lead to an underestimation of local fluxes calculated with equation (2.1).

To compensate for the uneven surface of thalli, measurements below the thallus surface are not shown on final transects. The depth axis on transects is therefore denoted z′ = z − z₀, where z is the z-coordinate from the sample data and z₀ is the z-coordinate of the thallus surface. The thallus surface position was determined from the intermittent sudden drop in O₂ concentration when the microsensor pushed against the thallus surface cortex. Maps of O₂ concentration, Z_δ and J were generated from measured transects and grids using data interpolation software (kriging gridding method using default settings, i.e. Linear Variogram and Point Kriging, Surfer v. 8; Golden Software Inc., USA).

2.7. Statistics

Two-way ANOVAs tested differences in the mean Z_δ over F. vesiculosus between flow rates and light conditions (light/dark). For significant main effects (flow rate and/or light condition) and interaction effects, Tukey's multiple comparisons post hoc test was applied. Two-way ANOVAs tested the differences between mean O₂ flux values between flow rates and thallus condition (single or multiple tufts). For significant main effects (flow rate and/or thallus condition) and interaction effects, Tukey's multiple comparisons post hoc test was applied. Statistical analysis was performed using Rstudio (Rstudio v. 0.99.491, 2016) with the level of significance set to p < 0.05.

3. Results

3.1. The diffusive boundary layer around single tufts of hyaline hairs

Isopleths of O₂ concentration in the water column above the thalli showed a local increase in effective DBL thickness, Z_δ, associated with the hyaline hair tuft (figure 1). The highest Z_δ values were located downstream from the hyaline hairs, either directly behind the tuft or even within the expanse of the hyaline hairs. In light (350 µmol photons m⁻² s⁻¹), and at a flow of 1.65 cm s⁻¹, Z_δ reached a maximum thickness of 1.2 mm downstream relative to the tuft at y = −0.5 mm (electronic supplementary material, figures S2 and S3A), while under a flow of 4.88 cm s⁻¹ the maximum Z_δ was reduced to 0.6 mm at y = −2.5 mm (electronic supplementary material, figure S3A). In darkness, the maximal Z_δ values were 0.9 mm and 0.4 mm at y = −2 mm and y = 0 mm under flows of 1.65 and 4.88 cm s⁻¹, respectively. The mean Z_δ did not change significantly (p < 0.05) between measurements in light and darkness (electronic supplementary material, figure S3b) under low flow (Z_δ(Light) = 0.72 mm, Z_δ(Dark) = 0.58 mm; p adj = 0.26) or high flow (Z_δ(Light) = 0.36 mm, Z_δ(Dark) = 0.18 mm; p adj = 0.13). However, flow velocity had a significant effect on the mean Z_δ, which was significantly thinner under high flow (4.88 cm s⁻¹) than under low flow (1.65 cm s⁻¹) (p adj < 0.001 for both main effects).

Figure 1.

Fine-scale mapping of DBL around a single tuft of hyaline hairs on an illuminated F. vesiculosus thallus (350 µmol photons m⁻² s⁻¹). (a) and (b) 3D plots of F. vesiculosus thallus surface (grey area) and upper extension of DBL (coloured area) around a single tuft of hyaline hairs, at flow velocities of 1.65 (a,c,e) and 4.88 cm s⁻¹ (b,d,f). Colour bars depict the effective DBL thickness, Z_δ (mm), and arrows indicate the direction of flow. (c) and (d) Transects in the x-direction (perpendicular to the flow), at position y = −2 mm from (a) and (b), respectively, normalized to the thallus surface showing the local O₂ concentration. The zero position (0, 0) indicates the position of the cryptostomata. Colour bars denote O₂ concentration (in µmol O₂ l⁻¹). (e) and (f) Transects of O₂ concentration (in µmol l⁻¹) measured across a single tuft of hyaline hairs in F. vesiculosus measured at flow velocities of 1.65 (e) and 4.88 cm s⁻¹ (f), in light (350 µmol photons m⁻² s⁻¹). The arrows indicate the flow direction. The zero position (0, 0) indicates the position of the cryptostomata, and transects were adjusted to the thallus surface. Colour bars denote O₂ concentration (in µmol O₂ l⁻¹).

The hyaline hairs affected Z_δ downstream from the hair tuft (figure 1a,b) and caused a thickening of the boundary layer that also expanded perpendicular to the flow direction (figure 1c,d), reaching a maximum expansion at y = −2 mm for both flows (Z_δMax = 1.8 mm for 1.65 cm s⁻¹ and Z_δMax = 0.9 mm for 4.88 cm s⁻¹). Beyond the local peak in boundary layer thickness, the DBL closely followed the contours of the thallus surface topography. Two transects measured at higher resolution along the x-axis at y = −2 mm showed that the increase in Z_δ was roughly identical and extended approximately 1 mm on both sides of the hyaline hair tuft (figure 1c,d). At distances more than 1 mm away from the local maximum, the DBL approached a lower, more homogeneous thickness over thallus areas unaffected by the hair tuft.

Unexpectedly, a transect of O₂ concentrations above the illuminated thallus at y = −2 mm in the x-direction, that is, perpendicular to the flow, showed a local area of increased O₂ concentration apparently separated from the DBL (figure 1c). A longitudinal transect at x = −0.5 mm along the y-direction, that is, in the flow direction, showed further indications of an apparent local ‘upwelling’ of O₂ into the transition zone between the DBL and the fully mixed water column downstream from the hyaline hair tuft (figure 2). We found such ‘upwelling’ zones most pronounced under low flow located approximately 2 mm downstream from the centre of the tuft and extending several millimetres into the water column with O₂ concentrations reaching up to more than two times air saturation in some cases (figure 2a).

Figure 2.

Local transects of O₂ concentration around single hair tufts from two different measurement series over an illuminated F. vesiculosus thallus (350 µmol photons m⁻² s⁻¹). Panel (a) shows a transect taken from figure 1a at x = −0.5 mm under a flow velocity of 1.65 cm s⁻¹, while panel (b) was measured similarly to figure 1e, also at a flow velocity of 1.65 cm s⁻¹. The hair tufts were 2.5–3 mm in diameter and protruded 3–3.5 mm from the thallus. Both transects were adjusted to the thallus surface. The black arrow indicates the flow direction. Colour bars denote O₂ concentration (in µmol O₂ l⁻¹).

3.2. The boundary layer around multiple tufts of hyaline hairs

To investigate the combined effects of multiple tufts on the boundary layer, 3D grid measurements of O₂ concentration were carried out over a F. vesiculosus thallus with several tufts of hyaline hairs spaced at approximately 2–5 mm distance. Such measurements showed that the smooth local thickening of the DBL around a single hyaline tuft relative to the DBL over the smooth thallus was altered in the presence of multiple tufts (figure 3). The boundary layer topography was more heterogeneous, with Z_δ varying by more than 1 mm and reaching a maximum thickness of more than 2.5 mm under low flow and more than 1.5 mm under high flow. The DBL topography was thus strongly determined by the interaction between flow and the tufts of hyaline hairs under low flow, while we observed local minima in Z_δ between individual tufts at higher flow velocity (figure 3b).

Figure 3.

Boundary layer mapping over several tufts of hyaline hairs on an illuminated F. vesiculosus thallus (350 µmol photons m⁻² s⁻¹). (a) and (b) 3D plot of F. vesiculosus thallus surface (grey area) and the upper extension of the DBL (coloured area) of multiple tufts of hyaline hairs under a flow velocity 1.65 (a,c) and 4.88 cm s⁻¹ (b,d). Colour bars depict the effective DBL thickness, Z_δ (mm). (c) and (d) Transects of O₂ concentration at position x = 1 mm (along the y-axis direction) normalized to the thallus surface. Colour bars denote O₂ concentration (in µmol O₂ l⁻¹).

Transects of O₂ concentrations at x = 1 mm (extracted from the 3D grids in figure 3a,b) gave detailed information on how the O₂ concentration varied over the thallus with distance along the thallus in the flow direction (figure 3c,d). In light, the thallus surface O₂ concentration reached more than 900 µM in both flows, while the O₂ concentration in the transient zone of the DBL (z = 0.7 mm) varied between 350 and 750 µM under low flow and between 300 and 550 µM under high flow. This demonstrated an overall compression of the boundary layer and a more effective O₂ exchange between thallus and water under higher flow.

However, the thickening of the 300–350 µM O₂ contour areas, for example at y = −11 mm and y = −6 mm (figure 3c), was due to gradually increasing O₂ concentrations from the bulk water towards the upper part of the DBL (data not shown). This creates an artefact in the precise determination of Z_δ by the method proposed by Jørgensen & Des Marais [4] that will overestimate the local DBL thickness, for example, compared with the local profile in y = −1 mm where a more steady O₂ increase was measured.

3.3. Diffusive O₂ fluxes over the Fucus thallus with single and multiple tufts

Although inconsistencies were found (e.g. in the area around x = −1.5 mm, y = 1.5 mm in figure 4a,b), increases in DBL thickness generally correlated with a decrease in O₂ flux, and the flow-dependent boundary layer topography strongly affected the O₂ flux from the illuminated F. vesiculosus thallus.

Figure 4.

(a) and (c) Isopleths of O₂ flux (in nmol O₂ cm⁻² s⁻¹) and (b) and (d) the effective DBL thickness, Z_δ (in mm), measured over an illuminated F. vesiculosus thallus (350 µmol photons m⁻² s⁻¹) around a single tuft of hyaline hairs at flow velocities of 1.65 cm s⁻¹ (a,b) and 4.88 cm s⁻¹ (c,d). The hyaline hairs were rooted in the cryptostomata located at the (0, 0) coordinate, as indicated by the black cross. Black arrows indicate the flow direction.

Comparing the O₂ fluxes in transects over the F. vesiculosus thallus with single and multiple tufts of hyaline hairs (figure 5a,b) showed an increased O₂ flux just upstream from the position of the hyaline hair tufts independent of the flow velocity. The flux values generally correlated with the boundary thickness and the apparent O₂ flux gradually decreased downstream relative to the hair tuft. However, local variations were found in areas exhibiting less uniform increases in O₂ concentration towards the thallus surface.

Figure 5.

Comparison of O₂ flux values (in nmol O₂ cm⁻² s⁻¹) calculated from transects of O₂ concentration profiles measured over an illuminated (350 µmol photons m⁻² s⁻¹) intact F. vesiculosus thallus with several tufts of hyaline hairs (a) and a thallus with only a single hair tuft (b) measured under flow velocities of 1.65 cm s⁻¹ and 4.88 cm s⁻¹. Note the difference in the x-scale. The black arrow indicates the flow direction. The individual position of the multiple hair tufts in (a) were not mapped and the zero position on the x-axis thus only reflects the starting point of the transect. In (b), the zero position indicates the centre of the cryptostomata. (c) and (d) The average O₂ flux (±s.e.m.) across (c) the intact thallus and (d) the thallus with a single hair tuft protruding.

The average O₂ flux calculated from transects over F. vesiculosus thalli with single and multiple hyaline hair tufts (figure 5c,d) showed that flow was the major determinant of gas exchange between the thallus and the surrounding seawater. The O₂ flux values were higher in high-flow treatments than in low-flow treatments (p adj < 0.001) in measurements over both single and multiple hair tufts. The O₂ fluxes measured around a single hair tuft under high flow were higher than the corresponding measurements over multiple tufts (figure 5c,d; p adj < 0.001). However, the apparent flux values in the multiple tuft measurements were averaged over a two times larger distance (figure 5c; 12 mm), and thus include the combined effect of multiple tufts and boundary layer variation, while the values of the single tuft treatments (figure 5d; 6 mm) only reflect boundary layer effects on the apparent O₂ flux immediately downstream from the hair tuft.

4. Discussion

In measurements around single hyaline hair tufts, the DBL followed the contour of the smooth thallus surface except around the tufts where a thickening occurred downstream and perpendicular to the flow direction with a concomitant decrease in the DBL thickness 1–2 mm away from the hair tufts, depending on the flow regime.

In measurements over multiple tufts, the smooth apparent thickening of the DBL observed around isolated single tufts was absent. This more dynamic boundary layer landscape was probably caused by the close vicinity of neighbouring hyaline hair tufts creating a more complex flow field. Interestingly, this suggests that the effect of multiple hair tufts apparently leads to an overall increased boundary layer thickness across the thallus. Intuitively, a thin DBL would create physical conditions that could better avoid high detrimental O₂ concentrations and inorganic carbon limitations in light and O₂ limitation in darkness. So why does Fucus expend metabolic energy on the production of hyaline hairs?

In early work by Raven [12], it was suggested that (i) hyaline hairs aid in nutrient uptake by having a highly decreased diffusion resistance over the plasmalemma when compared with the thick algal thallus and (ii) hairs could protrude through the viscous sublayer into the mainstream flow with better nutrient access. However, as pointed out by Hurd [15], the thin and flexible hairs are considered unlikely to disrupt the viscous sublayer and create turbulence themselves. Here we show that, across a thallus with multiple hair tufts, the overall DBL thickness is increased, which has a functional significance similar to the observed DBL effects of epiphytes on submerged macrophytes [24,28]. A thickening of DBLs creates a mass-transfer limitation that in light can lead to high thallus surface O₂ concentrations [24,29], potentially inducing photorespiration [30] and limiting the inorganic carbon supply [31] to the thallus. However, such a mass-transfer limitation would also maintain higher nutrient concentrations due to surface-associated enzyme activity that can aid in the uptake of, for example, phosphorus and other nutrients [15,32].

A thicker DBL over thalli with tufts of hyaline hairs could also create a niche for epibiotic bacteria, and the presence of bacteria on algal thalli is well known [33–35]. In light of the recently developed ‘holobiont’ concept [35,36], a physical structure facilitating an altered chemical microenvironment, such as demonstrated here for tufts of hyaline hairs, could provide a competitive advantage, for example by providing algal-associated bacteria with metabolic compounds or by enhancing extracellular hydrolytic activities on the thallus surface due to impeded mass transfer and thus removal of dissolved exoenzymes or hydrolysis products. Studies of the role of bacteria in the algal lifecycle and metabolism have shown that a strong host specificity of epiphytic bacterial communities exists, possibly shaped by the algal metabolites as the primary selective force [37]. Previous studies have, for example, demonstrated the presence of N₂-fixing cyanobacteria as part of the algal microbiome, and it has also been shown that native bacteria are required for normal morphological development in some algae [38]. However, the actual distribution and ecological niches of such macroalgae-associated microbes are not well studied. Spilling et al. [17] found more pronounced O₂ dynamics, reaching anoxia during darkness, in the cryptostomata cavities of F. vesiculosus, wherein the hyaline hairs are anchored. Cryptostomata could thus represent potential niches for aerobic and anaerobic bacterial degradation of organic substrates or O₂-sensitive N₂ fixation that warrant further exploration.

In some transects measured on light-exposed Fucus thalli, we observed areas of enhanced O₂ concentration detached from the boundary layer (figure 2). In a previous study, it was shown that nutrient uptake rates could increase 10-fold when the boundary layer was periodically stripped by passing waves [39]. However, in our case the flow upstream from the tuft was laminar and no waves or DBL stripping occurred. The observed phenomenon of enhanced O₂ above the DBL could be explained by a combination of factors. As flow is obstructed by a physical object, a differential pressure field is created where a local drop in pressure is created around the hyaline hairs due to the locally smaller cross-section of unobstructed flow. Such a pressure gradient could create a local advective upwelling around the area of low pressure, thus affecting the O₂ transport. This phenomenon is well described in, for example, sediment transport [40] and plumes of O₂ release have also been observed before in coral-reef-associated algae Chaetomorpha sp. using planar optodes [41]. In addition, so-called vortex shedding (von Kármán vortex sheets) could also be a factor influencing the observed O₂ release. Shedding of vortices can occur at certain Reynolds numbers at the transition between laminar and turbulent flow when the pressure increases in the direction of the flow, that is, in the presence of a so-called adverse pressure gradient [42]. In our study, the flow upstream from the hyaline hairs was laminar but a transition to turbulent flow can occur, even at low Reynolds numbers, when a certain surface roughness is present and vortex shedding can initiate at Reynolds numbers of approximately 50 [43]. Using characteristic scales from this study (hyaline hair tuft diameter = 2 mm; free-stream velocity = 1.65 or 4.88 cm s⁻¹; fluid density = 1 kg l⁻¹ and a dynamic fluid viscosity of 1.08 × 10⁻³ Pa s [44]), we calculated Reynolds numbers of approximately 30–90. Von Kármán vortices have previously been connected to flow patterns on the lee side of plant parts [45] and based on the calculated Reynolds numbers the theoretical basis for the generation of vortex shedding [43] due to tufts of hyaline hairs is present in our experimental set-up. However, it is important to note that if a combination of diffusive and advective mass transfer exists downstream from the hair tufts then the calculated O₂ flux from Fick's first law in this area will be subject to large uncertainties and will underestimate the combined diffusive and advective exchange of O₂ between the thallus and the overlying water. These values should therefore be interpreted with caution and we note that, in the case of such a mixed flow regime, an apparent increase in boundary layer thickness is not necessarily linked to reductions in flux. This underscores our findings that the O₂ dynamics in the presence of hyaline hairs on the Fucus thallus are more complex than hitherto assumed.

We speculate that a combination of pressure-gradient-mediated upwelling of O₂ and vortex shedding (figure 6) could explain the observed phenomena in figure 2, and the local mass transfer related to the presence of hyaline hair tufts on fucoid macroalgae may thus be more complex than previously thought. A more detailed investigation of such mixed diffusive and advective mass-transfer phenomena was, however, beyond the scope of this study and requires a more detailed characterization of the hydrodynamic regimes over thalli with and without tufts of hyaline hairs.

Figure 6.

Conceptual drawing showing possible scenarios for the observed upwelling of O₂ downstream of the hyaline hairs. Flow velocity (straight black lines) decreases from the free-stream velocity towards the thallus surface through the DBL. The hyaline hair tuft protruding from cryptostomata alters the local boundary layer thickness and creates a differential pressure field (shown in gradient blue and red colours) due to the smaller cross-section of unobstructed flow. Flow acceleration over tufts results in local low pressure driving local advective upwelling. In addition, an adverse pressure gradient downstream from the hair tuft potentially results in vortex shedding.

In conclusion, our study of the chemical boundary layer landscape over the thallus of F. vesiculosus revealed a strong local boundary layer heterogeneity over and around tufts of hyaline hairs anchored in cryptostomata. Single tufts showed an apparent thickening of the DBL downstream and horizontally relative to the thinner DBL over the smooth thallus surface, while areas with multiple tufts exhibited an apparent overall thickening of the boundary layer that may affect gas and nutrient exchanges between the alga and seawater. Furthermore, we also observed complex solute exchange phenomena that could be driven by local pressure gradients and/or vortex shedding over the hyaline hair tufts. Altogether, this study demonstrates that interactions between flow and distinct macroalgal surface structures give rise to local heterogeneity in the chemical landscape and solute exchange. This may allow for microenvironmental niches on the thallus that can harbour epiphytic microbes with a diversity of aerobic and anaerobic metabolism. Further microscale studies of such niches in combination with, for example, microscopy and molecular detection of microbes in relation to hyaline hairs and cryptostomata thus seem an important next step to reveal further insights into the presence and role of the microbiome of fucoid algae.

Authors' contributions

R.D.N. and M.K. designed the research; R.D.N. and M.L. performed the research; M.L., R.D.N. and M.K. analysed data; M.L. wrote the paper with editorial assistance from R.D.N. and M.K. All authors gave final approval for publication.

Competing interests

The authors declare no conflict of interest and no competing financial interest.

Funding

This study was supported by grants from the Danish Council for Independent Research, Natural Sciences (M.K.), and a PhD stipend from the Department of Biology, University of Copenhagen (M.L.).