Flow and epiphyte growth effects on the thermal, optical and chemical microenvironment in the leaf phyllosphere of seagrass (Zostera marina)

Abstract

Intensified coastal eutrophication can result in an overgrowth of seagrass leaves by epiphytes, which is a major threat to seagrass habitats worldwide, but little is known about how epiphytic biofilms affect the seagrass phyllosphere. The physico-chemical microenvironment of Zostera marina L. leaves with and without epiphytes was mapped with electrochemical, thermocouple and scalar irradiance microsensors as a function of four irradiance conditions (dark, low, saturating and high light) and two water flow velocities (approx. 0.5 and 5 cm s⁻¹), which resemble field conditions. The presence of epiphytes led to the build up of a diffusive boundary layer and a thermal boundary layer which impeded O₂ and heat transfer between the leaf surface and the surrounding water, resulting in a maximum increase of 0.8°C relative to leaves with no epiphytes. Epiphytes also reduced the quantity and quality of light reaching the leaf, decreasing plant photosynthesis. In darkness, epiphyte respiration exacerbated hypoxic conditions, which can lead to anoxia and the production of potential phytotoxic nitric oxide in the seagrass phyllosphere. Epiphytic biofilm affects the local phyllosphere physico-chemistry both because of its metabolic activity (i.e. photosynthesis/respiration) and its physical properties (i.e. thickness, roughness, density and back-scattering properties). Leaf tissue warming can lead to thermal stress in seagrasses living close to their thermal stress threshold, and thus potentially aggravate negative effects of global warming.

1. Introduction

Seagrasses are marine flowering plants that form extensive meadows in coastal waters at all latitudes except in the Antarctic region [1]. Seagrass meadows are ranked among the most productive and diverse coastal benthic ecosystems [2]. They play paramount ecological roles as habitat for fish and invertebrates [3] and hot spots for carbon sequestration [4]. They also provide various other ecosystem services by acting as recreational habitats and important nursery grounds for commercial species in coastal zones [5].

Globally, seagrass habitats are declining mainly due to anthropogenic disturbances [5]. While stressors arising from global warming [6], salinity changes [7] or the eelgrass wasting disease [8] are known to affect seagrass meadows, one of the greatest environmental impacts on seagrass are related to decreasing coastal water quality [9]. In particular, increasing nutrient loads cause eutrophication of seagrass habitats, which can impair seagrass growth and metabolism (e.g. [10]). Such eutrophication has intensified in vegetated coastal habitats since the 1950s [11]. Eutrophication promotes macroalgal blooms, which can competitively exclude seagrasses [12], and enhances the biomass and productivity of epiphytes colonizing seagrass leaves, in particular filamentous microalgae [13–16]. Depending on the environmental conditions (e.g. the presence of grazers, nutrient concentrations), the epiphytic cover can spread over 10–70% of the whole plant-associated dry mass [17]. Seagrass epiphytes modify the seagrass phyllosphere and can restrict access to light and carbon [18–20] particularly in light-limited environments [21], resulting in reduced seagrass productivity [17,18].

The seagrass phyllosphere is defined as the dynamic microhabitat forming at the tissue surface of above-ground parts of seagrasses. It is particularly important for the functioning and the fitness of the whole seagrass plant [19]. Indeed, it represents the main interface for gas and nutrient exchange [20]; in particular, inorganic carbon species for photosynthesis in light [18] and night-time uptake of O₂ from the surrounding water that is then transported to the below-ground tissues and the rhizosphere by aerenchymal tissues [22–24].

Within the phyllosphere, shear stress resulting from the water flowing above the leaf surface or epiphytic biofilm leads to the build-up of a velocity boundary layer reaching zero at the tissue surface [25]. Close to the leaf/epiphytic biofilm, turbulent transport is dampened and the flow turns laminar in the thin innermost water layer (less than 0.1–1 mm), where solute exchange between the seagrass leaf surface (with or without biofilm) and the surrounding water is governed by molecular diffusion [25]. This has the consequence that steep concentration gradients develop within this so-called diffusive boundary layer (DBL) [26,27] that are affected by flow, water-column solute contents, leaf/epiphytes metabolism and topography. At low flow, the mass transfer limitation in the DBL [28] can impede the exchange of solutes with the surrounding water column [27]. In seagrass meadows, this can potentially lead to insufficient O₂ supply during water column hypoxia at night-time (e.g. [29,30]), where the O₂ transfer cannot sustain the O₂ demand of (i) aerobic metabolism in below-ground plant tissue parts and (ii) the radial oxygen loss from root tips and the rhizome. The latter is essential for avoiding H₂S intrusion into the plant from the surrounding reduced sediment [19,31].

Analogous to molecular diffusion, heat transfer from the leaf surface to the bulk water is also impeded by the decreasing flow, leading to a thermal boundary layer (TBL) surrounding illuminated pigmented tissues and biofilms that dissipate excess absorbed light as heat [32]. The temperature increase at the tissue surface is counterbalanced by heat transfer to the surrounding water. Consequently, a temperature gradient is established in the TBL. There is ample evidence for TBLs in corals and photosynthetic biofilms that can heat up to 1°C above the surrounding seawater [32–34]. The presence of a TBL on seagrass leaves remains to be demonstrated.

Epiphytes are known to assimilate biologically important molecules (e.g. O₂) before they reach the seagrass surface [27] and they can alter nutrient and gas dynamics [18,21], as well as the light microclimate at the surface of seagrass leaves [19,35]. The epiphytic community composition exhibits seasonal variation. In Danish waters, epiphytic diatoms and brown algae are always present, green algae are mainly found during springtime, and cyanobacteria and red algae are mainly present in autumn [15]. Epiphytes shade the leaves, reducing the intensity of light received by photosynthetic surfaces [35,36]. Moreover, the photopigments of some of the species composing the epiphytic biofilm strongly absorb blue and red light, leaving the plant in a light environment dominated by green light, which is less effectively absorbed by chlorophyll a and b, the dominant seagrass photopigments [19]. The presence of an epiphytic biofilm can favour nutrient mobilization via N₂ fixation and/or via facilitating uptake of dissolved organic nitrogen [37] depending on the species present in the biofilm and the metabolic pathways involved. Nevertheless, unfavourable conditions could strongly hamper seagrass fitness as a result of a reduction in received light, a decrease in O₂ availability in darkness and/or production of reduced phytotoxic compounds, such as nitric oxide (NO) production via denitrification in anoxic parts of the epiphytic biofilm. However, the diel dynamics in the phyllosphere microenvironment are underexplored, and also depend on physical factors such as flow velocity, temperature and incident irradiance [38,39] that can mitigate or worsen the effects of epiphytes on seagrass health.

To better understand the complex physico-chemical conditions and dynamics taking place in the seagrass phyllosphere, we used electrochemical, thermocouple and scalar irradiance microsensors to investigate the O₂ microgradients, temperature variations and light microclimate at the surface of Zostera marina L. leaves with and without the presence of microalgal epiphytic biofilms. Oxygen microprofiling allowed calculations of O₂ fluxes across the surface of the leaves and the epiphytic biofilm. We also present first determinations of a TBL around seagrass leaves, and characterize in detail the light microclimate in epiphytic biofilms and at the seagrass leaf surface in different conditions of flow and irradiance. Based on previous studies showing a negative impact of epiphytes on seagrass leaves, we hypothesized that their presence would deteriorate the conditions for optimal productivity of Z. marina leaves, particularly because of inducing shading, impeding optimal photosynthesis and leading to phyllosphere anoxia during darkness.

2. Results

2.1. Oxygen microenvironment

The thickness of the DBL was similar above bare leaves and epiphytic biofilms with a mean of 0.29 ± 0.04 mm. The DBL thickness was only affected by the flow velocity (p_1,39.639 < 0.001, three-way ANOVA). The DBL was 1.9 ± 0.1 and 1.2 ± 0.05 times thicker under slow flow (0.5 cm s⁻¹) than under fast flow (5 cm s⁻¹) for bare leaves and leaves with epiphytes, respectively. The presence of epiphytes strongly enlarged the thickness of the hyperoxic/hypoxic area surrounding the leaves; proportionally to the thickness of the biofilm (figure 1). Within these dense biofilms, mass transport is likely determined solely by diffusion which is why the total diffusion distance (TDD; see [18]) was determined as the sum of the combined thickness of the epiphytic biofilm and the DBL above. The TDD between the seagrass leaf and the surrounding water reached a thickness sixfold greater than the TDD measured above bare leaves in the case of the thickest biofilm (table 1). As for the DBL, the TDD was also thinner (mean reduction of approx. 46 and 13% for bare and epiphyte-covered leaves, respectively) in fast than in slow flow conditions.

Figure 1.

Oxygen microenvironment in the seagrass phyllosphere. Depth profiles of O₂ concentration measured from the leaf surface to the bulk seawater at two different flow velocities, fast flow (5 cm s⁻¹; left panels) and slow flow (0.5 cm s⁻¹; right panels), for bare seagrass leaves and leaves with epiphytes. Profiles were measured in darkness (0 µmol photons m⁻² s⁻¹; blue symbols), at low (110 µmol photons m⁻² s⁻¹; orange symbols), saturating (500 µmol photons m⁻² s⁻¹; green symbols) and high (1050 µmol photons m⁻² s⁻¹; red symbols) photon irradiance levels of PAR (400–700 nm). The position of the seagrass leaf surface is illustrated in pale green (Y = 0), while the upper limits of the epiphytic biofilm and the effective DBL (i.e. the total diffusion distance, TDD) are represented by plain and dotted lines, respectively. Note that the y- and x-axis are different for bare and epiphyte-covered leaves. Data originate from three biological replicates.

Table 1.

Epiphytic biofilm thickness and total diffusion distance (TDD) (in millimetres) estimated from the O₂ concentration gradients measured at the leaf surface, above bare leaves and leaves with epiphytes. The mean ± s.e. values reported for fast (5 cm s⁻¹) and slow (0.5 cm s⁻¹) flow were calculated from the microsensor profiles (n = 4) measured at different incident photon irradiance levels (dark: 0; low: 110; saturating: 500 and high 1050 µmol photons m⁻² s⁻¹; PAR 400–700 nm).

		epiphytic biofilm (mm)	TDD thickness (mm)
			fast flow	slow flow
replicate 1	bare leaf		0.19 ± 0.01	0.38 ± 0.02
replicate 2	bare leaf		0.25 ± 0.02	0.40 ± 0.01
replicate 3	bare leaf		0.24 ± 0.01	0.50 ± 0.02
replicate 1	leaf with epiphytes	1.1	1.15 ± 0.05	1.50 ± 0.04
replicate 2	leaf with epiphytes	2.6	2.72 ± 0.05	3.11 ± 0.08
replicate 3	leaf with epiphytes	1.3	1.50 ± 0.00	1.53 ± 0.03

The O₂ concentrations measured at the leaf surface (electronic supplementary material, table S1) increased with the light irradiance regardless of the leaf type, i.e. with or without epiphytes (p_3,200.292 < 0.001, three-way ANOVA). In the light, surface O₂ concentrations were much higher than the bulk O₂ concentration, except in low light on leaves with epiphytes, i.e. 2.0–2.2-fold increase for bare leaves and 2.5–2.7-fold increase for leaves with epiphytes at 500 and 1050 µmol photon m⁻² s⁻¹ (p_3,200.292 < 0.001, three-way ANOVA, interaction leaf: light). Maximum surface O₂ concentrations were reached already at 500 µmol photon m⁻² s⁻¹ for bare leaves, when compared with 1050 µmol photon m⁻² s⁻¹ for leaves with epiphytes (figure 1). Considering the whole phyllosphere, the highest O₂ concentrations were observed in the middle of the thick epiphytic biofilm under slow flow (figure 1), reaching 620, 975 and 657 µmol O₂ l⁻¹ for replicates 1, 2 and 3, respectively. The respective fast flow values were 565, 880 and 628 µmol O₂ l⁻¹.

During darkness, the presence of epiphytes strongly exacerbated hypoxic conditions in the phyllosphere (figure 1), reducing the averaged leaf surface O₂ availability by approximately 79–97% under fast and slow flow, respectively. Remarkably, very low O₂ levels were observed at the leaf surface in the case of plant replicate 2 (0.11 µmol O₂ l⁻¹, slow flow, darkness), where almost the whole epiphytic layer was anoxic, except for the uppermost part of the epiphytic biofilm and at the leaf surface in fast flow. The latter was probably due to O₂ intrusion from the region below the leaf and/or lower epiphyte density at the leaf surface.

2.2. Light microenvironment

The presence of epiphytes changed the light microclimate in the phyllosphere in terms of both intensity (figure 2; p_1,53.210 < 0.001, linear mixed effect model) and spectral quality (figure 3; electronic supplementary material, figure S4) of photon scalar irradiance (PAR, 400–700 nm) at the seagrass leaf surface.

Figure 2.

Light availability in the seagrass phyllosphere. Depth profiles of photon scalar irradiance (PAR; 400–700 nm) measured from the bulk seawater to the leaf surface at low (110 µmol photons m⁻² s⁻¹; left panel), saturating (500 µmol photons m⁻² s⁻¹; middle panel) and high (1050 µmol photons m⁻² s⁻¹; right panel) photon irradiance. Note that the x-axes scales are different. Profiles were measured on bare leaves (blue lines and symbols) and on leaves with epiphytes, under fast flow (5 cm s⁻¹; orange lines and symbols) and slow flow (0.5 cm s⁻¹; green lines and symbols) conditions. The position of the seagrass leaf (y = 0) is illustrated in pale green, while the upper limit of the epiphytic biofilm is represented by a plain line. Data originate from three biological replicates.

Figure 3.

Light quality in the seagrass phyllosphere. (a) Spectral scalar irradiance (in % of the incident irradiance) measured above a Z. marina leaf with epiphytes (replicate 3). Scalar irradiances spectra (400–750 nm) are presented from the leaf surface (blue line, 0 mm from leaf surface) to the outer surface of the epiphytic biofilm (pale pink line, 1.2 mm from the leaf surface) in 0.2 mm vertical depth increments. (b) Scalar irradiance spectral attenuation coefficient, K₀, measured above a Z. marina leaf with epiphytes (replicate 3). The spectral attenuation coefficient was measured for five different layers, i.e. 0–0.2 mm, 0.2–0.6 mm, 0.6–1.0 mm, 1.0–1.2 mm and 1.2–1.7 mm distance from the leaf surface. Results are presented for three incident irradiance levels (110, 500 and 1050 µmol photons m⁻² s⁻¹), in fast flow (5 cm s⁻¹; left panel) and slow flow (0.5 cm s⁻¹; right panel), for replicate 3 only. (c,d) Control measurement on a leaf without epiphytes (bare leaves). Measurements on the two other biological replicates are shown in electronic supplementary material, figure S2.

In the presence of epiphytes, the photon scalar irradiance of PAR at the leaf surface represented only on average 43% of the incident PAR in fast flow, while it increased to 50% in slow flow (figure 2 and table 2). All replicates included, the degree of light attenuation was similar for the different incident light levels and a significant flow effect due to biofilm compression at high flow was detected (p_1,5.516 = 0.034, linear mixed effect model). However, biofilms on replicates 2 and 3 attenuated light slightly more than replicate 1, even though biofilms 1 and 3 had similar thickness. This could be due to density or community composition differences. We also observed that the light attenuation throughout the thickest biofilm (replicate 2) was more fluctuating than in the thinner biofilms, which could be related to a less compact structure of biofilm 2 (figure 2).

Table 2.

Light attenuation (in %) and photon scalar irradiance of PAR (400–700 nm) at the seagrass leaf surface (in µmol photons m⁻² s⁻¹), as calculated from scalar irradiance microprofiles measured above bare seagrass leaves and leaves with epiphytes under different incident photon irradiance levels (low: 110; saturating: 500 and high 1050 µmol photons m⁻² s⁻¹). As flow velocity can affect the shape of the microalgal epiphytic biofilm, the light attenuation was determined under fast and slow flow, independently. Mean ± s.e. (n = 3 biological replicates).

leaf	flow	incident irradiance (µmol photons m−2 s−1)	N	attenuation (%)	s.e.	photon scalar irradiance at the leaf surface (µmol photons m−2 s−1)	s.e.
bare leaf		110	3	12	11	89	17
bare leaf		500	3	0	3	545	43
bare leaf		1050	3	−1	1	1051	78
leaf with epiphytes	fast	110	3	53	6	51	9
leaf with epiphytes	fast	500	3	56	8	269	60
leaf with epiphytes	fast	1050	3	52	10	537	122
leaf with epiphytes	slow	110	3	44	11	64	18
leaf with epiphytes	slow	500	3	56	5	265	21
leaf with epiphytes	slow	1050	3	50	8	559	101

The presence of epiphytes also affected the spectral characteristics of light reaching the seagrass leaf (figure 3a,b). A gradual decrease in the spectral scalar irradiance was detected particularly across the range of photosynthetic active wavelengths (400–700 nm). Strongest light attenuation was observed at 673 nm, corresponding to the absorption maxima of chlorophyll (Chl) a (figure 3a). At this wavelength, the maximum light attenuation coefficients ranged from 1 ± 0.4 (fast flow, 110 µmol photons m⁻² s⁻¹) to 1.7 ± 0.4 mm⁻¹ (slow flow, 1050 µmol photons m⁻² s⁻¹). The strongest light attenuations occurred between 0.2 and 0.6 mm depth into the epiphytic biofilm under fast flow and between 0.6 and 1 mm depth into the epiphytic biofilm under slow flow (figure 3b). Moreover, the scalar irradiance in the first 0.2–0.4 mm of the epiphytic biofilm (0.8–1.2 mm from the leaf surface) increased up to 130% relative to the incident irradiance under slow flow, especially for wavelengths between 550 and 650 nm, (figure 3a,b). In comparison, only minor spectral light changes were observed on bare seagrass leaves (figure 3c,d).

2.3. Net O₂ gas exchange

For bare leaves, the calculated net O₂ fluxes at the leaf surface are representative of respiration in the dark and net photosynthetic rates in the light (figure 4). Respiration rates amounted to 411 ± 51 and 444 ± 37 nmol O₂ m⁻² s⁻¹, for slow and fast flow, respectively. In the light, the highest net photosynthetic rates were measured at a saturating incident photon irradiance of 500 µmol photons m⁻² s⁻¹, reaching 1145 ± 204 and 1484 ± 372 nmol O₂ m⁻² s⁻¹ in fast and slow flow, respectively (electronic supplementary material, table S2).

Figure 4.

Net O₂ fluxes in the seagrass phyllosphere. Net O₂ fluxes (nmol m⁻² s⁻¹), from the leaf/epiphytes community, i.e. measured at the epiphytic biofilm/seawater interface (light grey bars), the seagrass leaf surface with epiphytes (dark grey bars) and from bare seagrass leaves (black bars). Fluxes were calculated according to Fick's first law of diffusion and were measured in fast flow (5 cm s⁻¹; left panel) and slow flow (0.5 cm s⁻¹; right panel) conditions in darkness and under low, saturating and high photon irradiances (i.e. 0, 110, 500 and 1050 µmol photons m⁻² s⁻¹, respectively). Negative values denote an O₂ influx and positive values an O₂ efflux across the respective surface/interface. Fluxes are presented as means ± s.e. (n = 3, biological replicates).

The presence of an epiphytic biofilm drastically reduced the net O₂ fluxes at the seagrass leaf surface (figure 4; p_1,15.123 = 0.001, three-way ANOVA). Negative values were reached in almost all the investigated conditions, except for the leaf with the thinnest epiphytic biofilm (replicate 1). While the O₂ fluxes increased with the light intensity on bare leaves, peaking at 500 µmol photons m⁻² s⁻¹ (figure 4), they remained similar for all the light conditions in the presence of epiphytes (p_2,5.149 = 0.020, three-way ANOVA, leaf : light interaction). A negative net O₂ flux indicates a net influx of O₂ at the interface between the seagrass leaf surface and the epiphytic biofilm, even under illuminated conditions. Conversely, in the dark the influx of O₂ to the seagrass leaf was drastically reduced by approximately 90% in the presence of epiphytes.

The combined seagrass/epiphyte community flux (i.e. a result of net O₂ production by the epiphytic community and the leaf), followed a similar pattern as bare seagrass leaves with an O₂ influx in darkness and an O₂ efflux in light (figure 4); although reduced compared to bare leaves. Only light affected the community fluxes (p_3,11.125 < 0.001, two-way ANOVA), with rates increasing from −0.37 ± 0.11 nmol O₂ m⁻² s⁻¹ in the dark to 0.74 ± 0.23 nmol O₂ m⁻² s⁻¹ at 500 µmol photons m⁻² s⁻¹, regardless of the flow condition.

2.4. Temperature microenvironment

Temperature microprofiles showed that absorbed radiation can drive an increase in the surface temperature of seagrass leaves, leading to convective heat transfer to the surrounding seawater and thus to the establishment of a TBL. The temperature increase at the leaf surface relative to the bulk seawater temperature was affected by flow velocity (p_1,37.194 < 0.001), incident irradiance (p_3,71.779 < 0.001), the presence of epiphytes (p_1,19.506 < 0.001) and interactive effects between these factors (figure 5; p_3,3.483 = 0.026, three-way ANOVA, flow : light : leaf interaction).

Figure 5.

Temperature microenvironment in the seagrass phyllosphere. (a) Depth profiles of temperature shown as the temperature difference from the bulk seawater temperature. Temperature profiles from the leaf surface to the bulk seawater were measured for two different flow velocities, fast flow (5 cm s⁻¹; top panel) and slow flow (0.5 cm s⁻¹; bottom panel). Profiles were measured on bare leaves (left panel) and on leaves with epiphytes (right panel), in dark (blue symbols), low (orange symbols), saturating (green symbols) and high (red symbols) photon irradiance (i.e. 0, 110, 500 and 1050 µmol photons m⁻² s⁻¹, respectively). The position of the seagrass leaf surface (y = 0) is illustrated in pale green, while the upper limit of the epiphytic biofilm is represented by a plain black line. Results are shown for replicate 2, profiles for the other two biological replicates are reported in electronic supplementary material, figure S3. (b) Leaf surface warming, ΔT (°C), as a function of photon scalar irradiance (PAR; 400–700 nm) in the phyllosphere. Fast flow conditions (5 cm s⁻¹) are shown on the left panel, while the slow flow conditions (0.5 cm s⁻¹) are on the right panel. Yellow symbols represent values for leaves with epiphytes, while blue symbols are for bare leaves. Dotted lines are linear regression lines (r² = 0.37 and 0.75 for bare leaves; r² = 0.57 and 0.88 for leaves with epiphytes, for fast and slow flow conditions, respectively). Data originate from three biological replicates.

Negligible small temperature changes were observed under dark conditions (table 3), which then increased significantly with the incident irradiance. The leaf surface warming was lower in fast than slow flow (regardless of the presence of epiphytes), peaking at 0.19 ± 0.05°C in the presence of epiphytes at 1050 µmol photon m⁻² s⁻¹. In slow flow, the temperature increase at the bare leaf surface reached 0.24 ± 0.02°C under an incident irradiance of 1050 µmol photons m⁻² s⁻¹. These temperature differences remained lower than the increase measured on leaves with epiphytes and was dissipated by heat transfer over approximately 2 mm thick TBL above the bare leaf (figure 5a).

Table 3.

Temperature difference between the bulk seawater and the seagrass leaf surface (ΔT in °C), measured above bare leaves and leaves with epiphytes, as a function of flow velocity (fast: 5 cm s⁻¹ and slow: 0.5 cm s⁻¹) and incident photon irradiance (dark: 0; low: 110; saturating: 500 and high: 1050 µmol photons m⁻² s⁻¹). Values are means ± s.e. (n = 3 biological replicates).

leaf	flow	incident irradiance (µmol photons m−2 s−1)	N	leaf warming (Δ°C)	s.e.
bare leaf	fast	0	3	0.05	0.01
bare leaf	fast	110	3	0.04	0.01
bare leaf	fast	500	3	0.11	0.03
bare leaf	fast	1050	3	0.14	0.08
bare leaf	slow	0	3	0.01	0.03
bare leaf	slow	110	3	0.08	0.03
bare leaf	slow	500	3	0.19	0.01
bare leaf	slow	1050	3	0.24	0.02
leaf with epiphytes	fast	0	3	−0.03	0.02
leaf with epiphytes	fast	110	3	0.09	0.03
leaf with epiphytes	fast	500	3	0.16	0.03
leaf with epiphytes	fast	1050	3	0.23	0.03
leaf with epiphytes	slow	0	3	0.02	0.01
leaf with epiphytes	slow	110	3	0.11	0.04
leaf with epiphytes	slow	500	3	0.35	0.07
leaf with epiphytes	slow	1050	3	0.56	0.08

The presence of an epiphytic biofilm (thickening the TDD), increased the TBL thickness reaching approximately 4 mm (replicate 2, figure 5a) up to approximately 7.5 mm (replicate 3; electronic supplementary material, figure S5) in slow flow. The thicker TBL increased the leaf tissue surface warming, where the average surface temperature of leaves with epiphytes increased up to 0.56 ± 0.08°C with respect to the bulk seawater temperature (i.e. approx. 2.3-fold compared to bare leaves), under an incident irradiance of 1050 µmol photons m⁻² s⁻¹ in slow flow (electronic supplementary material, table S2). A maximum leaf tissue heating of up to +0.8°C was measured in replicate 1 (electronic supplementary material, figure S5), representing approximately 3.3-fold temperature increase compared to bare leaves.

A positive linear correlation (p < 0.05 for all the correlations) was found between the seagrass leaf surface warming and the scalar irradiance measured at the leaf surface (figure 5b). For fast and slow flow, respectively, we found R² values of linear regressions of 0.37 and 0.75 for bare leaves, and 0.57 and 0.88 for leaves with epiphytes. The correlation was stronger for leaves with epiphytes exhibiting a heating slope of 0.22°C (10³ µmol photons m⁻² s⁻¹)⁻¹ in fast flow, increasing to 0.87°C (10³ µmol photons m⁻² s⁻¹)⁻¹ in slow flow, when compared with only 0.08 and 0.17°C (10³ µmol photons m⁻² s⁻¹)⁻¹ in fast and slow flow for bare seagrass leaves (figure 5b).

2.5. Nitric oxide production

Anoxic conditions within the leaf epiphytic biofilm in darkness (replicate 2, figure 1), led to the accumulation of potentially phytotoxic NO in the epiphytic biofilm (figure 6). NO concentrations peaked at 3.9 ± 1.2 µmol NO l⁻¹ (mean ± s.d., n = 3, technical replication) at a depth of 0.4 mm above the leaf surface, while they reached 1.3 ± 0.3 µmol l⁻¹ at the leaf surface. We estimated a flux of NO into the seagrass leaf of −18 ± 5 nmol NO m⁻² s⁻¹. No production of NO was observed in seagrass replicate 1 and 3, correlating with increased O₂ availability in their epiphytic biofilm (figure 1).

Figure 6.

Nitric oxide (NO) production within the epiphytic biofilm. Depth profiles of NO concentration measured in darkness, from the seagrass leaf surface to the bulk seawater on plant replicate 2 (with and without biofilm) and plant replicates 1 (green profile) and 3 (red profile) with biofilms. Orange profiles are technical replicates made at different positions in the epiphytic biofilm of plant 2, while blue profiles were technical replicates measured on the bare leaf segment from the same plant. The position of the seagrass leaf surface is illustrated in pale green (y = 0). Values measured on blue, green and red profiles are <|0.3| µmol l⁻¹, very close to the sensor's detection limit and due to electrical noise affecting the zero current of the sensor.

3. Discussion

This study investigated the physico-chemical conditions and dynamics within the phyllosphere of Zostera marina L. in the presence or absence of leaf epiphytic biofilms, under different flow velocities and light intensities. As already demonstrated by other studies [18,19,36,40], epiphytic biofilms increased the thickness of the diffusion distance compared to bare leaves, which together with metabolic activity in the epiphytic biofilms, led to severe O₂ limitation in darkness, especially under slow flow. Epiphytes also attenuated incident light, especially at high flow, leading to substantial shading of the seagrass leaf surface that negatively affected the photosynthetic performance of the seagrass host. For the first time, phytotoxic NO formation associated with anoxic stress due to epiphytes was evidenced in darkness. As another original outcome, the presence of a TBL was demonstrated in the seagrass phyllosphere, which resulted in leaf surface warming with increasing irradiance, especially in the presence of epiphytes and low flow velocities. In summary, epiphytes strongly change the physico-chemical properties of the seagrass phyllosphere, which can induce microenvironmental stress. Depending on the total leaf area covered by epiphytes, this microenvironmental stress may ultimately affect the whole functioning of seagrass meadows in coastal areas subject to eutrophication.

3.1. Epiphytic biofilms shape the chemical conditions in the phyllosphere, especially under slow flow

The seagrass phyllosphere microenvironment is shaped by the dynamic interaction of the water column physico-chemical conditions, the seagrass metabolism and the activity of autotrophic and heterotrophic epiphytes inhabiting the leaf surface. On bare leaves, the DBL thickness decreased with the flow velocity from approximately 0.4 mm in slow flow to approximately 0.2 mm in fast flow, which fall within the range of DBL thickness measured for other aquatic macrophytes (e.g. 0.4 mm at a flow rate of 0.3 cm s⁻¹ and 0.2 mm at a flow rate of 3 cm s⁻¹; [23,41]). An epiphytic community colonizing a seagrass leaf affects the local microenvironment, both because of its metabolic activity and its physical properties. The porous structure of epiphytic biofilm generates a transport of solutes within them, mostly ruled by molecular diffusion [40]. Consequently, the ‘effective' diffusion distance between the mixed seawater and the leaf tissue surface beneath the epiphytic microalgal biofilm (TDD, total diffusion distance; [18]) is longer. In slow flow, the TDD was 3.4 to 6.9 times thicker than the DBL on bare leaves. Thus, epiphytic biofilms both increase diffusion distances for solute exchange between leaf and seawater, and affect the physiology of the seagrass leaf due to the build up or the depletion of solutes diffusing from and to the bulk seawater [40], which can further limit seagrass metabolism [27,42]. For instance, mass transfer limitation of inorganic carbon, soluble phosphorous and nitrogen species, can reduce seaweed performance [38]. This mass transfer limitation due to a thicker DBL might have caused the slight reduction in net photosynthetic rates at the surface of bare leaves in slow flow, when compared with fast flow conditions in our study. In addition, a stronger basification of the leaf microenvironment driven by the pH increase due to photosynthesis could have reduced the CO₂ and HCO₃⁻ availability at the leaf surface, inducing carbon limitation in slow flow conditions; although this has mainly been shown for epiphyte-covered leaves [18].

In the light, the phyllosphere around bare leaves exhibited higher O₂ levels than in the bulk seawater, reaching maximal O₂ concentrations of approximately 550 µmol l⁻¹ (for replicate 3 at 500 µmol photon m⁻² s⁻¹). The combined effect of increasing TDD and photosynthetic activity in the biofilm caused a twofold increase in O₂ concentration in the epiphytic biofilm microenvironment (approx. 1000 µmol l⁻¹; replicate 2) under a similar photon irradiance and regardless of the flow velocity. Such high O₂ build-up around leaves driven by the presence of epiphytes may stimulate photorespiration in the seagrass leaf by enhancing the oxygenase activity of Rubisco [18,23]. This competing activity of Rubisco [43] decreases photosynthetic carbon fixation [18,44]. It may also induce internal oxidative stress [45], accumulation of reactive oxygen species [46] and overall reduce the photosynthetic performance of the seagrass.

In darkness, the phyllosphere turned hypoxic in the presence of epiphytes because of a thicker TDD and epiphyte community respiration. Anoxia was even reached within one of the epiphytic biofilms (replicate 2). Hypoxic to anoxic conditions induced by the presence of epiphytic communities were previously observed for other macrophytes [40,47], affecting the O₂ supply to the leaves and the respiration efficiency of the whole plant [19,23,24]. At the leaf surface, O₂-depletion may create a favourable environment for fermentation processes, hypothetically leading to ethanol production, which is considered a phytotoxin [48]. Anoxic biofilm conditions can also promote the production of NO via NO synthases [49]. NO has diverse physiological functions but is also a phytotoxin that interacts with the plant signalling network [50], provoking both beneficial and harmful effects [51]. The present study reports the first direct measurements of NO production within the epiphytic biofilm covering seagrass leaves, with NO concentration reaching approximately 4 µmol l⁻¹ in the anoxic part of the microalgal epiphytic biofilm. NO has been shown to have cytotoxic properties at concentrations down to 0.5 µmol l⁻¹ [52], initiating chain reactions that can cause cell injury and death [53,54]. The concomitant production of these two metabolic end-products (i.e. NO and ethanol) under hypoxic to especially anoxic leaf microenvironment conditions can thus result in a critical cocktail that potentially could be lethal for the leaves and ultimately the whole plant when epiphytic biofilms cover large portions of the leaves (e.g. [55]). This apparent relationship between anoxic stress due to epiphytes and formation of NO in the biofilm deserves more attention in future studies.

3.2. Epiphytes modify the phyllosphere light microclimate and affect the leaf productivity

The presence of an epiphytic biofilm dramatically attenuated the photon scalar irradiance of PAR (400–700 nm) reaching the surface of the epiphyte-covered leaves (by more than 50%), in line with previous findings [19,35,56]. In fast flow conditions, the compression of the biofilm further decreased the light intensity reaching the leaf, likely reducing the photosynthetic efficiency of epiphyte-covered leaves when compared with bare seagrass leaves. Meanwhile, a locally increased scalar irradiance in the upper layers of the epiphytic biofilm, due to photon trapping, redistribution and scattering within the biofilm was observed; analogously to what has been observed in coral tissue [33,57], microbial mats [58] and sediments [34,59].

Besides the overall reduction in PAR, the epiphytic biofilm also changed the spectral quality of light reaching the leaf surface. As already shown by Brodersen et al. [19], epiphytic biofilms can markedly reduce the light availability around 440 and 670 nm, corresponding to the absorption maxima of chlorophyll a. Depending on the biofilm community composition and epiphyte pigment composition [15], the absorption spectra could also be seasonally influenced, with more absorption in the green part of the light spectrum, as cyanobacteria and red algae become dominant in the biofilm. By influencing the light quality, the epiphytes may even exacerbate the light limitation, likely leaving the plant for longer time periods near minimal light requirements for growth [60,61]. Incident irradiance in seagrass meadows can vary substantially, e.g. due to wind stress and eutrophication events over daily and seasonal cycles [62], and it is likely that epiphyte overgrowth likewise limits seagrass photosynthetic performance when they cover large percentages of photosynthetic surfaces.

In the present study, O₂ fluxes measured at the leaf surface showed that the presence of thick epiphytic biofilms on seagrass leaves strongly affected plant photosynthesis and respiration. The flux of O₂ at the surface of leaves with epiphytes was negative, indicating a net O₂ influx rather than O₂ efflux from the seagrass leaf in light, while a high efflux of O₂ was measured on bare seagrass leaves in light. Strong light attenuation by epiphytes, O₂ build-up in the phyllosphere, and a limitation in inorganic carbon and other solutes, associated with thick TDD and active biofilms, contribute to reduced seagrass productivity and performance. Moreover, we found a pronounced reduced O₂ influx in darkness to leaves with epiphytes when compared with bare leaves. Epiphytes thus can markedly reduce internal plant aeration in darkness, which also affects the belowground tissue ability to oxidize rhizosphere compartments and prevent H₂S intrusion [19,31], owing to weakened internal O₂ concentration gradients.

3.3. The temperature boundary layer can worsen the deleterious effects of epiphytes

In both bare leaves and leaves with epiphytes, the leaf surface temperature increased relative to the bulk seawater with increasing irradiance. This phenomenon is due to the presence of a leaf TBL slowing down heat exchange with the surrounding water. Such TBLs have previously been shown to affect the radiative energy budget of corals and other surface-associated microbial communities [32–34,63], but have hitherto not been demonstrated for seagrass leaves. Driven by increased light absorption and dissipation of light energy as heat, the presence of epiphytes increased the temperature difference between the surrounding water and leaf surface, reaching a leaf surface warming of 0.56 ± 0.08°C under 1050 µmol photon m⁻² s⁻¹. The heat was dissipated in a 2.0–7.5 mm thick TBL surrounding the leaf (measured for bare leaf and leaf with the thickest epiphytic biofilm, respectively, in slow flow condition); corresponding to an approximately four times thicker TBL than TDD. Increasing flow velocity, alleviated the leaf surface warming as the TBL thickness decreases with the inverse power of the flow rate [64].

The channelling of excess light energy into heat dissipation can protect the photosynthetic apparatus under high irradiance [44]. However, for seagrasses living close to their thermal stress threshold [65], the presence of a TBL could cause a critical temperature increase at the leaf surface under low flow and high irradiance conditions, which can exacerbate thermal stress, likely causing severe protein damage [66] and oxidative stress [67]. Already sensitive to warming [6], seagrass health could thus be further deteriorated by a large epiphytic coverage impeding convective heat dissipation from the leaf surface. As the presence of epiphytic biofilm could lead to a temperature increase in the leaf microenvironment relative to the ambient water, that could potentially aggravate the negative effects of global warming in tropical waters or in case of more frequent heatwaves.

3.4. The indirect effect of eutrophication on seagrass meadows: the threat of epiphytes

Coastal zones are strongly impacted by sea surface temperature increase [68] and are increasingly exposed to eutrophication events driving severe hypoxia [69]. Indeed, nutrients loads entering marine waters usually cause a bloom of fast-developing autotrophic species such as phytoplankton and opportunistic algae, often resulting in more turbid waters (i.e. reduced light intensity) and local O₂ depletion. Epiphytic biofilms on seagrass plants are typically dominated by opportunistic species, especially by filamentous algae [70], and a heavier epiphytic load is usually observed in seagrass meadows growing in warmer and eutrophic waters [17,36,71]. Overgrowth of seagrasses with epiphytic communities can drastically alter the metabolism of seagrass leaves, likely causing energetic trade-offs and impairing the general functioning of the whole plant (as summarized in figure 7). Here, we showed that thick epiphytic biofilms (greater than 1 mm) can lead to the presence of strong leaf TBLs in light, especially at low flow velocities, and the potential formation of phytotoxic NO in the anoxic epiphytic biofilms during darkness. Such leaf surface warming could aggravate detrimental effects of heatwaves, hot calm summer periods and global warming for seagrasses, putting them at risk of surpassing critical thresholds for leaf thermal stress. We also showed that epiphyte-induced shading of seagrass leaves can increase under high flow conditions presumably due to biofilm compression, leaving epiphyte-covered seagrasses with limited relief of good water movement. Under these unfavourable conditions created on large areas by epiphytic overgrowth, the general plant growth could be slowed down [72] and mortality may increase because of the reduced internal plant aeration and increased H₂S intrusion [19,31]. Hence, the increase in epiphytic loads on seagrass leaves could superimpose stressful environmental conditions already experienced by seagrass meadows, worsening the effects of climate change on these fragile ecosystems.

Figure 7.

Seagrass phyllosphere dynamics in the presence and absence of an epiphytic biofilm. Epiphyte overgrowth (right side of the schematic diagram) on seagrass leaves (green symbols) can lead to hostile microenvironmental conditions and dynamics in the seagrass phyllosphere. In light, leaf epiphytes can lead to a strong build-up of O₂ (dots, bolder and darker dots indicate increasing O₂ concentrations) and reduced light conditions (thinner arrows) in the leaf microenvironment. In darkness, leaf epiphytes can lead to reduced internal plant aeration and production of phytotoxic NO (gradient of purple hue means increasing NO concentrations) within local anoxic leaf microhabitats; thus impeding the plants performance and likely leading to synergetic negative effects on the plants overall O₂ balance (‘hampered' tissues). Moreover, slow flow conditions can lead to thick leaf TBLs (marked with a yellow to orange colour code) and TDDs on epiphyte-covered seagrass leaves in the light, which may push densely epiphyte-covered seagrasses beyond their thermal tipping point. This can lead to high plant mortality in seagrass living close to their thermal stress thresholds and/or in coastal waters frequently exposed to water-column hypoxia and high H₂S conditions in the sediment.

4. Material and methods

4.1. Seagrass sampling

Zostera marina (Linnaeus 1753) specimens with and without leaf epiphyte-cover were collected from shallow coastal waters (less than 2 m depth) at Løgstør (Limfjorden, Denmark; 56°58'33.1″N 9°16'38.8″E) and brought back to the nearby Rønbjerg Marine Biological Station (Aarhus University, Denmark) in early June. Water temperature measured on site was 22.8°C (Testo 110, Testo AG, Lenzkirch, Germany; accuracy ±0.2°C), salinity was 30, and incident photon irradiance (400–700 nm) measured with a calibrated light meter (ULM-500, Walz GmbH, Effeltrich, Germany) varied from 600 (15 cm deep down in the eelgrass meadow) to 1500 µmol photons m⁻² s⁻¹ (at the top of the leaf, close to the surface of the water column) during midday. Collected seagrass shoots were kept in aerated tanks under artificial light mimicking light conditions at the sampling site. Three leaves partly covered with similar epiphyte loads were selected from three independent seagrass plants (N = 3). Two mid-leaf sections (approx. 4 cm long) either with or without epiphytes were sampled from the same leaf (i.e. six leaf sections in total). They were mounted side by side on microscope slides, held at each extremity with electrical tape, and slightly bent upwards to enable water circulation both below and above the leaf. They were then transferred to the experimental set-up (electronic supplementary material, figure S1).

4.2. Experimental set-up and design

The mounted leaves were positioned in the centre of a custom-made flow-chamber (see [18,73] for more details), wherein a constant laminar flow was maintained by a water pump immersed in a seawater tank (23°C, salinity = 30, constant aeration). The section containing the leaf fragments was illuminated evenly by a fibre-optic tungsten halogen lamp (KL-2500 LCD, Schott GmbH, Mainz, Germany) equipped with a heat filter and a collimating lens, positioned above the flow-chamber (approx. 0.5 m away) with a slight angle. Micro-profiles of different physico-chemical parameters (see following sections) were measured at two different flow conditions (slow and fast; described below) and four different incident photon irradiance levels (0, 110, 500 and 1050 µmol photons m⁻² s⁻¹) on bare leaves, i.e. control leaves, and leaves with dense microalgal epiphytic biofilms. The measured thicknesses of the epiphytic biofilms were 1.1 ± 0.1 mm, 2.6 ± 0.1 mm and 1.3 ± 0.06 mm (mean ± s.d.; n = 3, technical replicates per leaf) for biological replicates 1, 2 and 3, respectively. The flow velocity was estimated by measuring the distance travelled by small, suspended sediment particles during a defined time period. Slow and fast flow conditions corresponded to approximately 0.5 and 5 cm s⁻¹, respectively. The irradiance levels were chosen based on variable chlorophyll fluorescence measurements of rapid light curves [74], i.e. relative PSII electron transport rates versus irradiance (Mini PAM, Walz GmbH, Pfullingen, Germany) (electronic supplementary material, figure S2), and represented dark (0 µmol photons m⁻² s⁻¹), low (110 µmol photons m⁻² s⁻¹), saturating (500 µmol photons m⁻² s⁻¹) and high (1050 µmol photons m⁻² s⁻¹) light conditions. The scalar irradiance (PAR, 400–700 nm) at the leaf surface was measured in the flow-chamber with a spherical quantum sensor (US-SQS/L, Walz GmbH) connected to a calibrated quantum irradiance meter (ULM-500, Walz GmbH).

4.3. Microsensor measurements of O₂, temperature and nitric oxide

Oxygen concentration gradients were measured from the bulk seawater towards the leaf surface and throughout the epiphytic biofilm (when present) with a fast-responding (t₉₀ < 1–2 s) Clark-type O₂ microsensor (tip diameter 25 µm, OX-25, Unisense A/S, Aarhus, Denmark) in vertical increments of 100 µm. We considered the electrode at the leaf surface when the microsensor tip reached the leaf tissue surface but did not exert any pressure on it. The O₂ microsensor was linearly calibrated from readings in 100% air saturated seawater and anoxic seawater at experimental temperature and salinity. Temperature microprofiles were measured with a thermocouple microsensor (tip diameter 50 µm, T50, Unisense A/S) that was calibrated against readings of a high precision thermometer (Testo 110, Testo AG; accuracy ± 0.2°C) in seawater at different temperatures. A total of 48 profiles were measured for both O₂ and temperature under different experimental conditions of light, flow and epiphyte cover.

NO concentration profiles were measured with a Clark-type NO microsensor (tip diameter 100 µm, NO-100, Unisense A/S, Denmark). To calibrate the NO microsensor, anoxic (i.e. N₂ flushed) seawater was mixed with H₂SO₄ and KI, each with a final concentration of 0.1 M. The sensor was then immersed in the calibration solution and a zero reading was recorded after a stable signal was reached. Then 100 µM NaNO₂ was added to the calibration solution in three steps to create a four-point calibration (electronic supplementary material, figure S3). See also the manufactures website, www.unisense.com, for further details on the NO microsensor calibration procedure. After control measurements on bare leaves showing no NO production, NO profiles were measured in dark conditions only, on leaves covered with epiphytes. Between two and three technical replicates (i.e. pseudoreplicates) were measured on each leaf.

The temperature microsensor was connected to a thermocouple meter (Unisense A/S, Denmark), while the O₂ and NO microsensors were connected to a microsensor multimeter (Unisense A/S, Denmark). All meters were connected via USB cables to a PC. All microsensors, one after another, were vertically mounted on a motorized micromanipulator (Unisense A/S, Denmark) that was also connected to the PC. Data acquisition and microsensor positioning were controlled by dedicated software (Sensor TracePro, Unisense A/S), after manually positioning the tip of the microsensor at the leaf surface (defined as 0 µm on the figures) while observing the leaf and microsensor tip with a dissection microscope (Stemi SV6, Zeiss, Oberkochen, Germany).

4.4. Diffusive boundary layers and flux calculations

The outer limit of the DBL was determined as the intersection of the linear regression model fitted on the O₂ concentration gradient at the outermost surface (leaf or epiphytic biofilm, total depth interval of 200 µm spanned by three measurement points) with the linear model fitted on the bulk water O₂ concentration (depth interval of 200 µm spanning the first three measurement points of each profile). The effective DBL thickness was then estimated from the distance of the intersection to the biofilm/leaf surface. The O₂ fluxes between the surrounding water and the surface of bare leaves or the uppermost limit of the epiphytic biofilm were calculated using Fick's first law of diffusion

$$J_{{\mathrm{O}}_2}=-D_{{\mathrm{O}}_2}\cdot \frac{\text{d}C}{\text{d}z},$$ 4.1

where D_O2 is the diffusion coefficient of O₂ in seawater at experimental temperature (23°C) and salinity (30), i.e. 2.2328 × 10⁻⁵ cm² s⁻¹, and dC/dz is the slope of the linear O₂ concentration gradient in the DBL (total depth interval of 200 µm, i.e. three measurement points). Oxygen fluxes at leaf surfaces covered with epiphytes were also calculated according to equation (4.1), replacing D_O2 by the diffusivity of the biofilm, ϕD_E, estimated as 0.7 × D_O2 [75].

4.5. Microscale light measurements

A scalar irradiance microsensor with a spherical tip diameter of 50 µm [76] was mounted on a manually operated micromanipulator at 45° angle relative to the incident light. The scalar irradiance microprobe was connected to a fibre-optic spectrometer (USB 2000+, Ocean Optics Inc., Dunedin, FL, USA), interfaced to a PC running spectral acquisition software (SpectraSuite, Ocean Optics Inc.). To calibrate the scalar irradiance microprobe, the tip of the sensor was first positioned over a black non-reflective light well at the exact same position and distance from the light source as the seagrass leaf surface. This allowed us to quantify the downwelling irradiance and the light spectra of the vertically incident light, as in a collimated light field the downwelling irradiance and the downwelling scalar irradiance are identical [59]. After calibration, vertical profiles of spectral scalar irradiance were recorded in vertical steps of 0.2 mm from above the epiphytic layer (when present) towards the surface of the leaf. A total of 36 profiles were measured at low (110 µmol photons m⁻² s⁻¹), saturating (500 µmol photons m⁻² s⁻¹) and high (1050 µmol photons m⁻² s⁻¹) irradiance levels. Only one flow condition was tested on bare leaves, as the flow does not impact the scalar irradiance. However, profiles were recorded for both flow regimes on leaves with epiphytes, to test for potential enhanced shading effects at high flow rates due to compressed and more compacted epiphytic biofilms.

For each depth position, the measured scalar irradiance spectrum was integrated over 400–700 nm to quantify the PAR photon irradiance, as well as, the fractions of incident PAR irradiance. The absolute photon scalar irradiance at each depth was then calculated by multiplying with the known incident photon irradiance (in µmol photons m⁻² s⁻¹), previously measured with a calibrated photon irradiance meter (ULM-500, Walz GmbH)

$$E_0{(\text{PAR})}_z=\frac{{\int }_{\lambda =400}^{700}{E}_0{(\lambda )}_z\hspace{0.17em}\text{d}\lambda }{{\int }_{\lambda =400}^{700}{E}_d(\lambda )\hspace{0.17em}\text{d}\lambda }\cdot \varphi ,$$ 4.2

where E₀(PAR)_z is the PAR photon scalar irradiance (400–700 nm) at depth z; E₀(λ) is the scalar irradiance measured for the wavelength λ; E_d(λ) is the downwelling spectral irradiance measured over the black light well and Φ is the absolute incident photon irradiance (in µmol photons m⁻² s⁻¹).

The light attenuation percentage (table 2) was calculated as

$$\text{attenuation}=\frac{(\mathrm{PARbulk}-\mathrm{PARleafsurface})\times 100}{\text{PARbulk}}.$$ 4.3

The spectral scalar irradiance, E₀(λ), was measured in vertical steps of 0.2 mm from the leaf surface to the uppermost layer of the epiphytic biofilms and was plotted as transmittance spectra, i.e. as the percentage of the incident downwelling irradiance E_d(λ), as measured with the scalar irradiance microprobe positioned over a black light well at similar distance and position in the light field as the seagrass leaf.

Spectral attenuation coefficients of scalar irradiance, K₀(λ) over defined zones in the epiphytic biofilms were calculated as [77]

$$K_0(\lambda )=\frac{-\text{ln}(E_0{(\lambda )}_1/E_0{(\lambda )}_2)}{z_2-z_1},$$ 4.4

where E₀(λ)₁ and E₀(λ)₂ are the spectral scalar irradiances measured at depth z₁ and z₂, respectively.

4.6. Data analyses

All statistical analyses were performed using the R software, v. 2.15.0 [78]. For all the data analysed, i.e. DBL and TDD thicknesses, O₂ fluxes, light attenuation at the leaf surface, and leaf surface warming (temperature increase at the leaf surface compared to the water column temperature), normality (Shapiro's test) and homoscedasticity (Levene's test) of the data and residuals were checked. Linear mixed effect models [79] were preliminary built considering the light intensity (0, 110, 500 and 1050 µmol photon m⁻² s⁻¹), the flow rate (fast or slow) and the leaf surface (bare or epiphytes) as fixed factors and the replicate number as a random factor. The latter controlled for any replicate effect that may have influenced our results, as we performed repeated profiles on the same replicates in different flow and light conditions. The significance of the random term was first tested with the function ‘rand' of the package lmerTest [80]. The random term was removed when it was non-significant, and two- or three-way ANOVAs were carried out instead of mixed effect models. When significant effects of the factors were found, post hoc Tukey tests were used to discriminate the different groups. p-Values are reported with degrees of freedom and the exact test statistic in subscript. Data in tables and graphs are presented as means ± standard error (s.e.).

Data accessibility

Data supporting these findings are made fully available in electronic supplementary material and upon request from the authors. Raw data of microprofiles will also be available from the PANGAEA database, linked to the article title and with a DOI to follow.

Authors' contributions

F.N. and A.D. carried out the experiments (supervised by K.E.B.), analysed the data (under supervision of K.E.B. and M.K.) and wrote the manuscript with editorial input from all authors. M.K. and K.E.B. designed experiments and provided essential infrastructure.

Competing interests

The authors declare no conflicts of interests.

Funding

The study was funded by a grant from the Carlsberg Foundation (CF16-0899; K.E.B.), the Villum Foundation (grant no. 00028156; K.E.B.) and by a grant from the Independent Research Fund Denmark (grant no. DFF-1323-00065B; M.K.). F.N. was supported by the European Union's Horizon 2020 research and innovation programme under the Marie Skłodowska-Curie grant agreement (no. 701366), as well as, the IMAP network from the Future Ocean Cluster. A.D. was supported by the Swiss National Science Foundation.