Introduction

Coastal waters and estuaries are highly productive and ecologically valuable ecosystems. These systems are under increasing stress from anthropogenic disturbances due to sediment dredging, catchment runoff and urbanisation. These disturbances and the increasing frequencies of natural disturbances (e.g. flooding and cyclones) (Preen et al. 1995; Campbell and McKenzie 2004) directly reduce the distribution of ecologically important primary producers and inhibits the maintenance of healthy marine ecosystems (Dennison 1987; Duarte 1991; Alcoverro et al. 2001).

Seagrasses are often the dominant primary producers in coastal ecosystems and contribute to maintaining water quality by increasing the stability of sediments, biogeochemical cycling and trophic dynamics. The distribution of seagrasses is widely accepted as a barometer of coastal water quality (Duarte 1991; Gallegos and Kenworthy 1996; Biber et al. 2005). However, loss of seagrass biomass and suboptimal seagrass growth are usually detected when coastal ecosystems are already degraded and their poor condition is unable to be reversed by management actions. Consequently, for seagrass ecosystems, the development of a set of rapid and predictable sublethal stress indicators that respond to low-light fluxes have been the focus of intense research efforts in recent times (Kraemer and Hanisak 2000; Biber et al. 2005).

In response to the need for measures of sublethal stress thresholds for seagrasses, there have been many efforts to quantify the physiological responses of seagrasses in response to light deprivation over short time scales (days to weeks) commensurate with timescales of turbidity events in coastal waters (Longstaff et al. 1999; Ibarra-Obando et al. 2004; Biber et al. 2005). In situ and rapid assessments of sublethal stress have generally been restricted to morphological and structural photo-adaptive responses of seagrasses to light reduction, including measures of canopy height, shoot densities and leaf area indices (Longstaff and Dennison 1999).

Chlorophyll fluorescence techniques, such as Pulse Amplitude Modulated (PAM) fluorometery, offer a rapid in situ tool to develop sublethal physiological indicators of stress, detecting a response in seagrasses to light reduction within days. Chlorophyll fluorescence (fluorescence yield) can be measured and used to calculate the proportion of photons absorbed by the photosynthetic PSII reaction centres (i.e. the quantum yield), since fluorescence is inversely correlated to photosynthetic efficiency. In light-adapted plants, these photons are used for photosynthetic electron transport, described as the effective quantum yield (ΔF/Fm′). The effective quantum yield of a plant’s photosystem II (PSII) is \( \Delta F/{\text{F{m}\ifmmode{'}\else$'$\fi }} = F - {\text{F{m}\ifmmode{'}\else$'$\fi }}/{\text{F{m}\ifmmode{'}\else$'$\fi;}} \) where the minimum fluorescence (F), as measured immediately before the saturating pulse, and subtracted from the maximal fluorescence (Fm′), measured immediately after a saturating pulse of light and then divided by Fm′. F − Fm′ is described as the variable fluorescence (ΔF) (Beer et al. 2001; Ralph and Gademann 2005).

Research has focused on maximum quantum yield in dark adapted plants, but this has not been found to be a sensitive indicator of plant stress (Longstaff and Dennison 1999; Biber et al. 2005). In naturally fluctuating light climates there are predictable changes in maximal electron transport rates (ETRs), photosynthetic efficiencies and saturating irradiances (Beer et al. 1998; Ralph et al. 1998; Campbell et al. 2003). Such measures can be simply derived from rapid light curves (RLCs), using PAM fluorescence, and offer a more effective means of evaluating the photosynthetic performance of seagrasses as affected by their previous light history (Beer et al. 2001; Ralph and Gademann 2005). Nonetheless, there is an absence of manipulative studies that have investigated photosynthetic derivatives of chlorophyll fluorescence (e.g.ETRmax, photosynthetic efficiency and saturating irradiance), as physiological indicators, that respond to changes in light availability over relatively short timescales of days to weeks, before seagrass mortality.

In recent times, seagrasses in Hervey Bay have fluctuated considerably in response to anthropogenic disturbance (Preen et al. 1995; Campbell and McKenzie 2004). Hervey Bay is a subtropical embayment containing ∼2,000 km2 of fast growing seagrass species including Halophila ovalis, Halodule uninervis and Zostera capricorni. These species have relatively small rhizomes and roots with low carbohydrate storage capacity (Abal et al. 1994), making them susceptible to light reduction over short timescales (i.e. days to weeks) (Longstaff and Dennison 1999).

Currently there are no rapid methods to detect stress, in seagrasses, in response to light reduction before morphological changes or mortality occurs. In this study, we examined differences in the photosynthetic responses of two co-existing species of seagrass to 2-week light reduction, a period that seagrass meadows are often subject to severe depletion of light caused by anthropogenic and natural disturbances. Photo-adaptation was also investigated as a mechanism potentially employed by both species to survive severe light depletion. Chlorophyll fluorescence and, in particular, RLCs were used to rapidly assess if photo-adaptation could be detected under different short-term shading manipulations and before seagrass mortality.

Methods

Shading experiment

Site location

Two, 0 m depth, intertidal sites were chosen in Hervey Bay. The sites were located at Urangan (25°18.249′ 152°54.394′), and Burrum Heads (25°11.349′ 152°37.559′) (Fig. 1). Both sites consisted of mixed meadows of H. ovalis and Z. capricorni and were exposed at low tide for 2–3 h during a tidal cycle. The maximum tidal height ranges from 2 to 3 m.

Fig. 1
figure 1

Location of experimental shading experiments in Hervey Bay, Australia

Full size image

At both sites three 20 × 20 cm2 replicate plots were exposed to three different shade treatments. Three replicate control plots with no shade treatment were also marked, giving a total of 12 plots. The shade treatments consisted of 100 × 100 cm2 mesh cloths suspended on an aluminium structure 15 cm above the sediment. The Z. capricorni canopy reached a maximum of 10 cm at Urangan and 6 cm at Burrum Heads, while H. ovalis plants had a maximum length of 3 cm. The shade cloths were positioned so that plants received attenuated light during daylight. A 25 cm × 25 cm square was cut into the centre of each shaded plot, 15 cm into the sediment, ensuring all rhizomes were cut. Cutting prevented the shaded seagrass receiving photosynthates translocated, via the rhizomes, from non-shaded parts of the plant (Longstaff and Dennison 1999). The cuts were repeated several times during the duration of shading. The three treatments consisted of 90, 60 and 5% shading. Control plots of seagrass, without shade cloth, were also marked. The shade cloths were deployed for a total of 14 days, being changed every 3 days to reduce the effect of algal fouling.

At day 0, between 11:00 and 13:00 h during low tide and under sunny conditions, measures of RLCs, using a diving PAM, were made on light adapted leaves of H. ovalis and Z. capricorni from the centre of nine replicate plots to be treated and each of three control plots. Healthy green leaves were placed in a plastic ‘leaf clip’ in situ (Walz, Effeltrich, Germany) to record RLCs. The PAM light probe was attached to the leaf clip at a fixed distance from the leaf. The diving PAM automatically generated RLCs using an incremental sequence of 10 s actinic illumination periods, with light intensities increasing in the eight steps 0, 50, 150, 340, 580, 850, 1,180 and 1,760 μmol quanta m−2 s−1 photosynthetically active radiation (PAR). For each illumination period initial fluorescence (F) and following a saturating pulse of white light (800 ms of 8,000 μmol quanta m−2 s−1 PAR), maximum fluorescence (Fm′), were measured. The same procedure was followed after 14 days (14 days) of continuous shading.

Depth gradient experiment

Site location

Two sites were chosen at each depth of interest (0, 5 and 10 m) for analyses of photosynthetic variables of H. ovalis along a depth gradient.

Using the diving PAM and the procedure described above, the photosynthetic performance of H. ovalis was examined at three depths (0, 5 and 10 m), with two replicate sites for each depth (total of six sites). At each site, RLCs were measured, as described above, for ten replicate light adapted H. ovalis plants, between 11:00 and 13:00 h, under sunny conditions.

In situ underwater light measurements

Ambient underwater light was measured as photon flux using the diving PAM, which was calibrated underwater with a Li-189 light meter (LiCoR, Lincoln, NE, USA). Five photon flux measurements were made under every shade cloth at days 0 and 14 during PAM measurements (n = 75). Three to five measurements were made at each depth during PAM measures along the depth gradient. All light measurements were made under sunny conditions, avoiding cloudy periods. Mean values of light measures are presented.

Season

The shade experiment and depth gradient photosynthetic measures were completed within the month of November 2003, immediately prior to the onset of the wet season. The water temperature was 23°C for the duration of the shade experiment and the depth gradient photosynthetic measures.

Calculation of photosynthetic variables

Apparent photosynthetic ETRs were calculated as the product of effective quantum yield (\( F - {\text{F{m}\ifmmode{'}\else$'$\fi }}/{\text{F{m}\ifmmode{'}\else$'$\fi }} = \Delta F/{\text{F{m}\ifmmode{'}\else$'$\fi;}} \) where F is initial fluorescence, Fm′, is maximum fluorescence and ΔF is variable fluorescence), the incubation irradiance (I) and the absorbance factor (AF), i.e. the fraction of light absorbed by the leaf. This product was further multiplied by 0.5 because it was assumed that half the photons required for the movement of electrons along the photosystem pathways are absorbed by PSII (Schreibers et al. 1995). Effective quantum yield (ΔF/Fm′) was calculated from the RLC for the first of the eight actinic illuminations. AF values for H. ovalis and Z. capricorni were derived by measuring the proportion of light absorbed by single leaves according to the method described by Beer et al. (2001). The maximum ETR (ETRmax) and photosynthetic efficiency (α) were calculated by fitting the RLC data to an exponential function; \( {\text{ETR}} = {\text{ETR}}_{{\max }} \times (1 - \exp [ - \alpha ({\text{I}})/{\text{ETR}}]) \) (Jassby and Platt 1976) where ETR = electron transport rate and I = irradiance. The onset of light saturation (Ek) was calculated as ETRmax/α. In the few cases where a down-regulation drop in ETR was recorded, measures from high photon flux (e.g. 1,760 μmol quanta m−2 s−1 PAR) were removed before fitting data to the exponential model.

Data analysis

Shading experiments

One-way ANOVAs were used to determine the effects of plots prior to shading treatments on ETRmax, photosynthetic efficiency (α), minimum irradiance at which the plant is photosynthetically saturated (Ek), initial fluorescence yield (F), maximum fluorescence yield (Fm′), effective quantum yield (ΔF/Fm′) and AFs. Effective quantum yield and AF data from Urangan and Burrum Heads were arcsin square-root transformed prior to analysis.

Three-way ANOVAs were used to determine the effects of species, shade treatment and site on ETRmax, α, Ek, F, Fm′, ΔF/Fm′ and AF. Photosynthetic efficiency, ΔF/Fm′ and AF data from Urangan and Burrum Heads were arcsin square-root transformed prior to analysis. ETRmax and Fm′ data from Burrum Heads were loge transformed.

Depth gradient and shading

ANOVAs were used to test for significant differences in mean photosynthetic variables in Halophila ovalis (ETRmax, α, Ek, F, Fm′, ΔF/Fm′ and AF) (n = 3–10) among both sites (Urangan and Burrum Heads) and depths (0, 5 and 10 m), and among all shade treatments (control, 5, 60 and 90%) at the two sites, at 14 days (n = 14). Post hoc Bonferroni tests were used to test for significant differences among means. ETRmax and Ek, data were loge (x + 1) transformed prior to analysis. All analyses were performed using SYSTAT (Version 10.2).

Results

Before shading

Immediately prior to the shading manipulations at day 0, there were no significant differences (P < 0.05) for photosynthetic variables among each of the treatment plots at either site, Urangan and Burrum Heads.

Shading manipulations

Three-way ANOVA revealed a significant interaction between site, species and shade treatment for ETRmax and saturating light intensity (Ek) (Table 1). Post hoc, Bonferroni, analysis showed the interaction was due to a significantly higher mean (±SE) ETRmax (178.24 ± 17.53) for non-shaded (control) Z. capricorni at Burrum Heads compared with the means of non-shaded Z. capricorni (96.87 ± 20.61) and Halophila ovalis plants (81.21 ± 5.32) at the Urangan site (Fig. 2). At Burrum Heads the mean ETRmax for Z. capricorni non-shaded plants (178.24 ± 17.53) was also significantly higher than 60% (94.62 ± 23.69) and 90% (83.78 ± 7.64) shaded plants (Fig. 2). At Urangan, means of ETRmax for 5% (103.95 ± 12.16) and 60% (91.26 ± 10.00) shaded Z. capricorni were higher than the mean of 90% shaded Z. capricorni (74.49 ± 2.29) plants (Fig. 2). Mean ETRmax of 90% shaded H. ovalis at Burrum Heads (78.02 ± 14.86) and Urangan (48.61 ± 1.92) was significantly lower compared with non-shaded and 5% shaded H. ovalis at both sites.

Table 1 Three-way ANOVA of the effect of site (Urangan and Burrum Heads), species (Halophila ovalis and Zostera capricorni) and shade treatment (0, 5, 60 and 90% light reduction) on photosynthetic variables in Hervey Bay, n = 3
Full size table
Fig. 2
figure 2

Effect of shading on photosynthetic variables ETRmax, α, Ek, F, Fm′, ΔF/Fm′ and AF, at Urangan (UG) and Burrum Heads (BH) sites. Values are shown at 0 day and after 14 days shading for control, 5, 60 and 90% shade treated plots for H. ovalis (white bars) and Z. capricorni (shaded bars) (mean ± SE, n = 3)

Full size image

For Ek the significant three-way interaction (Table 1) was best explained by significantly higher mean Ek of non-shaded Z. capricorni (733.07 ± 92.40) and H. ovalis (573.23 ± 51.51) at Burrum Heads compared with non-shaded Z. capricorni (449.21 ± 113.64) and H. ovalis plants (365.06 ± 50.58) at Urangan (Fig. 2). Species interactions were best explained by higher mean Ek for 60% (294.73 ± 54.93) and 90% (209.70 ± 9.07) shaded Z. capricorni compared with mean Ek for 60% (209.70 ± 21.32) and 90% (120.07 ± 3.88) shaded H. ovalis, respectively, at Urangan, but no such differences were found at Burrum Heads (Fig. 2). At Burrum Heads there was no difference between mean Ek values in 60% shaded (342.86 + 92.70) and 90% shaded Z. capricorni (331.03 ± 72.68), but at Urangan mean Ek of 60% shaded Z. capricorni (294.73 ± 54.93) was significantly higher than the mean Ek of 90% shaded plants (209.70 ± 9.07) (Fig. 2). At Burrum Heads the mean Ek of 5% shaded Z. capricorni (534.65 ± 69.62) was significantly higher than 60% shaded (342.86 + 92.70) Z. capricorni, whereas at Urangan no difference was found between mean Ek of 5% shaded (328.66 ± 10.49) and 60% shaded (294.73 ± 54.93) Z. capricorni (Fig. 2).

A shading effect was evident for all parameters measured, except F. At both sites, 90% shade treated H. ovalis (1,100.67 ± 69.20 at Urangan; 230.33 ± 12.35 at Burrum Heads) showed higher mean Fm′ compared with control H. ovalis (47.33 ± 113.87 at Urangan; 187.33 ± 42.29 at Burrum Heads) plants (Fig. 2, Table 1). At both sites, 90% shaded H. ovalis had significantly higher mean α(0.2813 ± 0.0003 at Burrum Heads; 0.4047 ± 0.0045 at Urangan) and mean ΔF/Fm′ (0.6943 ± 0.0068 at Burrum Heads; 0.7553 ± 0.0180 at Urangan) compared with non-shaded (α − 0.2120 ± 0.0189 at Burrum Heads and 0.2267 ± 0.0156 at Urangan; ΔF/Fm′ − 0.5143 ± 0.0088 at Burrum Heads and 0.6070 ± 0.0255 at Urangan) and 5% shaded (α − 0.2010 ± 0.0492 at Burrum Heads and 0.2650 ± 0.0420 at Urangan; ΔF/Fm′ − 0.4647 ± 0.0799 at Burrum Heads and 0.7423 ± 0.0234 at Urangan) H. ovalis plants, at respective sites. At Urangan, 90% shaded Z. capricorni had significantly higher mean α (0.3557 ± 0.0062) compared with non-shaded (0.2220 ± 0.0140), 5% (0.3147 ± 0.0278) and 60% (0.3207 ± 0.0296) shaded Z. capricorni. At Urangan, ΔF/Fm′ was significantly higher for 90% shaded Z. capricorni (0.7683 ± 0.0143) compared with non-shaded Z. capricorni (0.6940 ± 0.0381). For ETRmax and saturating light intensity (Ek), the opposite was generally observed for both species at both sites, where 90% shaded plants had significantly lower values than control and 5% shaded plants, especially in H. ovalis (Fig. 2). At Urangan 90% shaded Z. capricorni (0.6368 ± 0.0154) and H. ovalis (0.6632 ± 0.0023) had higher AFs than respective 5% shaded Z. capricorni (0.6120 ± 0.0297) and H. ovalis (0.5060 ± 0.0481) plants (Fig. 2).

Halophila ovalis: depth gradient and shading experiment

Plots of ETRmax, photosynthetic efficiency (α), saturating irradiance (Ek), effective quantum yield (ΔF/Fm′) and AF against depth and shading treatments revealed a decline in ETRmax and Ek and a general increase in α and ΔF/Fm′ with increasing depth and shade (Fig. 3). Significant differences among means of each variable were found (Table 2). Mean ETRmax of H. ovalis in non-shaded and 5% shaded plots at Urangan were significantly lower than plants at 0 m but not significantly different than ETRmax of plants at 5 m (Fig. 3). In contrast ETR max and Ek of non-shaded plants at Burrum Heads were no different to plants at 0 m, while 90% shaded H. ovalis had mean ETRmax and Ek values that were significantly lower than plants at 0 m but not significantly different compared with plants growing at 5 and 10 m (Fig. 3). Mean α and ΔF/Fm′ values in control and 5% shaded plots were not significantly different compared with mean values at 0 m. Similarly mean α and ΔF/Fm′ recorded for 60% shaded H. ovalis were not significantly different when compared with mean α and ΔF/Fm′ of H. ovalis at 5 m (Fig. 3). Mean α and ΔF/Fm′ values in 90% shaded H. ovalis were significantly lower than plants at 0 m but not significantly different compared with H. ovalis at 5 and 10 m.

Table 2 Halophila ovalis: ANOVA of the effect of shade plots (control, 5, 60 and 90% at two sites) and depths (0, 5 and 10 m) in Hervey Bay
Full size table
Fig. 3
figure 3

Effect of depth (0, 5 and 10 m) on photosynthetic variables (ETRmax, α, Ek, F, Fm′, ΔF/Fm′ and AF) for H. ovalis. Each graph includes the effect of shading treatments at 14 days (control, 5, 60 and 90% shade plots) at Urangan (grey bars) and Burrum Heads (white bars). Photon flux density at each depth and shade treatment are also shown (mean ± SE, n = 3). Different letters above bars indicate significant differences (Bonferoni) between treatment means at P < 0.05

Full size image

Discussion and conclusions

Light is a key factor controlling the photosynthetic performance of seagrasses. Both long-term and short-term light deprivation directly influence seagrass photosynthesis and survival (Longstaff et al. 1999; Ruiz and Romero 2001; Peralta et al. 2002). We found short-term changes in the physiology of seagrasses by manipulating light during experimental shading. Of all the photosynthetic variables tested (ETRmax, Ek, α, Fm′, ΔF/Fm′ and AF) ETRmax and Ek showed the most consistent responses among shade treatments and depth gradients, characterised by reductions in ETRmax and increases in Ek, with light depletion. Both Halophila ovalis and Z. capricorni exhibited typical responses to shading yet H. ovalis had a more conspicuous response, suggesting that this species is more sensitive to reduced transient light than Z. capricorni. The results suggest a tolerance and acclimation to light deprivation in both species of seagrass between 5 and 60% of surface irradiance and are consistent with findings that H. ovalis has a limited tolerance to light deprivation when compared with morphologically larger species of seagrass (Longstaff et al. 1999).

There was a consistent trend among species and sites, as shown by the decreasing ETRmax and Ek and increasing α and ΔF/Fm′ with light depletion, and a strong separation of photosynthetic responses between high and low light treatments. These photosynthetic responses are evidence that photosynthetic parameters derived from RLCs (Beer et al. 2001) are a useful tool that can be used to evaluate short-term responses of seagrasses between highly differentiated light climates. The parameters sensitive to 90% shading in H. ovalis were reduced ETRmax, increased α, reduced saturating irradiance (Ek) and increased effective quantum yield (ΔF/Fm′). While a reduction in both ETRmax and Ek suggests a limited photosynthetic capacity such reduction indicates photo-acclimation and improved photon capture for conversion to chemical energy. Similarly, the increase in both α and ΔF/Fm′ in response to shading indicates a greater proportion of photons were used in photosynthesis and a more efficient use of light (Beer et al. 2001). These photo-adaptive responses to irradiance or photon flux deprivation are analogous with other photo-adaptive responses such as chlorophyll increases under low light conditions (Dennison and Alberte 1986; Longstaff and Dennison 1999). These findings are consistent with reports of reductions in photosynthetic performance and mortality in seagrasses in response to shading (Longstaff and Dennison 1999; Longstaff et al. 1999) and depth (Schwarz and Hellblom 2002; Durako et al. 2003).

The differences in photosynthetic performance of both seagrass species among shade treatments at different sites are easily explained by site features influencing light fluxes and the species responses to these differences. At low tide there was little obvious difference in light climates between sites, yet during tidal inundation high amounts of suspended fine ‘silty’ sediments at Urangan were evident more so than at Burrum Heads, where the sediments have a high sand composition and are less inclined to be re-suspended during tidal flux. These high silt loads at Urangan resulted in lower light availability and more pronounced shade-type responses (e.g. lower Ek) in both seagrass species at Urangan, compared with seagrasses at Burrum Heads. In addition, the fouling of shade screens appeared heavier at the Urangan site, despite regular cleaning, and may also have contributed to lower photon fluxes under the shade screens. At both sites Halophila ovalis was more sensitive to changes in light availability than Zostera capricorni and thus exhibited the greatest reduction in ETRmax and Ek and the highest increases in photosynthetic efficiencies in response to shading and in situ light depletion at Urangan. A likely explanation is that H. ovalis is less structurally complex and faster growing than Z. capricorni, with low storage capacity for carbohydrates that can be used for growth during periods of low light (Longstaff et al. 1999).

The increase in effective quantum yield (ΔF/Fm′) of both seagrass species under 90% shading, relative to controls, was due to elevated Fm′ values. Elevation of Fm′ suggests an increase in photochemical efficiency, a response that is consistent with photo-acclimation of plants to shading. This contrasts with Longstaff et al. (1999) who found that quantum yield in dark adapted H. ovalis remained unchanged during complete light deprivation for 24 days. However, experimental shading units deployed by Longstaff and Dennison (1999) and Longstaff et al. (1999) showed photo-adaptive responses of reduced chlorophyll a and sugar concentration in H. ovalis leaves after 15 days of shading that are consistent with findings of reduced ETRmax, increased α and increased Ek in the present study. In the present study no change in ΔF/Fm′ was found for Z. capricorni. Biber et al. (2005) concluded that the maximum quantum yield of Z. marina in response to light deprivation was not a sensitive indicator of chronic stress as Z. marina was able to acclimate to ambient light conditions. A close examination of Biber et al. (2005) shows a clear stress response in healthy leaves, with quantum yield values of 0.5 at 4 weeks, 1 week prior to an observed decline in shoot numbers and leaf area. Similarly, they show that quantum yield values in Halodule wrightii, in dark conditions, declined to 0.6 after 3 weeks and 0.2 after 4 weeks, at least a week prior to noticeable declines in shoot numbers. In the context of the clear photosynthetic responses to light reduction found in the present study, this result demonstrates the importance of using a suite of photosynthetic variables, such as ETRmax, α, Ek and ΔF/Fm′, derived from RLCs to assess short-term responses of seagrasses to light deprivation, not measures of maximum quantum yield alone.

The influence of depth on ETRmax and Ek implies a strong separation of irradiance history between depths of 0 and 5–10 m. After 2 weeks of exposure to 90% reduced light availability, H. ovalis exhibited shade-type physiological responses, i.e. lower ETRmax and Ek, similar to plants growing at 5–10 m. The similarity in the light climates between 90% shaded plants and ambient light at 5–10 m is the likely explanation. The 40% lower ETRmax and 67% lower Ek values in 90% shaded H. ovalis compared with controls, was generally lower than depth related reductions of ETRmax and Ek from 0 to 5 m and from 0 to 10 m. Intertidal plants may be more susceptible to light reduction than subtidal plants and exhibit greater photo-adaptive responses in order to optimise light acquisition and maximise the use of low-light climates. In addition, fouling of shade cloths in between cleaning may have lowered available photon flux density (PFD) below recorded levels and contributed to a heightened photo-adaptive response. Comparative lowering of ETRmax and Ek values has been recorded for Halophila stipulacea (Schwarz and Hellblom 2002) and H. ovalis (Ralph 1996). In order to optimise light acquisition and maximise their use of low-light climates, these plants can diurnally regulate their saturating irradiance as detected in a number of seagrass species (Ralph et al. 1998; Campbell et al. 2003). We have also measured a three to fourfold increase in light saturating irradiance, between 0600 and 1200 hours, in deepwater H. ovalis (unpublished data).

Absorbance factors have been shown to vary considerably with geographic location (Silva and Santos 2003), depth (Schwarz and Hellblom 2002), species (Beer et al. 2001; Campbell et al. 2003; Lan et al. 2005) and leaf age (Enríquez et al. 2002). In Hervey Bay there was no change in AF with depth, but the high AFs of dark shaded (90%) H. ovalis at Urangan were indicative of a photo-adaptive response but were unlikely alone to be a reliable indicator of light reduction. The general lack of change in AFs with depth and shading suggests that factors (e.g. time of year, nutrient status) other than light attenuation may be influencing AFs over time scales of weeks to months.

The stronger responses to shading suggest that H. ovalis was more vulnerable to light deprivation than Z. capricorni and that H. ovalis at depths of 5–10 m would be more vulnerable to light deprivation than intertidal populations. Intertidal H. ovalis appears to be more vulnerable to reductions in irradiance than intertidal Z. capricorni, implicating local light conditions as important influences on species specific photosynthetic performance. The lower photosynthetic responses of H. ovalis at depths greater than 5 m compared with intertidal plants, also suggest a greater vulnerability to light deprivation with increased depth. Compared with large seagrasses, the more pronounced responses of H. ovalis to light reduction may be due to differences in the capacity of small seagrasses to store carbohydrates and maintain photosynthetic performance (Czerny and Dunton 1995; Kraemer and Alberte 1995; Longstaff et al. 1999). The short-term response of seagrasses to in situ reductions in light availability is likely to result in reduced growth rates and therefore can provide a rapid and precise indicator of sub-lethal stress. The photo-adaptive responses from experimental light reduction and the measures of light attenuation and saturating irradiances along depth gradients, particularly at lower depth limits of survival, could also be used to predict reductions in photosynthetic capacity before mortality eventuates.

Both species showed a photo-adaptive response to light depletion that may help them tolerate short-term yet severe reductions in light commonly encountered in Hervey Bay. Responses of ETRmax, saturating irradiance (Ek) and photosynthetic efficiency (α) to light reduction were the most consistent and are therefore likely to provide the best indicators of photo-adaptation and possible seagrass stress.