Introduction

Diatom blooms are observed in aquatic ecosystems worldwide and cause serious environmental problems (e.g., Hijnen et al. 2007; Leland et al. 2001). Many studies have focused on the role of environmental factors such as light (Marra and Heinemann 1982), turbulence (Thomas and Gibson 1990), and nutrients (Ferris and Lehman 2007; Sommer 1986) in the dynamics of diatom populations in aquatic ecosystems. Light intensity has been shown to be the main factor controlling the timing of the first occurrence of spring diatom blooms (e.g., Znachor et al. 2013; Letelier et al. 2004) and driving phytoplankton succession during the initial phase of the spring bloom in the north-western Mediterranean Sea (Brunet et al. 2013). Phytoplankton composition is highly related to the light regime of aquatic habitats (Robinson and Rushforth 1987). Light intensity is one of the important factors regulating phytoplankton community structure in Lake Nieuwe Meer, the Netherlands (Huisman et al. 2004). In China’s Yanghe Reservoir, light penetration through the water column was considered a major force in affecting phytoplankton morphology during seasonal succession (Su et al. 2014).

The selective role of light on the distribution of phytoplankton communities mainly induces the capacity of light adaptation in each species. Other studies have shown that photosynthetic and growth rate in many phytoplankton species can be affected by light conditions (Dimier et al. 2009; van Leeuwe et al. 2005; van de Poll et al. 2007). Some diatom species adapted to high light intensities showed high growth rates (Wang et al. 2013), high photosynthesis rates, high light saturation intensities (Cheah et al. 2013), and efficient responses to photoinhibition (Stamenkovic and Hanelt 2013) under high light regimes. Phytoplankton from a low-light lake (Grande Lake) were better adapted to a low-light environment in winter with higher maximum photosynthetic rates than plankton from a high-light lake (a relict oxbow lake) (Rodríguez et al. 2012), and phytoplankton from low-light sites had lower concentrations of cellular xanthophyll-cycling pigments than phytoplankton from high-light sites (Polimene et al. 2014). The Chla content which can reflect the photoacclimation status was high in low light and low in high light (Brunet et al 2011). Dimier et al. (2009) studied the xanthophyll cycle in six marine phytoplanktonic species and found that the high light-adapted species presented both a low quantity and diversity of photosynthetic accessory pigments and a high pool of xanthophyll-cycling pigments per cell compared to the low light-adapted species. Barnett et al. (2014) suggested that a balance between the physiological and behavioral photoprotective responses existed in intertidal marine diatoms and NPQ is clearly more efficient in non-motile epipsammon than motile epipsammon species.

However, the mechanisms which result in differences in photosynthetic adaptation to light intensity remain unclear. In this study, we hypothesized that diatoms from different light regimes would develop their specific photoacclimation characteristics and physiological characteristics in response to the light regime at a smaller, more local scale. To evaluate the adaptability of diatoms to light regime (low light and high light intensity), we isolated eight strains of diatoms from two high-light habitats (light intensity at a 0.5-m depth greater than 300 μmol photons m−2 s−1 in spring) and a low-light habitat (light intensity at a 0.5-m depth less than 100 μmol photons m−2 s−1 in spring) and compared diatom growth rates and chlorophyll fluorescence parameters under five light intensities (2 to 80 μmol photons m−2 s−1). We attempted to explore (1) the light-adaptive characteristics of low light-adapted (LLA) and high light-adapted (HLA) diatoms under various light intensities and (2) the physiological regulation of LLA and HLA diatom response to light stress. The results from this study will further our understanding of acclimation to the light regime and physiological strategies in response to light stress in these diatom species.

Six diatom species were cultured in this study: Stephanodiscus hantzschii, Fragilaria crotonensis, Nitzschia palea, Synedra ulna, Stephanodiscus parvus, and Synedra acus, and two Cyclotella meneghiniana strains were used in this study, which represent a wide range of diversity in terms of cell sizes and taxa (Table 1). These diatoms are either nanophytoplankton (2 to 20 μm) or microphytoplankton (20 to 200 μm) based on cell size. S. hantzschii, a non-motile species, is common in rivers (Kim et al. 2007; Wang et al. 2012); the distribution of S. hantzschii is mainly influenced by water column turbulence (Yang et al 2011). C. meneghiniana is often a dominant species in relatively calm water conditions (Qi et al 2012; Zeng et al. 2007) and is considered as a large niche breadth species (Wang et al 2010). S. parvus is found in many temperate lakes at low to modest relative abundance (Sokal 2007). N. palea is considered a highly adaptable species regarding light regime (Rodríguez and Pizarro 2007). F. crotonensis, S. aulna, and S. acus are also regarded as highly flexible to light intensity (Bouvy et al. 2006; Schmitt and Nixdorf 1999; Tuji 2000).

Table 1 Information of diatoms used in this study
Full size table

Materials and methods

We isolated eight strains (Table 1) of diatoms from three different habitats where diatom blooms have been frequently occurring. The diatoms were divided into two groups (high light-adapted (HLA) diatoms and low light-adapted (LLA) diatoms) according to the results of preliminary experiments. These experiments cultured diatoms at eight light intensities (2 to 300 μmol photons m−2 s−1) and found that diatoms from the Hanjiang River had high growth rates and high photosynthetic activities at light intensities lower than 25 μmol photons m−2 s−1 but showed photoinhibition at 80 μmol photons m−2 s−1. Diatoms from Lake Taihu and Guanqiao Pool showed light-limited growth rates at 2 μmol photons m−2 s−1, but their growth rates increased with increasing light intensities. According to the response of growth and chlorophyll a fluorescence parameters to light stresses in diatoms from different habitats, diatoms from Hangjaing River were defined as LLA diatoms and diatoms from Lake Taihu and Guanqiao Pool were defined as HLA diatoms.

During spring, S. hantzschii was the dominant species in the Hanjiang River (85 μmol photons m−2 s−1 at a 0.5-m depth) while C. meneghiniana was the dominant species in Lake Taihu (564 μmol photons m−2 s−1 at a 0.5-m depth) and Guanqiao Pool (485 μmol photons m−2 s−1 at a 0.5-m depth). The light intensity in Lake Taihu was estimated from an equation determined from long-term monitoring at the Lake Taihu Ecological Experiment Station (Zhang et al. 2004). Light intensity at Guanqiao Pool and Hanjiang River was measured by using an illuminometer (model wi97178, Ruida, China) in the spring of 2010. Sub-dominant species found in the Hanjiang River were F. crotonensis, N. palea, and S. ulna. In Lake Taihu, the dominant diatoms were S. parvus and S. acus. The Hanjiang River watershed includes some mountainous areas and receives a large amount of snowmelt from these areas in early spring. During this time, the Hanjiang River flows fast and turbid, and the range of measured discharge was 800 to 8000 m3 s−1 in 2010 (Yang et al. 2012). In Gonghu Bay (Lake Taihu), the average Secchi depth was 117 cm in the spring of 2010 (Qi et al. 2012).

Diatoms were isolated according to standard phycological methods (Stein and Hellebust 1979). Monoclonal diatoms of each strain were cultured in D1 medium (Zhang and Huang 1991). Concentrated diatom cells were cleaned using H2SO4 and observed under scanning electron microscope (SEM, JSM-6390LV, Japan) (Zhang and Huang 1991). Identification of algal species was based on the shape and pattern of the surface of the diatom shell (Hu and Wei 2006).

Growth experiments were conducted in illuminated incubators with a 14/10 light/dark cycle and 18 °C/16 °C temperature regimes. The light intensity was 2, 12.5, 25, 60, and 80 μmol photons m−2 s−1 with cool-white fluorescent light. The 80 μmol photons m−2 s−1 was the highest light treatment in the present experiment because 80 μmol photons m−2 s−1 was high enough to cause varying degrees of photoinhibition in all LLA diatoms in our preliminary experiment. Each 250-mL flask contained 200-mL culture at an initial algae concentration of about 5 × 107 ± 1.5 × 106 cells L−1. Triplicate cultures were used for each treatment. All cultures were regularly shaken two times per day to prevent diatoms from adhering to flasks. In order to eliminate the effects of nutrient depletion on the growth and physiological characteristics of diatoms, all the cultures were transferred weekly to fresh D1 medium. The experiment lasted 16 days. Samples were harvested every second day to monitor cell density and photosynthetic parameters. The results are given as the average value among the replicates with the standard deviation.

Growth rates

Diatom cells preserved with 1 % Lugol’s iodine solution were counted using counting chambers under an Olympus CX21 light microscope. The specific growth rate of each strain was calculated using Eq. (1):

$$ \mu =\left[ \ln {B}_{\mathrm{t}2}- \ln {B}_{\mathrm{t}1}\right]/\varDelta t $$
(1)

Where μ is the specific growth rate (day−1), B t1 is the initial cell density; B t2 is the cell density on the last day of the logarithmic growth phase.

Chlorophyll a (Chla) concentration

Chlorophyll a concentration was measured using the spectrophotometric method (Stein and Hellebust 1979). A 5-mL sample was collected from each treatment and concentrated by centrifugation at 100 × g for 10 min. Then the concentrates were dissolved in 5-mL 90 % acetone and kept in the dark at 4 °C overnight for Chla extraction. The Chla concentrations were determined from absorbance at 665, 645, and 630 nm with a spectrophotometer (Stein and Hellebust 1979) (Shimadzu UV2550, Japan).

$$ \mathrm{C}\mathrm{h}\mathrm{l}a\left(\mathrm{mg}\ {\mathrm{L}}^{-1}\right)=11.6\ {\mathrm{OD}}_{665}-1.31\ {\mathrm{OD}}_{645}-0.14\ {\mathrm{OD}}_{630} $$
(2)

Chlorophyll a are expressed as pg Chla per microgram of carbon (pg Chla μg−1 C). The carbon content of each strain was estimated by using biovolume (Warnaars and Hondzo 2006): log10 C = 0.758 log10 V − 0.422 (for diatoms), where C is cell carbon in micrograms and V is cell volume in μm3.

Chlorophyll a fluorescence parameters

Photosynthetic performance was measured using a pulse amplitude-modulated fluorescence monitoring system (Phyto-PAM, Walz, Germany). Subsamples of each culture were transferred to quartz glass cuvettes and positioned in the detector system. The actinic light level and saturating pulse level were set at 264 μmol photons m−2 s−1 at a length of 180 s and 4000 μmol photons m−2 s−1 at a length of 500 ms, respectively.

The effective photochemical efficiency

The effective quantum yield of PS II (ΔF/F’ m, Y) was measured using the quenching method (Schreiber et al. 1994). Values of ΔF/F’ m were calculated based on Eq. (3).

$$ \varDelta F/F{'}_{\mathrm{m}}=\left[F{'}_{\mathrm{m}}-{F}_{\mathrm{s}}\right]/F{'}_{\mathrm{m}} $$
(3)

Where F s is the fluorescence at actinic light, F’ m is the maximum fluorescence after saturating light pulses. ΔF/F’ m describes the actual quantum yield of photochemical conversion of excitons already migrating in the light-harvesting complex of PSII (Hartig et al. 1998).

Non-photochemical quenching (NPQ).

NPQ is a sensitive parameter to assess excess heat dissipation in algae (Lavaud et al. 2004). NPQ was calculated by Eq. (4) (Kashino et al. 2002).

$$ \mathrm{N}\mathrm{P}\mathrm{Q}=\left({F}_{\mathrm{m}}-F{'}_{\mathrm{m}}\right)/F{'}_{\mathrm{m}} $$
(4)

Where F m is the maximum fluorescence yield following dark adaptation of 15 min, and F’ m is the maximum fluorescence yield of light-adapted sample.

Rapid light response curve (RLC).

RLCs were used to demonstrate the relative change of photosynthetic performance under different light intensities (Ihnkena et al. 2010). Each sample was exposed to each of 12 increasing actinic light levels for 10 s (from 3 to 1480 μmol photons m−2 s−1), and the rETR vs PAR curve (RLC) was plotted using the model of Platt et al. (1980) to obtain photosynthetic parameters (α, rETR max, I k) by the least-squares method.

$$ \mathrm{r}ETR=\mathrm{r}ET{R}_{\max}\cdot \left(1-{\mathrm{e}}^{-\alpha \cdot \mathrm{P}\mathrm{A}\mathrm{R}/\mathrm{Pm}}\right)\cdot {\mathrm{e}}^{-\beta \cdot \mathrm{P}\mathrm{A}\mathrm{R}/\mathrm{Pm}} $$
(5)

where α is the initial slope of RLC and reflects the photon-capturing efficiency under light-limiting conditions. The rETR max represents maximum relative electron transport rate of the RLC (Genty et al. 1989). Effects on rETR max are in components of the photosynthetic electron transport chain (Behrenfeld et al. 2004; MacIntyre et al. 2002) and attributed to changes in Calvin cycle enzyme (RUBISCO) concentration and activity.

$$ {I}_{\mathrm{k}}=\mathrm{r}ET{R}_{\max }/\alpha $$
(6)

The light saturation point of photosynthesis (I k) indicates the PAR at which photosynthesis is saturated.

Cell ultrastructure

Samples (50 mL) of algal culture were collected under sterile conditions after a 5-h light exposure under low (LL, 2 μmol photons m−2 s−1), medium (ML, 25 μmol photons m−2 s−1), and high (HL, 80 μmol photons m−2 s−1) light intensities. Triplicate cultures were used for each treatment. Medium light intensity was used as control. Each sample was centrifuged at 50 × g for 6 min under 4 °C. The pellet was suspended in 2.5 % glutaraldehyde and kept at 4 °C overnight and then washed with 0.1 M phosphate buffer (pH 7.4). After post-fixation in 1 % osmium tetroxide for 2 h at 20 °C, the samples were dehydrated by serial transfers through a progressive aqueous–acetone series (30–100 %) and embedded in Spurr’s resin. Sections were cut and subsequently stained with uranyl acetate (Stempak and Ward 1964) followed by lead citrate. Sections were viewed by transmission electron microscopy (TEM, Hitachi H-7650) operating at 80 kV. From each sampling, 10 TEM micrographs were examined. All quantitative analyses were made on sections through the cell equator.

Statistical analysis

Origin 8.0 (Microcal Software, USA) was used for graphical analyses, and SPSS 16.0 (Windows Software, Inc., USA) was used for statistical analyses. Differences in growth rate, Chla content, and RLC parameters were compared using one-way ANOVA followed by Tukey post-hoc tests. Results were considered statistically significant using α-level of 0.05.

Results

Growth of diatoms

The average growth rate of LLA diatoms (0.034 ± 0.005 day−1) was significantly higher than that of HLA diatoms (0.010 ± 0.002 day−1) although all eight diatoms showed low growth rates at a light intensity of 2 μmol photons m−2 s−1 (Fig. 2c). The cell number of F. crotonensis and S. ulna showed a slight decrease on day 3 but recovered rapidly from day 5 (Fig. 1a). There was no obvious decrease in S. hantzschii and N. palea on day 3. The cell densities of all four HLA diatoms decreased on day 3 but increased during the later growth phase. Differences of growth rates were found between LLA and HLA diatoms at light intensity of 12.5 μmol photons m−2 s−1 (Fig. 1b). The average growth rate of LLA diatoms was 0.128 ± 0.023 day−1 (Fig. 2c), and S. hantzschii showed relatively high growth rate at 12.5 μmol photons m−2 s−1 (Fig. 2a). The cell density of HLA diatoms increased slightly, and no obvious decrease was found on day 3 at 12.5 μmol photons m−2 s−1. At 25 μmol photons m−2 s−1 (Fig. 1c, Fig. 2a), the growth rates of LLA diatoms were still higher than those of HLA diatoms and S. hantzschii was also the highest one. The average growth rate in LLA diatoms was 2.09 and 1.66 times that of HLA diatoms at 12.5 μmol photons m−2 s−1 (F = 16.672, P = 0.006) and 25 μmol photons m−2 s−1 (F = 9.912, P = 0.020), respectively (Fig. 2c).

Fig. 1
figure 1

Growth curves of eight diatoms at five light intensities. Data are means ± standard deviations n = 3

Full size image
Fig. 2
figure 2

Specific growth rates of LLA and HLA diatoms under five light intensities. Data are means ± standard deviations of specific growth rates, n = 3. a Specific growth rates in four LLA diatoms. b Specific growth rates in four HLA diatoms. c The average values of specific growth rates in LLA and HLA diatoms. Asterisks show statistical significance with P < 0.05 (*) and P < 0.01 (**)

Full size image

The cell density of HLA diatoms grew fast with increasing light intensities. The average growth rate of HLA diatoms exceeded that of LLA diatoms at 60 μmol photons m−2 s−1 (Fig. 1d, Fig. 2c). The cell density of C. meneghiniana(T) exceeded S. hantzschii on day 3. The cell density of S. parvus exceeded N. palea on day 11. At a light intensity of 80 μmol photons m−2 s−1 (Fig. 1e), HLA diatoms had high growth (Fig. 2c) and C. meneghiniana(T) showed the highest growth rate while LLA diatoms grew slowly (Fig. 2b). The average growth rate in HLA diatoms was 1.45 and 1.69 times that of the LLA diatoms at 60 μmol photons m−2 s−1 (F = 8.871, P = 0.025) and 80 μmol photons m−2 s−1 (F = 8.181, P = 0.029), respectively (Fig. 2c).

LLA and HLA diatoms showed distinct patterns of growth rate under various light intensities. LLA diatoms had their maximum specific growth rates at 12.5 or 25 μmol photons m−2 s−1 but exhibited a decreasing trend above 25 μmol photons m−2 s−1 (Fig. 2a). The maximum specific growth rates of HLA diatoms occurred at 60 or 80 μmol photons m−2 s−1 (Fig. 2b).

Chlorophyll a content

We did not find any significant differences in Chla content in LLA diatoms between LL and HL treatments (Fig. 3a). Chla contents at LL (Chla LL) of S. hantzschii were higher than Chla contents at HL (Chla HL) during the first 3 days. Chla HL exceeded Chla LL and showed a slight increase since day 5. Chla HL and Chla LL in F. crotonensis were similar during the first 5 days but exceeded Chla LL from day 7. Chla HL of N. palea was higher than Chla LL during the whole growth period. No significant differences were found in Chla LL and Chla HL of N. palea. Chla LL fluctuation was found in all HLA diatoms (Fig. 3b).A significant increase in Chla content at LL on day 3 was found in C. meneghiniana(T) and S. acus , (P < 0.01), followed by S. parvus and C. meneghiniana(G) (P < 0.05). Chla LL of C. meneghiniana(T) increased dramatically from day 3 and reached the maximum value on day 7 which was about two times as high as Chla HL in C. meneghiniana(T). Chla contents of S. parvus showed an obvious increase at LL from day 3 to day 7 but only a slight increase at HL. The ratios of Chla content at LL to Chla content at HL of S. acus were close to 2, which lasted from day 3 to day 7, but Chla HL was greater than Chla LL from day 11. Chla LL of C. meneghiniana(G) increased rapidly from day 3 to day 9 but started to decline from day 9.

Fig. 3
figure 3

Change of chlorophyll a content in LLA and HLA diatoms at LL and HL. Data are means ± standard deviations of the Chl a content, n = 3. LL low light intensity (2 μmol photons m−2 s−1), HL high light intensity (80 μmol photons m−2 s−1)

Full size image

Ultrastructure

Light intensity affected cell ultrastructure (Fig. 4). Sectioned chloroplast areas of S. hantzschii and C. meneghiniana(T) were 25 ± 3 % and 34 ± 4 %, respectively, for the ML (Fig. 4b, e). For C. meneghiniana(T), chloroplasts were the dominating organelles, comprising 43 ± 7 % of the sectioned area of cells at LL (Fig. 4d), while there was no distinct difference in sectioned chloroplast areas under HL treatment (32 ± 5 %) (Fig. 4f) compared to ML treatment (Fig. 4e). For S. hantzschii, the sectioned chloroplast areas were 28 ± 4 % and 17.5 ± 3 % under LL and HL treatment, respectively (Fig. 4a, c).

Fig. 4
figure 4

Ultrastructure of S. hantzschii and C. menegihiniana(T). a S. hantzschii cells at LL, b S. hantzschii cells at ML, c S. hantzschii cells at HL, d C. meneghiniana(T) cells at LL, e C. meneghiniana(T) cells at ML, f C. Meneghiniana(T) cells at HL. C chloroplast, N nucleus, M mitochondrion. Scale bar was 0.5 μm in a and d and 1 μm in b, c, e, and f. LL = 2 μmol photons m−2 s−1; ML = 25 μmol photons m−2 s−1; HL = 80 μmol photons m−2 s−1

Full size image

Effective quantum yield of PS II

The ΔF/F’ m values decreased gradually with increasing light levels and was lowest at 80 μmol photons m−2 s−1 in S. hantzschii (Fig. 5a). The maximum ΔF/F’ m value of the LLA diatom S. hantzschii was 0.573 and occurred at 2 μmol photons m−2 s−1 on day 7. The ΔF/F’ m values of S. hantzschii at 12.5 μmol photons m−2 s−1 were close to ΔF/F’ m values at 2 μmol photons m−2 s−1. The ΔF/F’ m values of S. hantzschii at 60 and 80 μmol photons m−2 s−1 decreased gradually from day 9. The ΔF/F’ m values of F. crotonensis increased from day 3 and reached the maximum value at 25 μmol photons m−2 s−1 on day 7, followed by the ΔF/F’ m values at 12.5 and 2 μmol photons m−2 s−1 (Fig. 5b). The ΔF/F’ m of F. crotonensis at 60 and 80 μmol photons m−2 s−1 decreased after day 7. Same as F. crotonensis, S. unla reached the maximum ΔF/F’ m at 25 μmol photons m−2 s−1, followed by 12.5 and 2 μmol photons m−2 s−1 (Fig. 5c). The ΔF/F’ m of N. palea at 12.5 and 25 μmol photons m−2 s−1 increased slowly and reached a maximum on day 13, followed by the ΔF/F’ m at 2, 60, and 80 μmol photons m−2 s−1, which decreased slightly in the later stage (Fig. 5d).

Fig. 5
figure 5

ΔF/F’ m values of eight diatoms under different light intensities during the experiment period. The unit of light intensity is μmol photons m−2 s−1. Data are medians ± standard deviations of ΔF/F’ m values, n = 3

Full size image

The ΔF/F’ m of C. meneghiniana(T) increased from day 3 at all light intensities except 2 μmol photons m−2 s−1 (Fig. 5e). The ΔF/F’ m of C. meneghiniana(T) at 80 μmol photons m−2 s−1 were higher than at the other four light intensities, while at 2 μmol photons m−2 s−1 ΔF/F’ m decreased dramatically from day 7. The maximum ΔF/F’ m of S. parvus (0.612 ± 0.031) was obtained at 60 μmol photons m−2 s−1 on day 11 (Fig. 5f). The ΔF/F’ m of S. parvus at 80 μmol photons m−2 s−1 were a little lower than at 60 μmol photons m−2 s−1, followed by ΔF/F’ m values at 25, 12.5, and 2 μmol photons m−2 s−1. A slight decrease was found in ΔF/F’ m of S. parvus at 2 μmol photons m−2 s−1 from day 9. S. acus had higher ΔF/F’ m values at 60 and 80 μmol photons m−2 s−1 (Fig. 5g). The ΔF/F’ m of S. acus showed a slight decrease at 25 and 12.5 μmol photons m−2 s−1 from day 11 and a sharp decrease at 2 μmol photons m−2 s−1 after day 7. C. meneghiniana(G) had the maximum ΔF/F’ m at 60 μmol photons m−2 s−1 on day 11, followed by 80 μmol photons m−2 s−1 (Fig. 5h). The decrease in ΔF/F’ m of C. meneghiniana(G) began from day 13 at 12.5 μmol photons m−2 s−1 and day 9 at 2 μmol photons m−2 s−1.

The ΔF/F’ m of S. hantzschii reached the maximum values at 2 μmol photons m−2 s−1, and the ΔF/F’ m of the other 3 LLA diatoms slightly decreased, while that of all four HLA diatoms decreased markedly. At 80 μmol photons m−2 s−1, the HLA diatoms reached relatively high ΔF/F’ m, while the LLA diatoms had the lowest ΔF/F’ m.

Non-photochemical quenching

Changes of NPQ value in LLA diatoms are shown in Fig. 6a. The NPQ of S. hantzschii at HL increased from day 3 and reached a maximum on day 7. A small fluctuation around the original value was found in NPQ at LL. NPQ of F. crotonensis at HL increased slightly to a maximum on day 9, while that at LL showed a slight decreasing tendency. NPQ of S. ulna and N. palea at HL increased from day 3 and decreased below NPQ values at LL on day 9 and day 11, respectively. The NPQ of HLA diatoms exhibited another trend (Fig. 6b). NPQ of C. meneghiniana(T) at HL was lower than NPQ at LL, and the ratio of NPQ at HL to NPQ at LL (NPQHL/NPQLL) was in the range of 0.516 ~ 0.774. NPQ of S. parvus at HL decreased, and the NPQHL/NPQLL ratios were lower than 0.5 from day 5. NPQ of S. acus at HL dropped in the initial stage and recovered from day 11. No obvious differences were found in NPQ of C. meneghiniana(G) between LL and HL, and the NPQHL/NPQLL was close to 1.

Fig. 6
figure 6

The NPQ values in LLA and HLA diatoms during the experiment period. Data are means ± standard deviations of NPQ values, n = 3

Full size image

Rapid light curves

Photosynthetic parameters photon-capturing efficiency (α), maximum electron transport rate (rETR max), and light saturation point (I k) in lotic and lentic diatoms were compared under low light intensity (LL, 2 μmol photons m−2 s−1) and high light intensity (HL, 80 μmol photons m−2 s−1). Because intense responses from diatom regulation of Chla and NPQ were almost found on day 5, we sampled on day 5 and day 13 to compare (Table 2). The α of S. hantzschii was the highest one in the LLA diatoms (0.293 ± 0.033 μmol photons m−2 s−1) at LL on day 5, followed by N. palea (0.2815 ± 0.017 μmol photons m−2 s−1), F. crotonensis (0.266 ± 0.021 μmol photons m−2 s−1), and S. ulna (0.227 ± 0.026 μmol photons m−2 s−1). The α values of LLA diatoms at HL were significantly lower than that at LL on day 5. On day 13, the α values of LLA at LL decreased slightly, while the α values of LLA at HL recovered. The rETR max values of S. hantzschii at HL were much higher than those at LL day 5, but decreased to near rETR max value at LL on day 13. A similar phenomenon was found in the other three LLA diatoms. The highest I k value of LLA diatoms at LL was observed in N. palea (55.3 ± 4.218 μmol photons m−2 s−1). The highest I k values of the LLA diatoms at HL were significantly higher than those at LL, but dropped on day 13 (P < 0.05).

Table 2 Photosynthetic parameters in eight diatoms under low and high light intensities. Data are means ± standard deviations, n = 3
Full size table

On day 5, the α values of S. parvus at LL were significantly higher than at HL, followed by C. meneghiniana(G), C. meneghiniana(T), and S. acus. On day 13, the α values of HLA diatoms at LL decreased to near the α values of the HLA diatoms at HL. The rETR max of HLA diatoms at LL were much lower than at HL on day 5, and a small fluctuation was found on day 13. I k values of HLA diatoms at LL were lower than those at HL on day 5, but increased on day 13.

Discussion

Growth characteristics of LLA and HLA diatoms under different light intensities

Growth rates of LLA diatoms increased with increasing light intensities in a lower range of light intensities (0 ~ 25 μmol photons m−2 s−1), but decreased above 25 μmol photons m−2 s−1, indicating that LLA diatoms were low light adapted. Growth rates of HLA diatoms increased with increasing light intensities (0 to 80 μmol photons m−2 s−1) without obvious photoinhibition, and the optimum irradiance for HLA diatoms was much higher than that for LLA diatoms. We assume that HLA diatoms in this study could adapt to much higher light intensities in the field than in the experimental conditions because the light saturation point for growth of HLA diatoms was not reached. The high growth rate of S. hantzschii at LL illustrated the superior ability of this species to adapt to low light. A similar pattern was found by Kolmakov et al. (2002). However, a high irradiance of 13 MJ m−2 was found when S. hantzschii was the dominant diatom in the lower Nakdong River (Kim et al. 2007). Some studies show that F. crotonensis was well adapted to a low light environment (e.g., Hartig and Wallen 1986; Harris 1978), and the optimum light intensity to maintain a high growth rate for F. crotonensis was 27.6 μmol photons m−2 s−1 (Rhee and Gotham 1981). However, other studies have found that when F. crotonensis was dominant in Lake Erie (USA) and Lake Aydat (France), the light intensity reached up to 95 μmol photons m−2 s−1 (Hartig and Wallen 1986) and 150 μmol photons m−2 s−1 (Dauta et al. 1990), respectively. The optimum light intensity for S. acus was in the range of 200 ~ 300 μmol photons m−2 s−1 in Lake Melangsee (Schmitt and Nixdorf 1999), but was 63.5 μmol photons m−2 s−1 in River Spree (Germany) (Köhler 1994). Thus, light-adaptive characteristics associated with local light intensities in the habitat where algae are growing might be important in addition to species-specific characteristics. For example, Kropuenske et al. (2009) found that Fragilariopsis cylindrus isolated from a low light regime exhibited low growth rates at higher light intensities from 75 to 125 μmol photons m−2 s−1. HLA diatoms (Skeletonema costatum, Thalassiosira weissflogii, and Chaetoceros curvisetus) exhibited significantly lower growth rates under low-light treatment than under high-light treatment in a laboratory experiment (Wang et al. 2013).

Habitat-specific differences in adaptation to light in diatoms lead to differences in light intensity for the formation of diatom blooms in low-light ecosystems and high-light ecosystems. Summarizing the results from previous surveys (Fig. 7) (e.g.,Gikuma-Njuru et al. 2013; Jaeger et al. 2008; Pannard et al. 2007; Rothenberger et al. 2009), it clearly indicated that there are significant differences in light conditions between high-light ecosystems and low-light ecosystems when diatoms dominate in the water column. Aquatic ecosystems with annual mean light intensity (at a 0.5-m depth) below 300 μmol photons m−2 s−1 are defined as low-light ecosystems, while above 300 μmol photons m−2 s−1, they are defined as high-light ecosystems. When diatom densities were greater than 2 × 107 cells L−1, the light intensity was in the range of 27.7 ~ 230.2 μmol photons m−2 s−1 in low-light ecosystems, while in high-light ecosystems, it was in the range of 142.0 ~ 920.4 μmol photons m−2 s−1. It suggested that the ideal light intensity for diatom dominance in high-light ecosystems is considerably higher than in low-light ecosystems.

Fig. 7
figure 7

The relationships between diatom densities and light intensities in the high-light and low-light ecosystems. Light intensity was the light intensity at a 0.5-m depth in water column. Circles and triangles represent data collected from low-light ecosystems and high-light ecosystems, respectively

Full size image

Photoacclimation properties of LLA and HLA diatoms

The effective quantum yield (ΔF/F’ m) is closely related to the CO2 assimilation rate that can affect the capacity to convert solar energy into organic material, which consequently impacts on the growth of algae (Edwards and Baker 1993). The photosynthetic rate (mg C [mg Chla]−1 h−1) is proportional to the effective quantum yield (Hartig et al. 1998). In our study, although the eight diatoms were isolated from field sites in different years, the growth rates and photosynthetic parameters of LLA and HLA diatoms exhibited distinct habitat-specific traits. The ΔF/F’ m values of the diatoms showed a similar trend as their growth rates. The higher ΔF/F’ m values of LLA diatoms were accompanied by higher growth rates at light intensities of 12.5 and 25 μmol photons m−2 s−1. Under low light conditions, the ΔF/F’ m values of S. hantzschii were higher than for the other 3 LLA diatoms, and it was consisted with the change in growth rate. The obvious decrease of ΔF/F’ m values in F. crotonensis, S. ulna, and N. palea during the latter period at 60 and 80 μmol photons m−2 s−1 reflected the decreasing growth rates in these three LLA diatoms. This was consistent with other LLA species which often exhibit a low light-saturated growth rate and high photosynthetic efficiency (Gameiro et al. 2011). Low light saturation and high photosynthetic efficiency suggest that S. hantzschii developed a high tolerance of low light stress, and this physiological characteristic could help S. hantzschii outcompete other species in the Hanjiang River where the high tolerance of low light could help it to achieve superiority over other species to form blooms in early spring.

The highest ΔF/F’ m values of C. meneghiniana(T) at 80 μmol photons m−2 s−1 and that of C. meneghiniana(G), S. parvus, and S. acus at 60 μmol photons m−2 s−1 correspond to the high growth rates in the HLA diatoms. Different degrees of decline of ΔF/F’ m values in the four HLA diatoms at LL can explain the low average growth rate of HLA diatoms at LL.

Chla regulation

Light history is considered as a determining factor for the kinetics of pigment variations (Anning et al. 2000). Previous studies have demonstrated that diatoms can regulate their Chla content with light intensity changes (e.g., Migné et al. 2004; Laviale et al. 2009). Chloroplast changes induced by light variations will affect light utilization (Raven and Geider 2003). In the present study, the large distribution area of C. meneghiniana(T) on day 3 cell exhibited an active response to LL. The obvious decrease of ΔF/F’ m values in the four HLA diatoms at LL which occurred after day 9 illustrated that Chla regulation can help HLA diatoms respond to LL stress effectively. The cell density of HLA at LL on day 3 decreased sharply but recovered quickly from day 3 to 9. The growth curve of HLA diatoms at LL became smooth after day 9 probably due to reduced Chla regulation after day 9. In summary, LL can cause and increase in Chla content and chloroplast volume and a simultaneous shart decrease in the average specific growth rate in HLA diatoms. At the early growth stage, the dramatic increase in Chla content at LL leads to the increase in the α values indicating an enhanced photon-capturing efficiency in HLA diatoms. At the latter growth stage, the Chla content of HLA diatoms at LL decreased sharply together with a decrease in α decrease and a slight decrease in rETR max and I k .

NPQ regulation

The xanthophyll cycle (XC) dependent non-photochemical quenching (NPQ) is an essential protective mechanism for protecting algae against excess light, which can help dissipate excess energy through thermal radiation (e.g., Krause 1988; Lavaud et al. 2004). The decrease in ΔF/F’ m of LLA diatoms at HL that began on day 7 or day 9 indicated that NPQ regulation can delay the decline in ΔF/F’ m although the growth rate and photosynthesis activity in LLA diatoms at 80 μmol photons m−2 s−1 was much lower than that at 12.5 and 25 μmol photons m−2 s−1. In addition, low growth rates of LLA diatoms in HL might be due to the large energetic cost of the xanthophyll cycle. At the early growth stage of HL treatments, LLA diatoms showed similar characteristics as HLA diatoms with low α and high rETR max and I k . However, at the latter growth stage when NPQ regulation became weak, the parameters of photosynthesis activity in LLA diatoms decreased. The low Chla content of LLA diatoms at HL before day 7 (except S. unla) indicated that Chla regulation also played a regulation role in LLA.

Kashino et al. (2002) pointed that phytoplankton growing on the surface of water column respond to high irradiance more actively than that from a 25 m depth with obviously high NPQ values. High light-adapted species can have a high pool of xanthophyll-cycling pigments per cell compared to low light-adapted species (Dimier et al. 2009). C. meneghiniana has a large and flexible NPQ potential because the regulatory components of C. meneghiniana (a fast transient component, a steady-state component, and a fast relaxing component) can help C. meneghiniana respond to high light fast and effectively (Grouneva et al. 2008; Su et al. 2012). However, there was no significant increase in NPQ values in HLA diatoms under high light, suggesting that the light intensities used were much lower than their saturated value. The higher NPQ values of HLA diatoms observed at LL may relate to the increase in diadinoxanthin (DD). High light-acclimated algae also show high NPQ values in the dark-adapted samples due to a larger DD signal (Serodio et al. 2005). Similar to the HLA diatoms at LL, the NPQ values of LLA diatoms at LL fluctuated around the initial NPQ values on day 1.

The results of the present study confirmed our hypothesis that diatoms had excellent light-adaptive capability to local light conditions and physiological strategies to manage light fluctuation in the short term. The results show that we should consider the photoacclimation of diatoms to their historic light environments when we study the characteristics of certain species and we can potentially use some convenient light control measures to control the dominance of some species in small eutrophic aquatic ecosystems.