www.elsevier.com/locate/jphotobiol

Journal of Photochemistry and Photobiology B: Biology 82 (2006) 161–172

Effects of nitrogen source and UV radiation on the growth,chlorophyll fluorescence and fatty acid composition of Phaeodactylum

tricornutum and Chaetoceros muelleri (Bacillariophyceae)

Ying Liang *, John Beardall, Philip Heraud

School of Biological Sciences, Monash University, Clayton, Vic. 3800, Australia

Received 15 June 2005; received in revised form 23 October 2005; accepted 2 November 2005Available online 4 January 2006

Abstract

Cultures of the marine diatoms Phaeodactylum tricornutum and Chaetoceros muelleri were grown in f/2 medium supplied with eithernitrate (N-Nt), ammonium (N-Am) or urea (N-Ur) as the nitrogen (N) source at the same final N concentration (0.88 mM). Exponentialgrowth phase cultures of the two diatoms were exposed to four different light regimes for 2 days: (UVAR) PAR (60 lmol quantam�2 s�1) plus 8.22 W m�2 (unweighted) UVAR; (high UVBR) PAR (60 lmol quanta m�2 s�1) plus 1.04 W m�2 (unweighted) UVBRplus 13.73 W m�2 (unweighted) UVAR; (low UVBR) PAR (60 lmol quanta m�2 s�1) plus 0.19 W m�2 (unweighted) UVBR plus2.76 W m�2 (unweighted) UVAR and (PAR) PAR (60 lmol quanta m�2 s�1) alone (control). No significant effects of N source onthe growth rates of the two diatoms were detected. The maximum effective quantum yield of PSII, UPSIIe-max, and the initial slopeof the light curve, a, of P. tricornutum and C. muelleri were all inhibited, whereas Ik was somewhat increased, as a consequence of 2 daysof exposure to all the UVR treatments. Multiple factor ANOVA revealed that all the major fatty acids, both in P. tricornutum andC. muelleri, were influenced more strongly by N source than by UVR. The composition of saturated fatty acids (SFA), monounsaturatedfatty acids (MUFA) and polyunsaturated fatty acids (PUFA) in P. tricornutum and C. muelleri exhibited almost the same pattern ofvariation with N source and UVR. The maximum value of SFA was found in the N-Am treatment, that of MUFA in the N-Nt treatmentand for PUFA in the N-Ur treatment irrespective of the UV radiation. On the other hand, the impact of UVR resulted in an increase ofPUFA and a reduction of SFA both in P. tricornutum and C. muelleri under all N sources.� 2005 Elsevier B.V. All rights reserved.

Keywords: Phaeodactylum tricornutum; Chaetoceros muelleri; Nitrogen source; Ultraviolet radiation; Chlorophyll fluorescence; Fatty acids

1. Introduction

The number and diversity of marine and estuarine ani-mals being reared commercially for human consumptionand other purposes increases each year. Microalgae havean important role in mariculture as food for larval stagesof crustaceans and fish, for all stages of bivalves and asfood for zooplankton, which are fed to late larval and

1011-1344/$ - see front matter � 2005 Elsevier B.V. All rights reserved.

doi:10.1016/j.jphotobiol.2005.11.002

* Corresponding author. Address: The Key Laboratory of Mariculture,Ministry of Education, Ocean University of China, 5 Yushan Road,Qingdao 266003, PR China. Tel.: +86 532 82032273; fax: +86 53282894024.

E-mail address: yliang@ouc.edu.cn (Y. Liang).

juvenile fish and crustaceans. Whether microalgae couldmeet the needs of these animals depends on their size,digestibility and, especially, their nutritional value, whichis mainly related to the content of essential fatty acids[1–3]. The fatty acid profiles of a given algal strain isnot only determined by its heredity, but also can be mod-ified by different growth conditions [4–8]. Among the dif-ferent components of the culture medium, the nitrogen(N) source and concentration can provoke importantchanges in the biochemical composition of microalgal spe-cies. However, comparatively few studies have analysedthe effects of different N sources on the fatty acid compo-sition of microalgae [4,7–9] and those that have been pub-lished are partly contradictory even in the same species of

162 Y. Liang et al. / Journal of Photochemistry and Photobiology B: Biology 82 (2006) 161–172

microalgae. The results of Lourenco et al. [8] indicatedthat seven species, including Phaeodactylum tricornutum

Bohlin and Isochrysis galbana Parke, seemed to accumu-late less PUFA (polyunsaturated fatty acids) in the cul-tures with urea as N source. In contrast, Dunaliella

tertiolecta Butcher exhibited remarkably high PUFA per-centages in cultures with urea as N source [8]. Similarresults have been obtained in P. tricornutum [4,9] and inI. galbana [7]. However, there have been no reports onthe varying effects of N source on the fatty acid composi-tion of Chaetoceros muelleri Lemmermann.

Anthropogenic depletion of the stratospheric ozonelayer has caused a marked increase in UVBR (280–315 nm) reaching the earth�s surface [10–13]. Such increasesare more significant in polar regions, though significantrises are also occurring at lower latitudes. For instance,ozone depletion in the Northern Hemisphere has led to sig-nificant increases in UVBR as far south as France andnorthern Italy and this has extended over parts of theNorth Atlantic [14,15]. There is thus growing concernabout the potential negative impact of UVBR on aquaticecosystems. Both UVAR and UVBR are known to havedeleterious effects on growth and photosynthetic capabilityof phytoplankton organisms. UVBR is particularly harm-ful and has been shown to be genetically, physiologically,and photosynthetically detrimental to many forms of lifein terrestrial and aquatic ecosystems [16–19].

UVBR has been shown to have significant influence on arange of biochemical processes within microalgae includingnutrient uptake, photosynthesis and cell division [20–23].Several studies have reported that UVR inhibits the maxi-mum effective quantum yield of PSII (UPSIIe-max, =Fv/Fm), in microalgae [24–28]. However, there are few pub-lished papers pertaining to the effects of UVR on the fattyacid composition of microalgae [2,21,22,29,30], and thosethat have been published are seemingly contradictory.The extant literature on the subject includes: two studiesreporting an overall increase in SFA (saturated fatty acids)and MUFA (monounsaturated fatty acids) and decrease inPUFA upon UVR [2,30]; two papers reporting no signifi-cant differences in the fatty acid profiles between UVBRtreatments and the control [21,23], and three reporting anincrease in PUFA after UVR exposure [23,29,31]. The dia-toms P. tricornutum and C. muelleri are two commonlyused marine unicellular algae in many mariculture systems[8,32]. They are both rich in PUFAs, and grow well inlarge-scale cultures, either indoors or outdoors[4,8,33,34]. However, to our knowledge, there have beenno reports on the varying effects of UVR on the fatty acidcomposition of P. tricornutum and C. muelleri culturedunder different N sources.

In this study we addressed this deficiency by assessingthe effects of three different N sources (nitrate, ammoniumand urea) on the growth, chlorophyll fluorescence and fattyacid composition of the two marine diatoms, P. tricornu-tum and C. muelleri, with and without exposure to UVRfor 2 days during the mid-exponential growth phase.

2. Materials and methods

2.1. Microalgal cultures

Microalgal cultures of P. tricornutum and C. muelleri

were obtained from the CSIRO culture collection, Hobart,Australia and are designated in the CSIRO catalogue ascultures CS-29C and CS-176, respectively. The culturesused were uni-algal but not axenic, although bacterial num-bers were minimal at all stages. Cultures were grown insterilized natural seawater enriched with f/2 medium [35]and supplied with nitrate (N-Nt), ammonium (N-Am) orurea (N-Ur) as nitrogen (N) source at the same final N con-centration (0.88 mM). Cultures were grown in 3 L glassflasks containing 2 L culture medium at 18 ± 1 �C. Aera-tion of the cultures was achieved via a constant stream ofair, passed through a 0.2 lm cellulose acetate filter at�1 L min�1. Cool white fluorescent lights provided contin-uous photosynthetically active radiation (PAR) with a pho-ton flux of 60 lmol quanta m�2 s�1 with no UVR. Celldensities were determined everyday using an improvedNeubauer haemocytometer. For each culture, the maxi-mum specific growth rate (expressed as divisions d�1) wascalculated from the slope of the linear portion of a curveshowing log cell number as a function of time.

When cultures reached mid-exponential growth phase,each parental culture was thoroughly mixed and 250 mlWhirl-Pac bags 11 · 23 cm (pre-sterilized, low-densitypolythene bags, transparent to UVR) were established fordifferent light treatments. All subcultures were grown at18 ± 1 �C with a 16:08 light–dark cycle and exposed to fourdifferent light regimes for 2 days: (UVAR) PAR (60 lmolquanta m�2 s�1) plus 8.22 W m�2 (unweighted) UVAR;(high UVBR) PAR (60 lmol quanta m�2 s�1) plus1.04 W m�2 (unweighted) UVBR plus 13.73 W m�2

(unweighted) UVAR; (low UVBR) PAR (60 lmol quantam�2 s�1) plus 0.19 W m�2 (unweighted) UVBR plus2.76 W m�2 (unweighted) UVAR (low UVBR treatment)and (PAR) PAR (60 lmol quanta m�2 s�1) alone (control).The subcultures were not aerated during the 2 days UVRexposure.

Following irradiation, each culture was well mixed and a10 ml subsample was obtained after 1 or 2 days UVR expo-sure for cell density and chlorophyll fluorescence determi-nation. The remainder of each experimental culture wasfiltered through 47 mm diameter Whatman GF/C filtersfor subsequent fatty acid analysis.

2.2. UVR and control treatments

PAR was provided by a fluorescent tube (Phillips TLD36W/950, Holland) and UVR was provided by a singleQ-Panel fluorescent tube (UVA-340, Q-PANEL CO.USA). All tubes were installed in a batten luminaire sus-pended so that the fronts of the fluorescent tubes were10 cm directly in front of the Whirl-Pac bags. Differentspectral environments were provided by covering the tubes

Fig. 1. Final cell density of Phaeodactylum tricornutum (a) and Chaetoc-

eros muelleri (b) cultured at different N source (N-Nt, N-Am and N-Ur)and exposed to different UVR treatments (UVAR, high/low UVBR, PAR)for 2 days. Error bars are standard deviations. UVAR: PAR (60 lmolquanta m�2 s�1) plus 8.22 W m�2 (unweighted) UVAR; high UVBR: PAR(60 lmol quanta m�2 s�1) plus 1.04 W m�2 (unweighted) UVBR plus13.73 W m�2 (unweighted) UVAR; low UVBR: PAR (60 lmol quantam�2 s�1) plus 0.19 W m�2 (unweighted) UVBR plus 2.76 W m�2

(unweighted) UVAR; PAR: PAR (60 lmol quanta m�2 s�1) alone(control).

Y. Liang et al. / Journal of Photochemistry and Photobiology B: Biology 82 (2006) 161–172 163

with either new cellulose acetate film (Kodacel; acetate filmbase, Kodak Eastman Company, New York; opaque towavelengths <280 nm) constituting the UVBR treatment,Mylar film (Du Pont type S, 375 lm thick, opaque to wave-lengths <315 nm) constituting the UVAR treatment, andUltraphan 395 film (Digefra, Munich, Germany, opaqueto wavelengths <395 nm) constituting the control. Differenttreatments were separated by specially designed darkenedplexiglass plates (opaque to wavelengths <400 nm) to pre-vent stray light reaching other Whirl-Pac bags. The UVBRirradiance was varied by attenuation with one perforatednickel screen to provide low or high UVBR.

2.3. Measurement of UVR and PAR

The spectral power distribution of the ultraviolet lightsource used in this study was measured using a spectrora-diometer system comprising an Acton Research Corpora-tion (Massachusetts, USA) Model 300i, 300 mm focallength monochromator, and an Acton Research Corpora-tion model PD-439 photomultiplier tube detector, withlight brought to the monochromator via a 3 m fibre opticincorporating a quartz glass diffuser element (ActonResearch Corporation). Data was acquired using an ActonResearch Corporation NCL model data acquisition mod-ule and SpectraSense software.

The spectroradiometer was calibrated against a 1000 Wtungsten-halogen standard lamp (Oriel Corporation, Con-necticut, USA), traceable to the National Institute of Stan-dards, USA. Wavelength calibration was achieved usingemission lines from a mercury lamp. Table 1 shows theunweighted UVBR and UVAR fluence-rates for theUVR treatments, and the irradiance weighted by six pub-lished biological weighting functions.

PAR fluxes were measured using a LiCor (Lincoln,Nebraska,USA)LI-188B IntegratingQuantumRadiometer-Photometer.

2.4. Fluorescence measurements

Chlorophyll fluorescence was measured using a pulseamplitude modulated fluorometer (Phyto-PAM, Walz,

Table 1UVBR fluence-rates in W m�2, at a distance of 10 cm from the fluorescent tucombined with one perforated nickel screen (filter), or cellulose acetate combinweighting functions to give biologically effective UVBR fluence rates

Cellulose acetate

Unweighted UVBR (W m�2) 1.04Weighted UVBRJones and Kok [37] 0.9Behrenfeld et al. [38] 0.6Setlow [39] 0.6Cullen et al. [40] 0.4Caldwell et al. [41] 0.6Quaite et al. [42] 0.5Unweighted UVAR (W m�2) 13.73

Weightings are relative and normalized at 300 nm except for Cullen et al. [40] wUVAR fluence-rates are also stated.

Effeltrich, Germany). Data were recorded using PhytoWinsoftware (version 1.45, Walz) on a PC. Before measure-ment, both the control and the UV-exposed samples were

be source, filtered with either cellulose acetate alone, or cellulose acetateed with Mylar film, unweighted and weighted by six published biological

Cellulose acetate + 1 filter Cellulose acetate + Mylar

0.19 0

0.16 00.11 00.11 00.07 00.11 00.09 02.76 8.22

hich is a dimensionless biologically weighted irradiance, E�inh. Unweighted

164 Y. Liang et al. / Journal of Photochemistry and Photobiology B: Biology 82 (2006) 161–172

dark adapted for 15 min. The processing software gener-ates values for minimum (Fo) and maximum (Fm) fluores-cence values from which variable fluorescence (Fv) wascalculated as Fm � Fo. The maximum effective quantumyield of PSII, UPSIIe-max, is then calculated as Fv/Fm. Inaddition, 11 PAR irradiances, from dark to 550 lmolquanta m�2 s�1 were applied in increasing order, with aduration of 60 s at each level, to produce the rapid lightcurve. The initial slope of the rapid light curve (rETR/PAR) is defined as a (alpha). rETRmax corresponds tothe apparent saturated electron transport rate (uncorrected

Fig. 2. Mean values (n = 4) and standard deviations of fluorescence parameter(lmol electron m�2 s�1) and (j–l) Ik (lmol quanta m�2 s�1) in Phaeodactylum

exposed to different UVR treatments (UVAR, high/low UVBR, PAR) for 1 dconditions to those described in Fig. 1.

for the cellular absorptance) and Ik represents the particu-lar PAR-value at the intersection of the lines defined by theinitial slope and rETRmax (parallel to PAR-axis). All mea-surements occurred in a darkened room.

2.5. Fatty acid analysis

For fatty acid analysis, samples of the two diatoms wereharvested after 2 days of UVR exposure. Cells were filteredonto 47 mm diameter GF/C Whatman Glass MicrofibreFilters. Freeze-dried biomass was used for fatty acid analysis

s, (a–c) UPSIIe-max, (d–f) a (lmol electrons/lmol photons), (g–i) rETRmax

tricornutum cultured at different N source (N-Nt, N-Am and N-Ur) anday (Day 1) and 2 days (Day 2). The UVR treatments represent identical

Table 2Summary of MANOVA (a) and SNK multiple comparison test results (b) on the fluorescence parameters of Phaeodactylum tricornutum exposed to UVRunder different N sourceA

Days of exposure Day 1 Day 2

Source of variance UPSIIe-max a rETRmax Ik UPSIIe-max a rETRmax Ik

(a) N source F 21.762 5.143 15.328 16.721 110.916 162.50 62.293 96.423UVR 699.222 242.810 21.725 9.407 229.011 285.50 54.197 53.557N source · UVR 24.746 14.667 9.610 1.474 37.003 43.000 32.540 5.069N source P-level <0.001 0.024 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001UVR <0.001 <0.001 <0.001 0.002 <0.001 <0.001 <0.001 <0.001N source · UVR <0.001 <0.001 0.001 0.267 <0.001 <0.001 <0.001 0.008

(b) N source UraAmbNtb UraAmbNtb UraAmbNtb UraAmbNtb NtaAmbUrc NtaAmbUrc AmaNtbUrc AmaUrbNtc

UVR CCaB2bAcB1d CaB2bAcB1d CaB1bB2bAb B1aAaB2bCb CaB2bAbB1b CaB2bAbB1b CaB1bB2bAb B1aAbB2bCc

N source · UVR DNt CaB2bAcB1c CaB2bAcB1c CaB1aB2aAa BNS CaB2bAbB1b CaB2bAbB1b CaB1aB2aAa B1aAaB2aCb

Am CaB2bAcB1c CaB2bB1bcAc CaB1bB2bAb CaB2bAbB1b CaB2bAbB1b CaB1bB2cAc B1aAbB2bCc

Ur CaB2bAcB1d CaB2bAcB1d CaB1bB2bAb CaB2bAbB1b CaB2bAbB1b CaB2bAbB1b B1aAbB2bcCc

A Student-Newman-Keul (SNK) multiple comparison test results are arranged in decreasing order from left to right: a > b > c > d. Group with theidentical superscript letters indicate that mean values are not significantly different. Significant probabilities are in bold (P < 0.05).B NS: no significant.C C: PAR; B1: high UVBR; B2: low UVBR; A: UVAR.D Nt: NaNO3; Am: NH4Cl; Ur: NH2CONH2.

Y. Liang et al. / Journal of Photochemistry and Photobiology B: Biology 82 (2006) 161–172 165

by gas chromatography, with prior direct transesterificationas described by Lepage and Roy [36].

The fatty acid methyl esters obtained by transesterifica-tion were analysed on a Hewlett–Packard HP6890 gaschromatograph provided with a fused silica capillary col-umn (J&W Scientific, DB-23, 0.25 mm film thickness,30 m · 0.25 mm i.d.) and a HP 5973 mass selection detectorfitted with a direct capillary inlet. It was equipped with asplit/splitless injector which was used in splitless mode.The samples were injected using an autosampler and a1 ll injection volume. Ultra high purity helium was usedas the carrier gas at a flow rate of 1 ml min�1, with an injec-tor and detector temperature of 280 �C. The oven was pro-grammed from 70 to 150 �C at 30 �C min�1, then rampedto 250 �C at 5 �C min�1 and held at 250 �C until all peakshad appeared. Peak retention times and areas are calcu-lated by integration of areas under the peaks. Fatty acidscomposition is calculated from the area of the peak (%).Mass spectra were acquired and processed with the Hew-lett–Packard workstation. Fatty acid methyl esters wereidentified by comparing the retention times of experimentalsamples with authentic references and by comparison ofthe mass spectra with the spectra present in the Hewlett–Packard NBS75K Library.

2.6. Statistical analysis

The results of N source on the maximum specific growthrates of P. tricornutum and C. muelleri were analysedby one-way analysis of variance (ANOVA). Student-Newman-Keul (SNK) multiple comparison test were usedto test the differences between treatment groups. Differenceswere considered significant at a probability level of 0.05.

Two factor ANOVA was used to examine the effect ofUVR (UVAR, high/low UVBR, PAR) and N source (N-Nt, N-Am and N-Ur) on the final cell density, fluorescenceparameters (UPSIIe-max, a, rETRmax and Ik) and fatty acid

compositions of P. tricornutum and C. muelleri. UVR andN source were all considered fixed factors. P < 0.05 wasaccepted for significant differences. Post hoc analyses weremade by the Student-Newman-Keul (SNK) test. All statis-tics were performed with SPSS (version 11.5) software.

3. Results

3.1. Growth

Changes in growth of P. tricornutum are illustrated inFig. 1. The maximum specific growth rates under PARonly were 1.38 ± 0.01 d�1 for N-Nt, 1.43 ± 0.01 d�1 forN-Am and 1.34 ± 0.01 d�1 for N-Ur. These differences ingrowth rates were not statistically significant (P > 0.05).Two factor ANOVA revealed that N source but notUVR had a significant influence on the final cell densityof P. tricornutum. The final cell density of P. tricornutumin N-Nt treatment was always higher than those in N-Am and N-Ur treatments whatever the UVR (Fig. 1).

For C. muelleri, the maximum specific growth rates were0.80 ± 0.01 d�1 for N-Nt, 0.75 ± 0.01 d�1 for N-Am and0.76 ± 0.03 d�1 for N-Ur. No significant differences amongthem were detected (P > 0.05). Two factor ANOVArevealed that UVR but not the N source had a significantinfluence on the final cell density of C. muelleri. In N-Nttreatment, the final cell density of C. muelleri was inhibitedby high UVBR when compared with the control; whereasin N-Am and N-Ur treatments, there were no significantdifferences between the four UVR treatments (Fig. 1).

3.2. Chlorophyll fluorescence

The fluorescence parameters UPSIIe-max, a, rETRmax

and Ik of P. tricornutum cultured under different Nsource and exposed to different UVR for 2 days are pre-sented in Fig. 2. Two-way ANOVA (Table 2) revealed

Fig. 3. Mean values (n = 4) and standard deviations of fluorescence parameters, (a–c) UPSIIe-max, (d–f) a (lmol electrons/lmol photons), (g–i) rETRmax

(lmol electron m�2 s�1) and (j–l) Ik (lmol quanta m�2 s�1) in Chaetoceros muelleri cultured at different N source (N-Nt, N-Am and N-Ur) and exposed todifferent UVR (UVAR, high/low UVBR, PAR) for 1 day (Day 1) and 2 days (Day 2). The UVR treatments represent identical conditions to thosedescribed in Fig. 1.

166 Y. Liang et al. / Journal of Photochemistry and Photobiology B: Biology 82 (2006) 161–172

that UPSIIe-max and a in P. tricornutum were influencedstrongly by UVR, with a much smaller influence of Nsource and their interaction N source · UVR. The valuesof UPSIIe-max and a in the control were always higherthan those in the three UVR treatments, whatever theN source and days of UVR exposure. For N source,the higher values of UPSIIe-max and a were recorded inN-Ur treatment on day 1 and in N-Nt treatmenton day 2. P. tricornutum displayed a significant effectof N source and UVR on rETRmax and Ik. The rETRmax

values in the control were higher than those in the three

UVR treatments whereas the lower Ik values alwaysoccurred in the control compared to the three UVRtreatments. For N source, the higher values of rETRmax

and Ik were occurred in N-Ur treatment on day 1 and inN-Am treatment on day 2 (Fig. 2 and Table 2).

The fluorescence parameters UPSIIe-max, a, rETRmax

and Ik of C. muelleri varied in the interaction with Nsource and UVR (Fig. 3 and Table 3). Two factorANOVA revealed that N source had a significant influ-ence on all the parameters. Their values in N-Ur treatmentwere always higher than those in N-Am and/or N-Nt

Table 3Summary of MANOVA (a) and SNK multiple comparison test results (b) on the fluorescence parameters of Chaetoceros muelleri exposed to UVR underdifferent N sourceA

Days of exposure Day 1 Day 2

Source of variance UPSIIe-max a rETRmax Ik UPSIIe-max a rETRmax Ik

(a) N source F 31.500 22.200 14.185 5.691 197.615 81.750 165.146 115.554UVR 227.048 115.933 3.058 11.692 368.077 185.667 6.356 64.196N source ·UVR 1.976 1.933 2.088 1.846 4.077 2.417 4.425 7.479N source P-level <0.001 <0.001 0.001 0.018 <0.001 <0.001 <0.001 <0.001UVR <0.001 <0.001 0.070 0.001 <0.001 <0.001 0.064 <0.001N source ·UVR 0.148 0.156 0.131 0.172 0.018 0.091 0.072 0.002

(b) N source UraAmbNtb UraAmbNtb UraAmbNtc UraAmaNtb UraNtaAmb UraNtaAmb UraAmbNtb UraAmbNtc

UVR CaB2bAcB1d CaB2bAcB1d NS B1aAaB2aCb CaB2bA cB1d CaB2bAcB1d NS B1aAbB2cCd

N source ·UVR Nt BNS NS NS NS CaB2bAbB1c NS NS B1aAbB2bCb

Am CaB2bAcB1d B1aAaB2bCc

Ur CaB2bAcB1d B1aAaB2aCa

Note. The letters (C, B1, B2 and A) and (Nt, Am and Ur) represent identical conditions to those described in Table 2.A Student-Newman-Keul (SNK) multiple comparison test results are arranged in decreasing order from left to right: a > b > c > d. Group with theidentical superscript letters indicate that mean values are not significantly different. Significant probabilities are in bold (P < 0.05).B NS: no significant.

Y. Liang et al. / Journal of Photochemistry and Photobiology B: Biology 82 (2006) 161–172 167

treatments. For UVR, UPSIIe-max and a were higher in thecontrol when compared with the three UVR treatments,whatever the N source and days of UVR exposure. Therewas no significant influence of UVR on rETRmax of C.

muelleri (P > 0.05). The values of Ik after UVR exposurehad the general trend that the lower values were recordedin the control compared with the three UVR treatments(Fig. 3 and Table 3).

3.3. Fatty acids

The fatty acid profiles of P. tricornutum cultured underdifferent N sources and exposed to different UVR treat-ments are given in Table 4. For clarity, only the six majorfatty acids (14:0, 16:0, 16:1n � 7, 16:3n � 4, EPA andDHA) were considered. Two factor ANOVA (Table 5)revealed that all the major fatty acids in P. tricornutum

were influenced strongly by N source, with a much smallerinfluence of UVR and their interaction N source · UVR;The percentage of 14:0 in P. tricornutum varied in its inter-action with N source and UVR (Table 4), from 5.95% to8.19%. In fact, the 14:0 content was lower in the N-Nttreatment than in the N-Am and N-Ur treatments. P.

tricornutum displayed a significant effect of N source andUVR on 16:1n � 7 content. In particular, the percentageof 16:1n � 7 in the control was always found to be higherthan those observed in the three UVR treatments, exceptin the N-Nt treatment, where there was no significant dif-ference between the four light treatments. 16:0 and SFAin P. tricornutum exhibited almost the same pattern of var-iation with N source and UVR. The values in N-Am treat-ment were higher than those observed in N-Nt and N-Urtreatments. On the other hand, the 16:0 and SFA contentin the control treatment were higher than two or three ofthe UVR treatments (Table 5). The MUFA was influencedstrongly by N source with a much smaller influence ofUVR and no significant influence of their interaction N

source · UVR, with the maximum value occurring in theN-Nt treatment and the minimum in the N-Ur treatment.For UVR, the percentage of MUFA in the control wasalways found to be higher than those observed in the threeUVR treatments. Finally, EPA and PUFA exhibitedalmost the same pattern of variation with N source andUVR. The percentages of these fatty acids in the N-Urtreatment were higher than those observed in the N-Ntand N-Am treatments. On the other hand, EPA and PUFAwere lower in the control treatment when compared withthe three UVR treatments.

The fatty acid compositions of C. muelleri are given inTable 6. For clarity, only the six major fatty acids (14:0,16:0, 16:1n � 7, 20:4n � 6, EPA and DHA) were consid-ered. Two factor ANOVA (Table 7) revealed that N sourcehad a significant influence on all the major fatty acids in C.muelleri. UVR and the interaction between N source andUVR had no significant effects on 16:0, 16:1n � 7 andMUFA. The 16:0, 16:1n � 7 and MUFA content werealways higher in the N-Nt treatment in comparison withthe N-Am and N-Ur treatments. SFA content variedaccording to N source and UVR treatment. In fact, SFAcontent was higher in the N-Am treatment than in the N-Nt and N-Ur treatments. On the other hand, SFA contentwas higher in the control when compared with the threeUVR treatments. The 14:0 content of C. muelleri was sig-nificantly affected by the N source and UVR. The 14:0 con-tent was higher in N-Am treatment than in N-Nt and N-Urtreatments and also variable depending on the UVR in N-Am and N-Ur treatments. C. muelleri displayed a signifi-cant effect of N source and UVR on 20:4n � 6 content.In particular, the mean value of 20:4n � 6 in the control,as expressed by the percentage of total fatty acids, wasoften found to be lower than those observed in the threeUVR treatments with an exception in the N-Nt treatment,where there was no significant difference among the fourlight treatments. Finally, EPA and PUFA were strongly

Tab

le4

Fatty

acid

composition(%

oftotalfattyacid)ofPhaeodactylum

tricornutum

exposedto

UVR

under

differentN

sourcea

Nsource

N-N

tN-A

mN-U

r

UVR

UVAR

HighUVBR

Low

UVBR

PAR

UVAR

HighUVBR

Low

UVBR

PAR

UVAR

HighUVBR

Low

UVBR

PAR

14:0

6.38

(0.10)

6.57

(0.04)

6.18

(0.24)

5.95

(0.10)

7.88

(0.05)

8.17

(0.18)

7.87

(0.09)

8.09

(0.03)

7.69

(0.17)

7.95

(0.15)

7.85

(0.23)

8.19

(0.41)

16:0

13.50(0.16

)14

.32(0.07

)14

.03(0.39

)15

.54(0.35

)17

.37(0.06

)18

.03(0.57

)17

.51(0.64

)18

.26(0.81

)12

.93(0.30

)13

.37(0.33

)13

.61(0.26

)14

.17(0.68

)16

:1n�

727

.51(0.19

)26

.89(0.39

)27

.71(0.75

)27

.23(0.40

)28

.25(0.16

)28

.96(0.25

)28

.53(0.72

)29

.78(0.84

)23

.51(0.35

)23

.60(0.26

)24

.20(0.49

)24

.90(0.95

)16

:3n�

48.01

(0.09)

7.58

(0.08)

7.69

(0.14)

7.10

(0.02)

3.11

(0.06)

3.10

(0.11)

3.29

(0.22)

3.13

(0.23)

7.01

(0.17)

6.75

(0.08)

6.87

(0.17)

6.79

(0.11)

20:5n�

320

.26(0.12

)19

.98(0.23

)19

.51(0.69

)18

.86(0.61

)21

.01(0.22

)20

.84(0.71

)21

.34(0.55

)21

.07(0.44

)23

.42(0.56

)22

.16(0.57

)22

.48(0.54

)21

.18(1.18

)22

:6n�

31.78

(0.05)

1.67

(0.03)

1.63

(0.11)

1.60

(0.11)

1.61

(0.02)

1.52

(0.11)

1.65

(0.07)

1.49

(0.11)

2.52

(0.07)

2.33

(0.15)

2.36

(0.11)

1.99

(0.29)

SFA

24.90(0.34

)25

.98(0.17

)25

.05(0.67

)26

.21(0.17

)31

.46(0.18

)32

.51(1.23

)31

.19(0.81

)31

.63(0.86

)25

.46(0.47

)26

.17(0.48

)26

.21(0.28

)26

.66(0.52

)MUFA

34.93(0.24

)34

.05(0.44

)35

.52(0.88

)35

.87(0.65

)31

.35(0.32

)32

.10(0.33

)31

.87(0.58

)32

.74(0.65

)25

.69(0.41

)25

.84(0.35

)26

.49(0.51

)27

.10(1.01

)PUFA

40.17(0.28

)39

.97(0.30

)39

.43(1.55

)37

.92(0.68

)37

.19(0.44

)34

.89(1.06

)36

.94(1.38

)35

.63(1.49

)48

.85(0.66

)47

.99(0.37

)47

.30(0.44

)46

.24(1.53

)

Note.Theletters(N

t,Am

andUr)

representidenticalconditionsto

those

described

inTab

le2.

aValues

areav

erag

esoffourmeasurements.Standarddeviationsareshownin

parentheses.

168 Y. Liang et al. / Journal of Photochemistry and Photobiology B: Biology 82 (2006) 161–172

influenced by N source with a smaller but significanteffect by UVR. The content of these fatty acids in theN-Ur treatment was higher than one or both of the N-Ntand N-Am treatments. On the other hand, the EPA andPUFA content were all significantly lower in the controltreatment than in the three UVR treatments (Table 7).

It appears that SFA, MUFA and PUFA in P. tricornu-

tum and C. muelleri exhibited almost the same pattern ofvariation according to N source and UVR. The maximumvalue of SFA was found in the N-Am treatment, that ofMUFA in the N-Nt treatment and for PUFA in theN-Ur treatment irrespective of the UV radiation. On theother hand, the impact of UVR resulted in an increase ofPUFA and a reduction of SFA both in P. tricornutumand C. muelleri.

4. Discussion

The present set of experiments on P. tricornutum andC. muelleri grown under different N sources and exposedto different UVR treatments for 2 days induced markedchanges in chlorophyll fluorescence and fatty acid compo-sition. The study examined the changes that occur in thepresence of artificial sources of UVR and PAR. While wemust point out that the ratio of fluence rates in the PAR,UVAR, UVBR portions of the spectrum in any of thetreatments did not simulate those in natural sunlight, nev-ertheless, these artificial sources are of a kind commonlyused and described in the photobiological literature andtherefore allow direct comparisons to be made betweenthe current study and previous work.

4.1. Growth and chlorophyll fluorescence

In our experiments, nitrate, ammonium and urea were allsuitable N sources for growth of P. tricornutum andC. muelleri. No significant effects of N source on the maxi-mum specific growth rates of the two diatoms were detected(P > 0.05). Our findings are in accordance with the resultsof Fidalgo et al. [7] who reported that I. galbana grew inthe three different N sources at a similar rate, with no signif-icant differences among them. The data in Lourenco et al.[8] also showed the same result and reported that therewas no significant effect of N source on the specific growthrate of 10 marine microalgae including P. tricornutum.

In the present study, 2 days of UVR exposure caused asignificant decline both in UPSIIe-max and a in P. tricornu-

tum and C. muelleri. This result is consistent with the previ-ous literature showing UVR inhibits UPSIIe-max inmicroalgae [24–28]. The marked decrease of UPSIIe-max

under UVAR and UVBR stress shows an inhibiting effectof these wavebands of UVR on photosynthetic efficiency,while the observed decline in a seems correlated to anincrease in the optimal photosynthetic irradiance valueand indicates an overall decrease in the photosyntheticquantum yield of photosynthesis. rETRmax were inhibitedby UVR in P. tricornutum but not in C. muelleri in the

Table 5Summary of MANOVA (a) and SNK multiple comparison test results (b) on the fatty acid composition of Phaeodactylum tricornutum exposed to UVRunder different N sourceA

Source of variance 14:0 16:0 16:1n � 7 20:5n � 3 SFA MUFA PUFA

(a) N source F 458.942 409.075 327.699 79.474 509.853 958.081 584.353UVR 5.278 20.128 6.577 8.186 7.572 12.788 9.818N source ·UVR 5.720 1.860 3.370 3.049 1.725 1.870 2.413N source P-level <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001UVR 0.004 <0.001 0.001 <0.001 <0.001 <0.001 <0.001N source ·UVR <0.001 0.115 0.010 0.016 0.141 0.113 0.046

(b) N source AmaUraNtb AmaNtbUrc AmaNtbUrc UraAmbNtc AmaUrbNtc NtaAmbUrc UraNtbAmc

UVR B1aCbB2bAb CaB1bB2bAc CaB1bB2bAb AaB1aB2aCb CaB1aB2bAb CaB2bB1cAc AaB1aB2abCb

N source ·UVR Nt B1aAabB2bCc BNS CaB1aB2aAa AaB1aB2abCb NS NS AaB1aB2aCb

Am B1aCaB2bAb CaB2abB1bAb AaB1abB2abCb AaB1aB2abCb

Ur B1aCaB2aAa CaB2abB1bAb AaB1 abB2abCb AaB1a B2abCb

Note. The letters (C, B1, B2 and A) and (Nt, Am and Ur) represent identical conditions to those described in Table 2.A Student-Newman-Keul (SNK) multiple comparison test results are arranged in decreasing order from left to right: a > b > c. Group with the identicalsuperscript letters indicate that mean values are not significantly different. Significant probabilities are in bold (P < 0.05).B NS: no significant.

Y. Liang et al. / Journal of Photochemistry and Photobiology B: Biology 82 (2006) 161–172 169

present study (Figs. 2 and 3). The decrease of rETRmax indi-cated inhibition and/or downregulation of photosynthesis.

4.2. Effect of N source and UVR on fatty acid composition

Several studies have reported that UVBR has qualitativeeffects on the distribution of FA in algae [2,21,23,29–31].Our data showed that the impact of UVR resulted in anincrease of PUFA and a reduction of SFA both in P. tri-

cornutum and C. muelleri. These observations suggest thatsuch UVR treatment enhanced PUFA and suppressed SFAformation in the two diatoms, which agrees with the find-ings of De Lange and Van Donk [31] who showed thatthe content of EPA for Cryptomonas pyrenoidifera Geitlerincreased under UVR. Our results also support the viewof Skerratt et al. [23] who reported that relative amountsof PUFA in Phaeocystis antarctica Karsten increased andthe SFA decreased slightly in a low UVBR treatment.The data in Dohler and Biermann [29] also showed thesame trends and these authors reported that the impactof UVBR resulted in an increase of long-chained FA (C-18 and C-20) of Ditylum brightwellwii Grunow, while therewas a reduction of short-chained FA (C-14, C-16). On theother hand, our data are different from the results of Goeset al. [30] (who showed that the formation of PUFA in thegreen algae Tetraselmis sp. was suppressed in the presenceof UVBR) and the data of Skerratt et al. [23] who reportedthat PUFA percentages decreased in Chaetoceros simplex

Ostenfeld under high UVBR. The fatty acid compositionof P. tricornutum after 2 days UVR exposure in our studyalso differs from the results of Wang and Chai [2] in thesame species, which showed that UVBR exposure causesa reduction in the levels of n � 3 fatty acids in P. tricornu-

tum. This difference could be related to growth conditionsbecause the data presented by Wang and Chai [2] wereobtained from cultures grown in open plastic dishes, orthey could result from the different culture medium (10ppt salinity river water enriched with modified f mediumwas used by Wang and Chai [2] whereas we used 35 ppt

salinity natural seawater enriched with f/2 medium), orthe differences could be related to different UVR fluxesand different times of UVR exposure.

Comparatively few studies have analysed the effects ofdifferent N source on the fatty acid composition of mic-roalgae [4,7–9]. Our results showed that P. tricornutum

and C. muelleri exhibited almost the same pattern of var-iation in SFA, MUFA and PUFA according to N source.The maximum value of SFA was found in the N-Amtreatment, for MUFA in the N-Nt treatment and forPUFA in the N-Ur treatment irrespective of the UVRexposure. Our findings are in accordance with resultsfor P. tricornutum [4,9] and I. galbana [7] which showedthat the highest fatty acid contents in the early stationaryphase were produced in cultures grown on urea, but aredifferent from the results of Lourenco et al. [8] whoshowed that P. tricornutum seemed to accumulate lessPUFA in the cultures with urea as the N source. The dif-ference could result from the lower concentration of Nnutrient in the f/2 medium (0.88 mM) we used comparedwith the Conway medium (1.18 mM) used by Lourencoet al. [8].

In conclusion, the biochemical properties of microalgaechange when the culture conditions change. The effect of Nsource and UVR on the microalgal fatty acid compositionsdepends on different microalgal species, the range of nitro-gen concentration, different UVR fluxes and wavelengthsand different time of UVR exposure. Clearly, the variationin fatty acid composition with both N source and UVRexposure require consideration of these factors when opti-mizing the nutritional value of microalgae as feedstuff formariculture.

5. Abbreviations

EPA eicosapentaenoic acidFm maximal fluorescence (dark adapted)Fo minimal fluorescence (dark adapted)Fv variable fluorescence (dark adapted)

Table 6Fatty acid composition (% of total fatty acid) of Chaetoceros muelleri exposed to UVR under different N sourcea

N source N-Nt N-Am N-Ur

UVR UVAR High UVBR Low UVBR PAR UVAR High UVBR Low UVBR PAR UVAR High UVBR Low UVBR PAR

14:0 12.92(0.17) 13.17(0.19) 12.75(0.08) 13.62(0.81) 15.60(0.17) 15.87(0.34) 15.41(0.17) 17.49(0.09) 11.30(0.13) 11.93(0.15) 11.26(0.24) 11.50(0.15)16:0 19.38(0.92) 19.39(0.66) 19.06(0.18) 20.05(2.13) 17.25(1.55) 18.07(0.52) 16.64(1.42) 18.23(0.94) 18.64(0.58) 16.88(1.11) 18.38(0.64) 18.11(0.06)16:1n � 7 29.29(1.28) 29.48(0.28) 28.91(0.33) 29.58(1.55) 27.08(1.65) 28.70(0.51) 26.98(1.63) 28.39(0.67) 28.43(0.41) 28.24(1.05) 28.21(0.93) 27.34(0.23)20:4n � 6 5.22(0.09) 5.40(0.22) 5.35(0.04) 5.04(0.70) 6.47(0.25) 6.89(0.24) 6.67(0.18) 5.58(0.03) 5.08(0.04) 5.32(0.09) 5.05(0.06) 4.89(0.05)20:5n � 3 13.12(0.60) 12.98(0.36) 13.47(0.13) 12.32(1.87) 11.74(0.72) 11.15(0.34) 12.39(0.88) 9.61(0.37) 13.39(0.47) 14.23(1.48) 13.65(0.71) 13.22(0.13)22:6n � 3 1.13(0.12) 1.02(0.14) 1.21(0.03) 0.97(0.29) 1.23(0.11) 1.11(0.13) 1.30(0.14) 0.75(0.09) 1.42(0.06) 1.41(0.06) 1.40(0.11) 1.29(0.05)SFA 34.22(1.00) 34.27(0.77) 33.59(0.23) 35.40(2.91) 35.57(1.66) 36.68(0.82) 34.71(1.60) 38.29(0.90) 32.86(0.76) 31.17(1.68) 32.37(0.85) 32.56(0.15)MUFA 33.34(1.02) 33.64(0.49) 32.90(0.19) 33.62(1.67) 31.03(1.19) 32.15(0.42) 30.85(1.16) 32.45(0.70) 32.23(0.35) 32.07(1.12) 32.04(0.84) 31.45(0.24)PUFA 32.44(2.01) 32.10(1.26) 33.51(0.40) 30.98(3.23) 33.40(2.84) 31.17(1.24) 34.44(2.76) 29.26(1.61) 34.91(1.08) 36.76(2.77) 35.59(1.69) 35.99(0.26)

Note. The letters (Nt, Am and Ur) represent identical conditions to those described in Table 2.a Values are averages of four measurements. Standard deviations are shown in parentheses.

Table 7Summary of MANOVA (a) and SNK multiple comparison test results (b) on fatty acid composition of Chaetoceros muelleri exposed to UVR under different N sourceA

Source of variance 14:0 16:0 16:1n � 7 20:4n � 6 20:5n � 3 SFA MUFA PUFA

(a) N source F 1020.359 14.488 10.241 138.291 34.915 37.880 17.526 17.954UVR 31.843 1.302 1.228 17.679 6.592 4.287 1.497 3.061N source · UVR 10.735 1.726 1.515 4.109 1.926 1.863 1.431 2.028N source P-level <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001UVR <0.001 0.289 0.314 <0.001 0.001 0.011 0.232 0.040

N source · UVR <0.001 0.143 0.201 0.003 0.103 0.114 0.230 0.087

(b) N source AmaNtbUrc NtaAmbUrb NtaAmbUrb AmaNtbUrb UraNtbAmc AmaNtbUrc NtaAmbUrb UraNtbAmb

UVR CaB1bB2cAc BNS NS AaB2abB1bCc B1aB2aAaCb CaB1bB2bAb NS B2aB1abAabCb

N source · UVR Nt CaB1aB2aAa NS NS B1aB2aAaCa NS NS NS NSAm CaB1bB2cAc B1aB2abAbCc

Ur B1aCbB2bAb B1aB2bAbCc

Note. The letters (C, B1, B2 and A) and (Nt, Am and Ur) represent identical conditions to those described in Table 2.A Student-Newman-Keul (SNK) multiple comparison test results are arranged in decreasing order from left to right: a > b > c. Group with the identical superscript letters indicate that mean values arenot significantly different. Significant probabilities are in bold (P < 0.05).B NS: no significant.

170Y.Lianget

al./JournalofPhotochem

istryandPhotobiologyB:Biology82(2006)161–172

Y. Liang et al. / Journal of Photochemistry and Photobiology B: Biology 82 (2006) 161–172 171

Ik represents the PAR-value at the onset of light sat-uration

MUFA monounsaturated fatty acidsPAM pulse amplitude modulatedPUFA polyunsaturated fatty acidsrETRmax corresponds to the apparent light-saturated elec-

tron transport rateSFA saturated fatty acidsUVAR ultraviolet-A radiation, 315–400 nmUVBR ultraviolet-B radiation, 280–315 nmUVR ultraviolet radiationa the initial slope of the light curve (rETR/PAR)UPSIIe-max the maximum effective quantum yield of PSII

(=Fv/Fm)

Acknowledgements

This work was supported, in part, by the AustralianResearch Council. We also thank the China ScholarshipCouncil (CSC) provide the opportunity for the first authorto visit and carry out these experiments in Australia.

References

[1] P.A. Thompson, P.J. Harrison, J.N.C. Whyte, Influence of irradianceon the fatty acid composition of phytoplankton, J. Phycol. 26 (1990)278–288.

[2] K.S. Wang, T.-J. Chai, Reduction in omega-3 fatty acids by UV-Birradiation in microalgae, J. Appl. Phycol. 6 (1994) 415–421.

[3] M.A. Borowitzka, Microalgae for aquaculture: opportunities andconstraints, J. Appl. Phycol. 9 (1997) 393–401.

[4] W. Yongmanitchai, O.P. Ward, Growth and omega-3 fatty acidproduction by Phaeodactylum tricornutum under different cultureconditions, Appl. Environ. Microbiol. 57 (1991) 419–425.

[5] X.Q. Xu, J. Beardall, Effect of salinity on fatty acid composition of agreen microalga from an Antarctic hypersaline lake, Phytochemistry45 (1997) 655–658.

[6] J. Zhu, Y.K. Lee, T.M. Chao, Effect of temperature and growth phaseon lipid and biochemical composition of Isochrysis galbana TKI, J.Appl. Phycol. 9 (1997) 451–457.

[7] J.P. Fidalgo, A. Gid, E. Torres, A. Sukenik, C. Herrero, Effects ofnitrogen source and growth phase on proximate biochemical compo-sition, lipid classes and fatty acid profile of the marine microalgaIsochrysis galbana, Aquaculture 166 (1998) 105–116.

[8] S.O. Lourenco, E. Barbarino, J. Mancini-Filho, K.P. Schinke, E.Aidar, Effects of different nitrogen sources on the growth andbiochemical profile of 10 marine microalgae in batch culture: anevaluation for aquaculture, Phycologia 41 (2002) 158–168.

[9] J.P. Fidalgo, A. Gid, J. Abalde, C. Herrero, Culture of the marinediatom Phaeodactylum tricornutum with different nitrogen sources:growth, nutrient conversion and biochemical composition, Cah. Biol.Mar. 36 (1995) 165–173.

[10] R.C. Smith, B.B. Prezelin, K.S. Baker, R.R. Bridigare, N.P. Boucher,T. Coley, D. Karentz, S. MacIntyre, H.A. Matlick, D. Menzies, M.Ondrusek, Z. Wan, K.J. Waters, Ozone depletion: ultravioletradiation and phytoplankton biology in Antarctic waters, Science255 (1992) 952–959.

[11] J.B. Kerr, C.T. McElroy, Evidence for large upward trends ofultraviolet-B radiation linked to ozone depletion, Science 262 (1993)1032–1034.

[12] T. Shindell, D. Rind, P. Lonergan, Increased polar stratosphericozone losses and delayed eventual recovery owing to increasinggreenhouse-gas concentrations, Nature 392 (1998) 589–592.

[13] D. Karentz, I. Bosch, Influence of ozone-related increases inultraviolet radiation on Antarctic marine organisms, Am. Zool. 41(2001) 3–16.

[14] M. Blumthaler, W. Ambach, Indication of increasing solar ultravi-olet-B radiation flux in alpine regions, Science 248 (1990) 206–208.

[15] R. Stolarski, R. Bojkov, L. Bishop, C. Zerefos, J. Staehelin, J.Zawodny, Measured trends in stratospheric ozone, Science 256 (1992)342–349.

[16] A.T. Davidson, D. Bramich, H.J. Marchant, A. McMinn, Effects ofUV-B irradiation on growth and survival of Antarctic marinediatoms, Mar. Biol. 119 (1994) 507–515.

[17] J.J. Cullen, P.J. Neale, Effect of UV on short-term photosynthesis ofnatural phytoplankton, Photochem. Photobiol. 65 (1997) 264–266.

[18] J. Beardall, S. Beer, J.A. Raven, Biodiversity of marine plants in anera of climate change: some predictions on the basis of physiologicalperformance, Bot. Mar. 41 (1998) 113–123.

[19] J. Beardall, J.A. Raven, The potential effects of global climate changeon microalgal photosynthesis growth, and ecology, Phycologia 43(2004) 31–45.

[20] D.O. Hessen, H.J. De Lange, E. Van Donk, UV-induced changes inphytoplankton cells and its effects on grazers, Freshwat. Biol. 38(1997) 513–524.

[21] K. Sundback, S. Odmark, A. Wulff, C. Nilsson, S.-A. Wangberg,Effects of enhanced UV-B radiation on a marine benthic diatom mat,Mar. Biol. 128 (1997) 171–179.

[22] S. Odmark, A. Wulff, S.-A. Wangberg, C. Nilsson, K. Sundback,Effects of UV-B radiation in a microbenthic community of a marineshallow-water sandy sediment, Mar. Biol. 132 (1998) 335–345.

[23] J.H. Skerratt, A.D. Davidson, P.D. Nichols, T.A. McMeekin, Effectof UV-B on lipid content of three Antarctic marine phytoplankton,Phytochemistry 49 (1998) 999–1007.

[24] W.D. Larkum, W.F. Wood, The effect of UV-B radiation onphotosynthesis and respiration of phytoplankton, benthic macroalgaeand seagrasses, Photosynth. Res. 36 (1993) 17–23.

[25] R.M.Forster,H. Schubert,Wavelength dependence of photoinhibitionin the green algaChlorella vulgaris, Photosynthetica 33 (1997) 541–552.

[26] H. Herrmann, D.-P. Hader, F. Ghetti, Inhibition of photosynthesisby solar radiation in Dunaliella salina: relative efficiencies of UV-BUV-A and PAR, Plant Cell Environ. 20 (1997) 359–365.

[27] J. Nilawati, B.M. Greenberg, R.E.H. Smith, Influence of ultravioletradiation on growth and photosynthesis of two cold ocean diatoms, J.Phycol. 33 (1997) 215–224.

[28] P. Heraud, J. Beardall, Changes in chlorophyll fluorescence duringexposure of Dunaliella tertiolecta to UV radiation indicate a dynamicinteraction between damage and repair processes, Photosynth. Res.63 (2000) 123–134.

[29] G. Dohler, T. Biermann, Impact of UV-B radiation on the lipid andfatty acid composition of synchronized Ditylum brightwellii (West),Grunow. Naturforsch. 49C (1994) 607–614.

[30] J.I. Goes, N. Handa, S. Taguchi, T. Hama, Effect of UV-B radiationon the fatty acid composition of the marine phytoplankter Tetrasel-mis sp.: relationship to cellular pigments, Mar. Ecol. Prog. Ser. 114(1994) 259–274.

[31] H.J. De Lange, E. Van Donk, Effects of UVB-irradiated algae on lifehistory traits of Daphnia pulex, Freshwat. Biol. 38 (1997) 711–720.

[32] C.C. Parrish, J.S. Wells, Z. Yang, P. Dabinett, Growth and lipidcomposition of scallop juveniles Placopecten magellanicus, fed theflagellate Isochrysis galbana with varying lipid composition and thediatom Chaetoceros muelleri, Mar. Biol. 133 (1999) 461–471.

[33] G.A. Dunstan, J.K. Volkman, S.M. Barrett, J.-M. Leroi, S.W. Jeffrey,Essential polyunsaturated fatty acids from 14 species of diatom(Bacillariophyceae), Phytochemistry 35 (1994) 155–161.

[34] F. Pernet, R. Tremblay, E. Demers,M. Roussy, Variation of lipid classand fatty acid composition of Chaetoceros muelleri and Isochrysis sp.grown in a semicontinuous system, Aquaculture. 62320 (2003) 1–14.

[35] R.R. Guillard, J.H. Ryther, Studies of marine planktonic diatoms. I.Cyclotella nana Hustedt and Detonula confervacea (Cleve) Gran, Can.J. Microbiol. 8 (1962) 229–239.

172 Y. Liang et al. / Journal of Photochemistry and Photobiology B: Biology 82 (2006) 161–172

[36] G. Lepage, C.C. Roy, Improved recovery of fatty acid through directtransesterification without prior extraction or purification, J. LipidRes. 25 (1984) 1391–1397.

[37] L.W. Jones, B. Kok, Photoinhibition of chloroplast reactions. 1.Kinetics and action spectra, Plant. Physiol. 41 (1966) 1037–1043.

[38] M.J. Behrenfeld, J.W. Chapman, J.T. Hardy, H. Lee, Is there acommon response to ultraviolet-B radiation by marine phytoplank-ton, Mar. Ecol. Prog. Ser. 102 (1993) 59–68.

[39] R.B. Setlow, The wavelengths of sunlight effective in producing skincancer: a theoretical analysis, Proc. Natl. Acad. Sci. USA 71 (1974)3363–3366.

[40] J.J. Cullen, P.J. Neale, M.P. Lesser, Biological weighting function forthe inhibition of phytoplankton photosynthesis by ultraviolet radia-tion, Science 258 (1992) 646–650.

[41] M.M. Caldwell, L.B. Camp, C.W. Warner, S.D. Flint, Action spectraand their key role in assessing biological consequences of solar UV-Bradiation change, in: R.C. Worrest, M.M. Caldwell (Eds.), Strato-spheric Ozone Reduction, Solar Ultraviolet Radiation and Plant Life,Springer-Verlag, Berlin, 1986, pp. 87–111.

[42] F.E. Quaite, B.M. Sutherland, J.C. Sutherland, Action spectrum forDNA damage in alfalfa lowers predicted impact of ozone depletion,Nature 358 (1992) 576–578.