Essential polyunsaturated fatty acids from 14 species of diatom (Bacillariophyceae)

Phymchrmistry. Vol. 35, No. 1, pp. 155-161. 1994 hinted in Great Britain.

ESSENTIAL

GRAEME

A.

0

0031~9422/W S6.00+0.00 1993 Pergamon Press Ltd

POLYUNSATURATED FATTY ACIDS FROM 14 SPECIES OF DIATOM (BACILLARIOPHYCEAE)

DUNSTAN,

JOHN K. VOLKMAN,STEPHANIEM. BARRETT,JEANNIE-MARIELEROI* and S. W. JEFFREY+

CAIRO Division of Oceanography. G.P.O. Box 1538, Hobart, Tasmania 7001, Australia; lCSIRO Division of Fisheries, G.P.O. Box 1538, Hobart, Tasmania 7001, Australia

(Receivedin revisedform 5 July 1993) Key Word Index-Bacillariophycae; polyunsaturated fatty acids.

diatoms; biotechnology;

chemotaxonomy;

mariculture; lipids,

Abstract-The lipid class and total fatty acid compositions of 14 species of diatom (class Bacillariophyazae) were examined. The major lipid class in these microalgae was polar lipid (50.1~90.8%), although in some species, triacylglycerols (0.3-47.7%) and free fatty acids (O-29.2%) were also abundant in the logarithmic phase cultures. The major fatty acids in most species were 14:0,16:0,16: l(n-7) and 20: 5(n-3). The polyunsaturated fatty acids 16:2(n-4), 16:3(n-4), 16:4&l), 18:4(n-3) and 22:6(n-3) also comprised a significant proportion (>4% each) of the total fatty acids in some species. The characteristic fatty acid composition of diatoms is readily distinguishable from those of other classes of microalgae. The mean proportion of 22:6(n-3) was significantly higher in representatives from the order Centrales (4.9%) than the Pennales (1.6%), but there was no consistent difference in the proportions of the other fatty acids. Four of the pennate diatoms contained significant proportions of arachidonic acid, 20:4(n-6) (3.5-5.6%), but most species examined contained very low proportions of (n-6) polyunsaturated fatty acids. Based on the lipid compositions, all of the diatom species studied are potentially suitable sources of some of the longer chain polyunsaturated fatty acids.

INTRODUCIION

Microalgae from the class Bacillariophyceae (diatoms) are important in marine food webs. Some species which grow well in mass culture have been used successfully as microalgal feedstocks for larval animals reared intensively [l]. Since the primary storage material in diatoms is lipid, they are of potential value to the biotechnology industry for the production of lipids and, in particular, polyunsaturated fatty acids (PUFA). Relative to the large number of species within this group (> 1300) [2], very few species have been examined for lipid class and fatty acid composition [3-73, and only a few comparative studies have been performed using modem analytical techniques [8]. Diatoms examined so far generally contained high concentrations of the (n-3) PUFA 20:5(n-3) [3, 43 and smaller proportions of 22: 6(n-3) [S]. Both of these PUFA are considered to be essential to the diets of many marine animals [l, 9, 101, which suggests that these microalgae may be a suitable source of PUFA for the supplementation of mariculture fecdstocks. Some species of diatom have also been reported to contain significant proportions of the (n-6) PUFA 20:4 (n-6) [8, 111. In the present paper, we report details of the fatty acid and lipid compositions of 14 species of diatom to examine possible chemotaxonomic trends and their potential nutritional suitability as feedstocks for the mariculture

industry and as a source of PUFA for the biotechnology industry. Data on most of the species examined in the present study have not been reported previously, although data on some well-studied species are included for comparison.

REWLm

AND DL!SCUSSlON

Diatom species (14) from the orders Centrales and Pennales were examined for lipid class and total fatty acid composition. The species name-s, CSIRO Culture Collection of Microalgae strain codes, culture temperature and growth media are shown in Table 1. Although culture conditions am induce quantitative changes in the lipid and fatty acid composition of microalgae [S, 12, 133, all species examined were grown under conditions known to produce health9 ails and vigorous growth [ 143; all were sampled at late-logarithmic growth phase. The present study was designed to examine the lipid compositions of healthy cells at similar physiological stages and to show that the diatoms as a group are quite different with respect to qualitative fatty acid composition compared with microalgae from other classes. Other desirable features, e.g. good growth characteristics and lack of toxins need to be established 155

G. A. DUNSTAN et al.

156

Table I. Diatom species examined, growth temperatures, growth media, total cellular lipid content (pg cell-‘) and per cent compositions (%) of individual lipid classes

CSiRO stram

Species

Culture Culture temp.” mediaf

Total lipids (pgceff-‘)

Lipid class as a percentage of total lipids* --.-- -- -. - -. - -- - .- ------PL PG DG+ST FFA TG HC

Cent rafes sp. Wuzosolenia serigeru Skeleronema costafum Ske~eion~rnusp.

cs-151 CS-62 CS-181 cs-2s2

20

20 20 25

f/, ijz f/l 0,

1196 114 IO 3.2

84.3 72.4 63.6 81.9

--: 4.1 13.3 6.2

2.3 3.1 0.7 1.7

Tha~assi~)sira stellaris

CS-16

20

G

30

84.7

--

2.1

Coscinodiscus

.9.4 17.4 7.9 2.1

11.9 7.1 4.5 1.3 9.3

1.5 3.9 0.3 0.6 1.8

Pennafes Amphiprora Amphora

hyalina

sp.

Cylindrotheca Fragilaria

jiisiformis

pinnata

Haslea osrrearia Naricuia sp.

CS-285

20

Y,

16

81.4

6.3

1.4

6.8

2.1

2.0

cs-1% cs-13(3

20 20

cs-121

2.5

cs-250-2 cs-4%

20 20

r/, fi, f/, f/J, f/z

20 8.2 5.4 67 If

92.4 90.7 89. I 53.9 50.1

trll tr -.-. 4.6 tr

0.9 2.1 3.7 1.7 0.1

2.1 2.9 0.9 14.7 -

3.1 2.5 5.0 14.5 47.7

1.5 1.8 1.3 10.1 2.0 0.3

CS-q

20

VI

9.7

89.7

8.4

0.3

1.0

0.3

~alassianema

n~tzsc~io~des

CS-146

28

V,

If 4’

85.2

.-

1.8

5.1

2.5

1.4

Thalassiothrix

heteromorpha

CS-132-8

20

fit

45

59.4

3.2

3.6

5.9

2.0

Nirzschia

ciostarium

25.9

lI’L = polar lipids and chlorophylls; PG = unidentified peak includes some pigments DG + ST =diacyfgfycerofs and sterofs; FFA = free fatty acids: TG = triacyfgfycerofs; HC= hydrocarbons; total lipid includes minor contributions from alcohols. tSee ref. [ 143 for details. $Not detected. $Axenic. lITrace fess than 0.1%. Tfotaf lipid included an unidentified peak (4%) efuting between free fatty acids and triacyfgfycerofson TLC-FID.

Lipid

class

composition

TLC-FID data showed that the total cellular lipid (3-f 196 pg cell- ‘) and the lipid class compositions varied significantly (Table I), even though all cultures were harvested at late-logarithmjc growth phase. There was no consistent difference in lipid class composition between the representatives from the orders Centrales and Pennafes. Total cellular lipid was generally dependent on cell size (species known to have larger cells had more lipid, e.g. Coscinodiscus sp.; Table 1) and the accumulation of storage lipid. Polar lipids (mostly phospholipid and glycolipid. and in diatoms, sulpholipid 1153) are major components of cellular membranes and typically the most abundant lipid class in microatgal cells, as they were in the diatoms examined in the present study (SO-91%). Significant amounts of triacylglycerols were also evident in some species (0.348%). In logarithmic phase cultures of microalgae, the triacylglycerols are usually present in low amounts [S, 163, but they are known to accumulate as storage lipid in stationary phase cells 1133. However, non-senescent diatom cells have been noted to contain elevated levels of t~acylgly~rois [17], as was confirmed by the present study. High proportions of free fatty acids were found in most species examined (up to 26%; Table I), as reported in other species of diatom [S, IS]. These are thought to result from the release of enzymes which hydrolyse ester bonds and free the fatty acids [S]. The free fatty acids were probably derived from hydrolysis of triacyiglycerob, with diacylglycerols

(Table I), and possibly monoacylglycerols and glycerol also being produced. Because the solvent system used did not differentiate lyso-phospholipids or lyso-glycolipids from other polar lipids, it could not be established whether the free fatty acids were also in part derived from polar lipids. Previous studies of the diatom Nirzschiu pungens showed that the high free fatty acid content was not due to hydrolysis from poor sample handling [IS]. Alternatively, the high content of free fatty acids observed in some species of diatom may reflect their presence as cellular constituents, although elevated concentrations of free fatty acids are not typical of other microalgae [S, 131. Small amounts of hydrocarbons were also detected (0.3-10% of total lipids; Table 1) and they were predominantly n-C*, 15 and n-C,,,, in all species, except in Hastea ostreuria, and Rhizosolenia setigera where C,, and C,, isoprenoid alkenes, respectively, were abundant [191* Fatty acid composition The total cellular content of fatty acids in the diatoms examined was pro~rtiona~ to the total cellular lipid and ranged from 1.6 to 52.4 pg cell - ’ and 898 pg ceII - t for the large celled Coscinodiscus sp. (Tables 2 and 3). Saturated fatty acids represented 16-37% of the total fatty acids. Palmitic acid (16:O) was abundant in most species, although N. closterium, Thalassionema nitrschioides, Thaiassiothrix heteromorpha and the two species of Skeletonemu, contained extremely low proportions (less than

Polyunsaturated

fatty acids from diatoms

157

Table 2. Relative proportions (%) of individual fatty acids and cellular content (pgcell- ‘) of total fatty acids from five species of centric diatom (Centrales)

Species strain Saturated fatty acids 14:o 15:o 16:O 18:0 24:0 Subtotal* Monoenoic fatty acids 16: l(n-7) 16: l(n-5) 16: l(n-13)t 18: l(n-9) 18: l(n-7) 22:l 24:l Subtotal* Polyunsaturated fatty acids 16:2(n-7) 16: 2(n-4) 16: 3(d) 16:4(n-1) 18 : 2(n-6) 18:3(d)

18:3(n-3) 18 :4(n-3) 20: 3(n-6) 20:4(n-6) 20:4(n-3) 20: 5(n-3) 22:4(n-6) 22: 5(n-6) 22: 5(n-3) 22: 6(n-3) Subtotal* Total Cellular content of fatty acids (Pgcell- ‘)

Coscinodiscus sp. cs-151

Rhizosolenia setigffa CS-62

Skeletonma

Skeletonema

Thalassiosira

cos1awn G-181

SP. CS-252

slellaris

11.0

10.5 0.4 16.5 0.4 -t 28.0

17.6 0.4 16.4 1.3 0.1 36.6

11.9 0.3 5.1 0.1 17.5

19.6 0.5 3.9 0.3 0.1 24.5

14.6 0.2 1.3 1.0 3.2 trS 21.0

17.7 0.3 0.9 0.4 7.6 27.2

19.1 0.3 0.7 0.3 0.2

19.8 0.6 0.4 0.3 2.2 0.2 0.5 23.9

0.6 2.1 4.9 4.2 1.6 0.8 0.6 3.5 0.1 1.1 0.5 26.0 0.7 0.1 4.6 51.0 100.0

1.4 1.8 1.0 3.5 1.9 0.5 0.6 1.7 0.3 1.0

900

17.5 tr 6.1 36.2 100.0 52

20.6 1.0

0.9 16.9 8.1 1.4 0.8 2.8 0.2 26.1

-

4.1 13.3 7.8 5.6 0.5 0.2 0.4 1.0 0.1 tr 0.1 18.3 -

4.7 62.0 100.0

tr 4.2 51.6 loo.0

5.6

1.6

CS-16

0.5 8.3 0.4 0.1 20.3 17.1 0.2 1.0 0.4 1.9 0.8 21.9 2.2 4.4 14.9 1.1 0.5 0.3 0.2 5.2 0.1 0.1 0.1 25.3 0.6 tr 4.8 57.8 100.0 19

*Subtotals include some minor fatty acids not listed (~0.8% each). tNot detected. ITrace; less than 0.1%.

10%; Tables 2 and 3). Such low proportions of this fatty acid have been reported for only a few diatoms [4, 5,7], but this is unusual for healthy microalgal cells; low contents have been noted for some species grown under extremely low light intensities [20]. There were high proportions of 14:0 in most species (mostly S-20%; Tables 2 and 3) and large amounts were present in the non-aerated cultures of T. nitzschioides (23.3%) and T. hereromorpha (29.0%). Small proportions (~2%) of 24:0 and 24: 1 were found in a few species (Tables 2 and 3). The higher proportions of cis-vaccenic acid [ 18 : I (n7)] relative to oleic acid [18: l(n-9)] in the non-axenic cultures (Tables 2 and 3) are likely to have derived from low concentrations of contaminating bacteria [8, 163.

Polyunsaturated fatty acids represented from 35 to 62% of the total fatty acids. The major PUFA was 20: 5(n-3) (12.2-30.2%), although 22:6(n-3) (O-6.1%), 20:4(n-6) (O-5.6%), 16:2(n-4) (0.9-13.3%), 16:3(n-4) (l.O-21.6%), 16:4(n-1) (tr-18.6%) and 18:4(n-3) (O.l-5.5%) were significant constituents in many species. Minor contributions from other PUFA are shown in Tables 2 and 3. The tropical species, Fragiloria pinnata, lacked both 22:5(n-6) and 22:6(n-3); instead the major C,, PUFA present were the presumed precursors of these PUFA, 22:4 (n-6) (0.2%) and 22:5(n-3) (l%, Table 3). This suggests that either the A4-desaturase was not operating or, since the hydrocarbons n-C,, :5 and n-C, I :6 were present [19], both 22:5(n-6) and 22:6(n-3) PUFA

of fatty acids

fatty acids

16.4

9.6

1.9 55.8 100.0

0.1

tr

I1

0.2 0.7 61.7 100.0

2.0 5.1 21.6 0.2 0.8 0.2 0.2 0.9 0.1 0.5 0.8 30.2

-

25.0 2.1 2.8 9.5 2.7 1.4 0.3 0.5 5.5 0.1 0.8 0.2 30.0

13.6 0.1 1.4 0.6 0.2

20. I 0.1 0.7 I.1 2.3 0.1 -

Cylifufrotheca

2.2

0.8 0.3 1.1 48.5 100.0

2.8 0.1 2.0 0.5 4.9 0.7 20.3

I.5

I.1 3.0 5.0 5.4

21.4

19.7 0.6 0.1 0.6 0.3 -

8.7 0.6 20.0 0.4 0.3 30.1

CS-13

JiiS~OTt?li.S

*Subtotals include some minor fatty acids not listed (~0.8% each). tNot detected. $Trace; less than 0.1%.

(Pgcell- ‘)

18: l(n-9) 18: l(n-7) 22:1 24:1 Subtotal* Polyunsaturated 16: 2(n-7) 16 : 2(n-4) 16: 3(A) 16:4(n-I) 18 :2(A) 18:3(h) 18:3(n-3) 18 :4(n-3) 20: 3(n-6) 20: 4(n-6) 20:4(n-3) 20: 5(n-3) 22 :4(d) 22 : 5(n-6) 22: 5(n-3) 22 : 6(n-3) Subtotal* Total Cellular content

6.0 0.4 14.6 0.3 0.7 21.9

sp. CSlO

Amphora

7.1 0.3 10.7 0.2 0.8 19.2

CS-28

strain

Saturated fatty acids 14:o IS:0 16:O l8:O 24:0 Subtotal* Monoenoic fatty acids 16: l(n-7) 16: l(n-5) 16:l(n-l3)1

hyalina

Species

Amphiprora

3.0

40.2 1tM.O

0.8 0.4 0.1 1.1 0.6 5.6 0.2 20.7 0.2 1.0

tr

0.8 4.5 4.8

31.6

26.3 0.1 0.2 0.6 4.3 -

1.4 0.4 25.2 0.8 0.1 28.2

CS-121

pinnala

Fragilaria

27

0.2 3.3 35.1 lcQ.0

tr

1.1 4.1 10.6 0.3 0.5 0.7 1.0 0.9 tr 0.2 0.7 12.2

28.6 0.3 0.1 0.9 I.6 31.7

8.6 0.4 20.4 0.8 0.2 33.2

CS-250-2

ostrervia

Has&a

Nilrschia

3.5

4.3

2.4 58.7 100.0

2.6 35.8 100.0 7.0

tr

4.0 0.1 24.2 1.1

-

-

0.9 9.4 9.8 3.2 0.5 0.5

22.8 0.8 1.4 0.1 0.2 25.4

‘;’ 15.9

8.2 0.4 7.2

closterium cs-5

tr

0.2 2.4 4.1 0.2 2.9 0.7 0.2 0.9 0.1 0.5 0.1 21.0 -

34.4 0.4 1.5 0.4 0.3 37.0

4.9 0.4 21.3 0.4 0.1 27.2

Navicula sp. CS-46 7Massiothrix

0.1

3.8

1.0 48.8 100.0

-

3.9 3.1 14.1 tr 0.6 0.7 0.1 0.1 0.2 3.5 0.1 25.2

18

1.9 1.6 3.2 18.6 1.4 0.7 0.1 0.6 0.3 2.2 0.1 12.9 0.2 0.1 1.6 43.6 100.0

14.6 0.1 0.5 0.7 3.9 1.4 1.9 23.7

0.6 2.9 32.7

::: 30.9 17.1 0.3 0.7 0.3 1.4 trS 20.3

28.9 0.3

CS-132-8

heteromorpha

23.3 0.4

nittschioides CS-146

Thalassionema

Table 3. Relative proportions (%) of individual fatty acids and cellular content (pgcell- ‘) of total fatty acids from nine species of pennate diatom (Pennales)

159

Polyunsaturated fatty acids from diatoms were efficiently decarboxylated to the corresponding alkenes [21]. Four of the pennate diatoms examined contained elevated proportions of arachidonic acid 20:4(n-6) (3.5-5.6% of total fatty acids; Table 3), as has been found in other species from the same genera (e.g. Cylindrotheca sp. [3], Nitzchia spp. [6, 123). Other microalgae that contain significant amounts of this PUFA include centric diatoms of the genus Chaetoceros 4-6% [8, 113, eustigmatophytes of the genus Nannochloropsis spp. with 4-9% [22, 231 and the rhodophyte Porphyridium cruentum in which 36% of its fatty acids is 20:4(n-6) [24]. Most marine microalgae contain less than 1% of the total fatty acids as (n-6) Czo and Cz2 PUFA [3, 5, 81. Commercial applications The (n-3) PUFA, 20:5 (n-3) and its elongation and desaturation product 22:6(n-3) are major membrane components of marine animals and are considered to be essential dietary components [l, 9, lo]. As well as these (n-3) PUFA, 20:4(n-6) is a significant component of the lipids of certain benthic macroalgae and animals which feed directly on them [27 and refs cited therein]. These include commercially valuable species, e.g. abalone (Hdiotis spp.) and spiny lobsters (Pam&us spp.). Whether 20:4(n-6) is essential to these marine animals needs to be examined for artificial feed formulation. In most marine animals, 20:5(n-3) and 20:4(n-6) are obtained directly from the diet and small amounts of both C2,, PUFA are converted into the anti-inflammatory leucotrienes [25] and the prostaglandins necessary for growth and reproduction [26]. Microalgae containing elevated contents of 20:4(n-6), e.g. those examined in the present study, may prove to be a valuable source of this PUFA. Lipid content and composition can be manipulated by changing the culture conditions [ 12, 18,281; as a result, the favourable PUFA content of diatoms may be further enhanced. The decarboxylase enzyme which is thought to form the n-&r :6 hydrocarbon from 22: 6(n-3) PUFA in microalgae [21] may reduce the amounts of this ‘essential PUFA. In spite of the reduced proportions of 22:6(n-3) due to decarboxylation, the diatoms still contained very high proportions of the other ‘essential’ (n-3) PUFA, 20:5(n-3) (12.2-30.2%; Tables 2 and 3); many species have been shown to be valuable feeds for larval animals [l]. Some diatoms may also be a ready source of free (non-esteritied) fatty acids and monoacylglycerols (Table 1).

from these orders (Table 1). Qualitatively, the fatty acid compositions of the diatoms were similar to each other in the present study and to other analyses of diatoms reported in the lit. [4-8, 12, 183. In diatoms, the proportion of 16: l(n-7) usually exceeds that of 16:0 (Tables 2 and 3) [3-5, 7, 81. Similar proportions of both fatty acids are also evident in the Eustigmatophyceae [22, 231 and some Prymnesiophyceae (e.g. Paoloua spp. formerly Monochrysis [S, 8, 291). All species of diatom examined contained high proportions of C,, PUFA (7-30% of total fatty acids, Tables 2 and 3) which had terminal double bonds at positions (n7), (n-4) and (n-l). The (n-7) and (n-4) Cl6 PUFA positional isomers occur in other microalgae of the Chromophyta, although in significantly reduced proportions relative to the diatoms (generally total less than 5% of fatty acids) [3,5,7,8,22,30]. In comparison, the Chlorophyta (Chlorophyceae and Prasinophyceae) also contain significant proportions of C,, PUFA (15-25%), but in these green microalgae the isomers are predominantly (n-6) and (n-3) [8, 163. The presence of 16:4(n-1) in nearly all species of diatom studied (up to 18.6%, Tables 2 and 3), suggests this PUFA should not be used as a specific marker for the toxic diatom N. pungens [18]. However, 16:4(n-1) may be useful as a more general marker for diatoms in complex communities of microalgae [31,32]. Nevertheless, it cannot be used quantitatively to estimate diatom biomass because the abundance and proportions of this and other fatty acids vary with culture conditions and between species [ 18,281. The diatoms studied contained high proportions of the long chain PUFA 20:5(n-3) (12.2-30.2% of total fatty acids), while only a few species contained significant proportions of 22:6(n-3) (up to 6.1%). The use of capillary column GC enabled the detection of small amounts of 22:6(n-3) not reported in previous studies which utilized packed column GC [3, 6, 7, 123. Species of microalgae which produce 18: S(n-3) (up to 28%, some species of Dinophyceae [33], Prymnesiophyceae [30] and Prasinophyceae [16]), tend to have lower proportions of 20:5(n-3) (tr-12%) than C,, PUFA (10-19%,X presumably due to chain shortening of 20:5(n-3) to 18: 5(n-3). The remaining Chromophyta (including the diatoms) and those prasinophytes (e.g. Tetraselmis spp. [16]) and prymnesiophytes (e.g. Paulova spp. [29]) lacking 18: 5(n-3), tend to have significant proportions of Cl0 PUFA (5-28%) and relatively less Ct2 PUFA (0- 13%) [S, 8,16,22,23,29], whereas microalgal Chlorophyceae generally have very little of either PUFA type [8, 163.

Chemotaxonomy

The mean proportion of 22:6(n-3) in the representatives from the order Centrales was 4.9% (s.e. =0.33), from the Pennales 1.6% (s.e. = 0.33). These means were significantly different (P c 0.001) using a t-test assuming unequal variances (Rehrens-Fisher statistic). There was no consistent difference in the proportions of the other fatty acids (Tables 2 and 3). There was also no consistent difference between the lipid class compositions in species

EXPERIMENTAL

Micro&al cultures. Cultures of the 14 species of diatom were obtained from the CSIRO Culture Collection of Microalgae [14]. All cultures were grown in medium f2, except T. stellaris which was grown in medium G (Table 1). Media were prepd from autoclaved sea water with filter-sterilized nutrients added aseptically; media compositions are detailed elsewhere [ 143. All equipment

160

G. A. DUNSTAN et al.

for the cultures (glassware and tubing) was sterilized by autoclaving and all inoculations were performed under asepetic conditions to avoid contamination. Cultures of the tropical species, F. pinnata and Skeletonema sp., at 28’; the were grown at 25”, T. nitzschioides remaining temperate species were grown at 20” (kO.5’). All cultures were illuminated from beneath with white fluorescent light (Daylight 70-1OO~Em-*sec~~’ tubes) on 12: 12 hr light : dark cycles. After 3-5 days, the light used to illuminate the cultures of R. setigera, Amphiprora hyalina and T. nitzschioides was filtered through 4OpEm-’ glass (‘blue light’, signal Belgian set- ‘) to improve cell densities or cell morphology [ 143. Cultures were grown in 700 ml of media in l-l Erlenmeyer flasks and aerated with filtered air (0.2 pm membrane filters) supplemented to 0.5% (v) with CO,, except for T. nitzschioides and T. heteromorpha which grew significantly better without aeration in 150ml of media in 5 x 250-ml Erlenmeyer flasks. All cultures were harvested towards the end of logarithmic phase. Lipid extraction and fractionation. Cells were harvested from 600 ml of culture medium by filtering through precombusted 47-mm-diameter glass libre filter (GFC). and extracted immediately with CHCI,-MeOH-H,O (5: lO:4, 5 x 5 ml) [34]. Samples were ultrasonicated between each extraction. The combined extracts were partitioned with CHCI,-H,O (1 : 1)to give a final solvent ratio of CHCI,-MeOH-HZ0 (10:10:9). Lipids were recovered from the lower CHCI, phase by removing solvents under vacuum and were then stored in CHCI, under N, at -20”. Concns of major lipid classes were determined by analysing a portion of the total lipid extract (in triplicate, with variation not exceeding f5% for each component) with a TLC-FID analyser using hexaneEt,O-HOAc (60: 17:0.2) [35]. Total fatty acid Me esters (FAME) were formed directly by heating a portion of the total extract with 3 ml of MeOH--CHCl, -HCI (10: 1: 1) at 80” for 2 hr under N,. After cooling and addition of 1 ml of Milli-Q H,O, FAME were extracted with 3 x 3 ml hexane-CHCI, (4: 1). After removal of the solvent with N,, FAME were dissolved in CHCI, containing a known amount of Me heptadecanoate (int. standard) and were stored at -20’ until analysis l-2 days later. Fatty acid analysis. Samples were analysed by GC-FID with cool on-column inj.; high purity H, was the carrier gas and FID temp. was at 270’. Samples (0.4~1) were injected on a polar 70% cyanopropyl siloxane (BP-X70) fused-silica column (50 m x 0.32 mm i.d.). After inj., the oven was held at 45. for 2 min and then temp. prog. to 120” at 30” min- ‘, then to 260’ at 3‘ min. ’ and held isothermally until all peaks had eluted. To distinguish between co-elutions and verify some identifications, all samples were analysed further on a non-polar Mesilicone (HP-l) fused-silica capillary column (50 m x 0.32 mm i.d.) under similar conditions, except that the oven was temp. programmed to 320” and the FID was at 330”. FAME standards of known composition were chromatographed to ensure accurate quantitation. Peak areas were quantified with integrating software on a PC. Fattv acid identifications were confrmed bv CC-MS.

Acknowledgements-This work was funded by FIRDC grants 1988/69 and 1991/59, and a grant from Chevron Oil Field Research. We thank MS T. O’Leary, Dr G. Maguire and an anonymous referee for useful comments on the manuscript, and MS Kathy Haskard for statistical analyses.

REFERENCES

I. Enright, C. T., Newkirk. G. F., Craigie, J. S. and Castell, J. D. (1986) J. Exp. Mar. Biol. Ecul. 96, 1. 2. Sournia, A., Chretiennot-Dinet, M.-J. and Ricard, M. (1991) J. Plankton. Res. 13, 1093. 3. DeMort, C. L., Lowry, R., Tinsley, I. and Phinney, H. K. (1972) J. Phycol. 8, 211. 4. Orcutt. D. M. and Patterson, G. W. (1975) Comp. Biochem. Physiol. SOB, 579. 5. Ackman, R. G., Tocher, C. S. and McLachlan, J. (1968) J. Fish. Res. Bd. Can. t5, 1603. 6. Kates, M. and Volcani, B. E. (1966) Biochim. Biophys. Acta 116, 264. 7. Cheucas, L. and Riley, J. P. (1969) .I. Mar. Biol. Ass. U.K. 49,97. 8. Volkman, J. K., Jeffrey, S. W., Nichols, P. D., Rogers, G. I. and Garland, C. D. (1989) J. Exp. Mar. Biol. Ecol. 12.8, 219. 9. Jones, D. A., Kanazawa, A. and Ono, K. (1979) Mar. Biol. 54, 261. 10. Watanabe, T., Kitajima, C. and Fujita, S. (1983) Aquaculture 34, 115. 11. Napolitano, G. E., Ackman, R. G. and Ratnayake, W. M. N. (1990) J. World Aquaculture Sot. 21, 122. 12. Opute, F. I. (1974) J. Exp. Bot. 25, 823. 13. Dunstan, G. A., Volkman, J. K., Barrett, S. M. and Garland, C. D. (1993) J. Appl. Phycol. 5, 71. 14. Jeffrey, S. W. (1980) in CUR0 Dioision of Fisheries and Oceanography, Annual Report (1977-1979) pp. 22-43. CSIRO, Australia. 15. Kates, M. (1989) in Marine Biogenic Lipids, Fats and Oils (Ackman, R. G., ed.), Vol. I, pp. 389-427. CRC Press, Florida. 16. Dunstan, G. A., Volkman. J. K., Jeffrey, S. W. and Barrett, S. M. (1992) J. Exp. Mar. Biol. Ecol. 161, 115. 17. Parrish, C. C. and Wangersky, P. J. (1990)J. Plankton Res. 12, 1011. 18. Parrish, C. C., deFreitas, A. S. W., Bodennec, G., Macpherson. E. J. and Ackman, R. G. (1991) Phytochemistry 30, 113. 19. Volkman, J. K.. Barrett, S. M. and Dunstan, G. A. (1993) Ory. Geochem. (in press). 20. Thompson, P. A., Harrison, P. J. and Whyte, J. N. C. (I 990) J. Phycol. 26, 278. 21. Lee, R. F. and Loeblich, A. R. III (1971) Phytochemistry 10, 593. 22. Volkman, J. K., Brown, M. R., Dunstan, G. A. and Jeffrey, S. W. (1993) J. Phycol. 29, 69. 23. Mourente, G., Lubian, L. M. and Odriozola, J. M. (1990) Hydrobiologia 203, 147. 24. Nichols, B. W. and Appleby, R. S. (1969) Phytochemistry 8, 1907.

Polyunsaturated fatty acids from diatoms 25. Tocher, D. R. and Sargent, them. Physiol. 87B, 733.

J. R. (1987) Comp. Bio-

26. Ruggeri, B. and Thoroughgood, C. A. (1985) Mar. Ecol. Prog. Ser. 23, 301. 27. Dunstan, G. A., Sinclair, A. J., O’Dea, K. and Naughton, J. M. (1988) Comp. Biochem. Physiol. 91B, 165. 28. Gillan, F. T., McFadden, G. I., Wetherbee, R. and Johns, R. B. (1981) Phytochemistry 20, 1935. 29. Volkman, J. K., Dunstan, G. A., Jeffrey, S. W. and Keamey, P. S. (1991) Phytochemistry 30, 1855. 30. Volkman, J. K., Smith, D. J., Eglinton, G., For&erg, T. E. V. and Corner, E. D. S. (1981) J. Mar. Biol.

Pllno d:l-L

161

Assoc. U.K. 61, 509. 31. Mayzaud, P., Claustre, H. and Augier, P. (1990) Mar. Ecol. Prog. Ser. 60, 123. 32. Morris, R. J., McCartney, M. J., Joint, 1. R. and Robinson, G. A. (1985) J. Exp. Mar. Biol. Ecol. 86, 151. 33. Nichols, P. D., Jones, G. J., De Leeuw, J. W. and Johns, R. B. (1984) Phytochemistry 23, 1043. 34. Bligh, E. G. and Dyer, W. J. (1959) Can. J. Biochem. Physiol. 37,911. 35. Volkman, J. K. and Nichols, P. D. (1991) J. Planar Chromat. 4, 19.