Journal of Paleolimnology1: 215-227, 1988 9 KluwerAcademicPublishers, Dordrecht- Printed in the Netherlands

215

Experimental determination of carotenoid degradation P. R. Leavitt 1 Department of Biology, Queen's University, Kingston, Ontario, Canada, K7L-3N6; ~Current address: Department of Biological Sciences, University of Notre Dame, Notre Dame, IN, USA, 46556

Received 6 May 1988; accepted 14 September 1988

Key words: Carotenoids, degradation, myxoxanthophyll, t-carotene, sediment, paleolimnology

Abstract

Reversed-phase thin layer chromatography was used to quantify carotenoid degradation resulting from the in vitrodecomposition of Oscillatoria utermi~hlii. Laboratory conditions simulated lacustrine sediments. The effects of light, oxygen, temperature and the presence of a sedimentary bacterial flora on the rate and extent of degradation were evaluated. Under anaerobic conditions, bacterially-mediated decomposition of O. utermOhliidid not result in statistically significant (P > 0.05) declines in the concentrations of most carotenoids at either 6 ~ after 37 weeks or 21 ~ after 26 weeks. Light, in the absence of oxygen, did not promote carotenoid degradation. Carotenoid concentrations declined linearly with time (25 ~o-62 ~o lost by 37 weeks) in a dark environment exposed to the atmosphere at 6 ~C, but only if algae were exposed to lake sediments or water containing natural, lacustrine bacterial populations. No distinct difference between the rates of carotene and xanthophyll destruction was noted, although myxoxanthophyll was more labile than other cyanophyte carotenoids, especially at 21 ~ (85~o loss by 26 weeks). Based on these results and published descriptions of algal decomposition kinetics, I conclude that; 1)the high sedimentary carotenoid concentrations typical of productive lakes may reflect a preferential loss of nonpigmented organic matter and, 2) sedimentary bacterial activity alone may not affect the reliability of carotenoids as paleolimnological indicators of past algal abundances in lakes with completely anoxic sediments.

Introduction

Fossil carotenoids have long been used as biochemical markers of photosynthetic prokaryotes (Ztlllig, 1961; Brown & Colman, 1963). However, interpretation of sedimentary pigment stratigraphies has been hampered because the factors that produce the observed patterns of abundance are poorly understood. Effects of lake trophic status on the relative preservation of different pigments may be particularly important (Swain, 1985).

A number of processes may degrade pigments prior to deposition in sediments: herbivore grazing (Repeta & Gagosian, 1982; Carpenter & Bergquist, 1985), photo-oxidation (Welschmeyer & Lorenzen, 1985; Carpenter etal., 1986, 1988), and cell senescence (Daley & Brown, 1973a). Bacterial decomposition of algae affects carotenoid concentrations under experimental conditions (Fox et al., 1944; Cranwell, 1976a) and may modify pigment levels both in the water column and in the sediments. The specific pigment content of lacustrine sediments may also be in-

216 fluenced by the initial pigment concentrations in the sedimenting material or by the influx of pigment-poor aUochthonous material (Swain, 1985). Although both phytoplankton decay (Depinto, 1977 and references within) and pigment degradation (Sanger & Gorham, 1970) occur during algal sedimentation, post-depositional carotenoid diagenesis may also affect fossil carotenoid concentrations (Fogg & Belcher, 1961; Moss, 1968, but see Brown, 1968). I report the use of reversed-phase thin layer chromatography to quantify carotenoid degradation resulting from in vitro decomposition of Oscillatoria utermOhlii (Utermohl) J. de Toni. Oscillatoria species often form deep-water blooms that are represented in lake sediments by the characteristic carotenoids myxoxanthophyll and oscillaxanthin (Brown & Colman, 1963; Griffiths, 1978; Ztlllig, 1981, 1982). Because these filamentous cyanophytes frequently dominate the phytoplankton assemblages of eutrophic lakes (e.g. Edmondson, 1968), the presence of myxoxanthophyll or oscillaxanthin in lacustrine deposits has been interpreted as a signal of increased productivity (Ztlllig, 1981, 1982; Swain, 1985). This paper examines the effects of light, oxygen, temperature and the presence of a lacustrine bacterial flora on the rate and extent of cyanophyte carotenoid degradation.

Materials and methods Oscillatoria utermrhlii were grown in 1.0 1cultures of half-strength mineral medium defined by Daley & Brown (1973b). Cultures were maintained at 19 ~ + 2 ~ in continuous fluorescent light (165/~E m - 2 s - 1). Algae from 144 hold cultures were harvested by centrifugation (3000 g, 10 min., 10 ~C). The concentrated suspension was lyophilized and stored in the dark at - 2 0 ~ under a nitrogen atmosphere until use. Lyophilized cells were used to prevent changes in carotenoid concentration that result from changes in light and nutrient regimes (Goodwin, 1980). Nine experiments were performed to determine the effects of bacterial decomposition, oxygen and

light on the carotenoids of O. utermrhlii (Table 1). Experiments containing lake mud and monimolinmetic water simulated sedimentary conditions both under anaerobic conditions (closed trials) and in the presence of oxygen (open). Experiments with monimolimnetic water alone controlled for the presence of a lacustrine sedimentary flora, whereas sulfide solution approximated in situ redox conditions, while removing the influence of natural microbial populations. Room temperature sulfide experiments allowed quantification of the influence of temperature on carotenoid degradation. Homogenized suspensions of lyophilized O. utermghliiin either monimolimnetic lake water or a solution of 20 mg 1- 1S = (as Na2 S 9H2 O) in deionized water were prepared and loaded into 25 ml glass vials under a nitrogen atmosphere and indirect lighting. Where appropriate, vials also contained a known volume of anaerobic lake mud, obtained from meromictic Little Round Lake, Ontario (44 ~ 47' N, 76 ~ 41' W) (Table 1). Closed series vials were capped and sealed with paraffin wax, whereas open vials were not. Sample vials were placed in the dark at the appropriate temperature, except for those of the sulfide-light experiment which received a continuous fluorescent illumination of 40/~E m - 2 s - 1 . Open series vials were shaken twice a week to aid in oxygen penetration and were made up to 24 ml with distilled water biweekly to compensate for water evaporation. Triplicate vials were removed and sampled at each of 15 times over 37 weeks for each cold (6 ~C) experiment and 11 times over 26 weeks for room temperature trials ( ~ 21 ~ Contents of each vial were suspended and temperature, pH and redox potential (Eh) were determined. Measurements of pH were made using a glass combination electrode. Redox determinations were made after a 5 minute equilibration period with a polished, fiat-surface platinum electrode and a saturated calomel reference electrode. All Eh values were normalized to pH 7.0 (Cole, 1975) and corrected to 25 ~ (Skoog & West, 1976). Slopes of Model I regressions of carotenoid concentration versus time were tested for the null hypothesis that the decay slope was zero using the

217 Table 1. Carotenoid diagnenesis experiments. Standard deviation of sediment wet weight indicated in parentheses. Experimental series

Wet weight sediment (g)

Dry weight algal suspension (mg/ml)

Volume algal suspension (ml)

[S =] (mg/1)

Temp ( ~C)

Sediment closed Sediment open Monimolimnetic closed Monimolimnetic open Sulfide closed Sulfide open Sulfide light** Sulfide (room) closed Sulfide (room) open

2.021 (0.140) 1.838 (0.070) -

0.870

22.0

?

0.870

22.0

?

0.870

24.0

15-20*

-

0.870

24.0

15-20*

-

1.111

24.0

20.0

-

1.111

24.0

20.0

-

1.111

24.0

20.0

-

1.000

24.0

20.0

-

1.000

24.0

20.0

6.0 + 1.0 6.0 + 1.0 6.0 + 1.0 6.0 _+ 1.0 6.0 + 1.0 6.0 _+ 1.0 6.0 _+ 1.0 21.0 + 4.0 21.0 + 4.0

* Estimated from McNeely (1973). * * 4 0 / ~ E m - 2 s 1.

Statistical Analysis System (SAS) computer package for the determination of homogeneity-ofslopes (Ray, 1982a, 1982b).

Extraction and pigment quantification

Samples were quantitatively transferred to glass centrifuge tubes, centrifuged (1400g, 30min, 10 ~ and decanted. Pellets were suspended in a known volume of acetone/methanol (84/16, by vol) by mild ultrasonic disruption for 1 min. Samples containing lake sediments were suspended in 15 ml of the solvent mixture, cooled to 1 ~ and disrupted by sonication following the procedures of Daley et al. (1973a). Extracts from exhaustive extraction of each sample were combined and stored at - 4 ~C, in the dark and under a nitrogen atmosphere until pigment analysis. All other samples were disrupted for 1 min in 5 ml of the extraction mixture and allowed to stand under nitrogen gas, in the dark, at room temperature for 6 h before being sealed and stored.

Reversed-phase thin layer chromatography was used for carotenoid separation following the procedures of Leavitt & Brown (1988). Handmade chromatographic plates were prepared with MN-kieselgur, impregnated with either light petroleum, castor or triolein oil and were spotted and developed in the dark. Pigment quantification was immediately carried out using a modification of the procedures of Daley et al. (1973b). Densitometric scanning of the developed chromatograms was performed using a Turner 111 filter fluorometer equipped with a Turner Model II TLC scanner door, a Hamatsu HTV-136 photomultiplier, a Balazers K-3 broad range primary filter and a Balazers B-40 narrow pass secondary filter. The densitometer was calibrated using a dilution series of each chromatographically pure carotenoid developed on the appropriate chromatographic system. The amount of pigment in each standard was determined spectrophotometrically (Davies, 1976). The specific extinction coefficients were obtained from Davies (1976) and Foppen (1971). Carotenoid identifications

218 reducing conditions ( - 150 mY) were produced through algal decomposition and samples smelled strongly of hydrogen sulfide within 4-6 weeks. Final redox potentials produced in the closed sulfide experiments were similar to those of the corresponding sediment and monimolimnetic trials, although both the time required for maximally reducing conditions and the minimum redox potential produced were variable. Despite the diffusion of atmospheric oxygen into the open vials, final redox potentials (0.0 mV) indicated a reducing environment, although less so than in the corresponding closed series vials. Room temperature sulfide trials achieved reducing conditions more rapidly than did corresponding 6 ~C experiments, but final redox potentials did not differ. The pH in the closed vials fell rapidly, concurrent with the production of reducing condi-

were based on comparisons of the spectral characteristics of the pigments and their LiAIH4 reduction products with published values and by cochromatography with carotenoids isolated from eight other blue-green algae(Hertzberg et aL, 1971).

Results

Eh, pH Initial redox potentials (Fig. 1) in the sediment and monimolimnetic experiments were indicative of moderately oxygenated lake sediments (Hutchinson, 1957) indicating that algal suspensions may have become oxidized during the vial loading procedure. In the closed trials of both series, highly A. SEDIMENT Eh

B. MONIMOLIMNION Eh

200"

400 I~ 200 -4IW

0"

10

.,oo] E

-2000

.,oo1 \ -,001 g 1'0

2'0

3'0

40

-300 I 0

-

, 10

, 20

30

10

20

30

D. ROOM Eh

C, SULFIDE Eh

100:

o

z_

~

_~

-100 -200

-200

-300'

-,,oo] "r-lr .-6O0

0

,

i

10

i

20

30

40

-400: I -soo: -600 0

TIHE (WEEKS) Fig. 1. Changes in the oxidation-reduction potentials during carotenoid degradation experiments. Each point is the mean of triplicate determinations. Standard deviations of the mean are indicated for each sample. [] = closed, 9 = open, and + = light experiments.

219 B. MONIMOLIMNETIC pH

A. SEDIMENT pH

81

8"

6

i

0 ~-

10 C. SULFIDE pH

i

20

]

J

30

0

40

l

10 D. ROOM pH

i

20

30

40

9 8 7'

7 6 5

J

0

10

i

20

i

6

30

40

0

10

20

30

TIME (WEEKS) Fig. 2. Changes in pH during the carotenoid degradation experiments. Symbols, means and standard deviations as in Fig. 1.

tions (Fig. 2). Regardless of the initial values, the final pH of all closed suspensions was approximately 6.4. The initial pH of the water in open series vials also declined slightly at the onset, however this period was followed by one in which pH increased to stable values of 7.4-7.6

Carotenoid degradation The patterns of pigment loss in the sediment (Fig. 3), monimolimnetic (Fig. 4) and sulfide experiments, both at 6 ~ (Fig. 5) and room temperature (Fig. 6), show that the rates of carotenoid degradation were low. Declines in pigment levels were statistically significant (P < 0.05) in only nine instances and with the exception of myxoxanthophyll degradation in the closed monimolimnetic trial (30 ~o loss by 37 weeks), all of the

significant carotenoid losses occurred in the open experiments (Table 2). Generally, myxoxanthophyll was the most labile pigment, although there was some variability in the extent of carotenoid loss. For instance, the concentration of zeaxanthin was reduced 62~o over the course of the open sediment experiment, while myxoxanthophyll and echinenone concentrations declined 41~ and 39~, respectively (Fig. 3). /~-carotene loss (11~o) was not statistically significant (P < 0.05). In the open monimolimnetic experiment, however, myxoxanthophyll showed the greatest loss (54 ~o) while zeaxanthin and/~-carotene concentrations were reduced to a lesser extent (42 ~o and 25 ~ , respectively, Fig. 4). No significant carotenoid degradation occurred in the 6 ~ sulfide experiments (Fig. 5). Light, in the absence of oxygen, does not appear to influence the rate of carotenoid destruction,

220 A. SEDIMENT-CLOSED

60 50-

40 j 30--I

<

I-. 2: uJ

20 10

:E

0

o

!

i

|

10

20

30

40

20

30

40

B. SEDIMENT-OPEN

40 302010, o 0

10

TIME (WEEKS) Fig. 3. Changes in carotenoid concentration of the closed and open sediment experiments. Means and standard deviations as in Fig. 1. 9 = B-carotene, 9 = echinenone, + = zeaxanthin and [] = myxoxanthophyll.

although increased temperature accelerated the degradation of some carotenoids in the open sulfide trial (Fig. 6). Despite the shorter experimental period, the losses of myxoxanthophyll (85 ~o over 26weeks) and echinenone (73~) were greater than the losses of the carotenoids from the cold experiments. Cis-stereoisomers were not separated from their native all-trans carotenoids on the reversedphase chromatograms and, therefore, it is not possible to determine if stereoisomerization is an important step in carotenoid degradation. Other coloured carotenoid derivatives (e.g. epoxides) were not detected on the chromatograms.

Discussion

Eh, pH Redox potential changes in the closed series experiments conducted at 6 ~ (Fig. 1) indicate anaerobic conditions and are within the range of values reported for lacustrine sediments (Hutchinson, 1957; Baas Becking et al., 1960). Variability in the length of time required to produce stable, low ( - 50 mV) potentials probably reflects differences in the bacterial innoculum size and metabolic activity, as influenced by experimental temperature (Hargrave, 1972; Foree & McCarty, 1970; Depinto, 1977). The reductions of pH in closed series experiments to values of 6.4 probably result from the

221 A. MONIMOLIMNETIC-CLOSED 50 40 30 ,-I

20 10

I-Z UJ

=E 0

0 0

Q.

10 20 B. MONIMOLIMNETIC-OPEN

30

40

30

40

50

30

10 0 0

10

20 TIME (WEEKS)

Fig. 4. Changesin carotenoidconcentrationsof closed and open monimolimneticexperiments.Symbols,means, and standard

deviations as in Fig. 3.

release of organic acids from decomposing algae and the dissolution of C O 2 into, and combination with, water (Otsuki & Hanya, 1972a). The patterns o f p H change in all open experiments are also similar to those noted for the aerobic decomposition of Scenedesmus by Otsuki & Hanya (1972b).

Carotenoid degradation

Carotenoid degradation proceeds most rapidly in treatments open to atmosphere. The enhancement of pigment degradation by oxygen may be a direct chemical action (oxidation) or mediated through bacterial processes (increased bacterial

metabolism, establishment of different microbial populations). In vitro studies show that molecular oxygen can act directly to degrade carotenoids (EI-Tinay & Chichester, 1970; Davies, 1976), however the importance of molecular oxygen in in vivo systems is less well known. Carotenoid degradation in higher plant tissues begins with the formation of detectable quantities of 1,2- or 5,6-epoxy-carotenoids (Simpson et al., 1970), using an oxygen atom from molecular oxygen rather than water (Yamamoto & Chichester, 1965). A similar mechanism may be expected to be involved during algal decomposition. In this study, two lines of evidence indicate that oxygen is acting on the carotenoids through an

222 A. SULFIDE-CLOSED 80 60 40 20 0

~ 0

10

20

30

40

20

30

40

B. SULFIDE-OPEN

80 ,-I

,< >

60

I-U.I

40

:S m

(:l. O)

0 0

10 C. SULFIDE-LIGHT

80 60 40 20 0 0

10

20

30

40

TIME (WEEKS) Fig. 5. Changes in carotenoid concentrations in the closed, open and light sulfide experiments. Symbols, means and standard deviations as in Fig. 3.

223

A. ROOM-CLOSED

so t 60 : 40 20 .-I

_< 00 Z LU

'0

2'0

30

20

30

B ROOM-OPEN

=i a.

1

80 60 40 2O 0 0

10

TIME (WEEKS) Fig. 6. Changes in carotenoid concentrations in the closed and open room temperature sulfide experiments. Symbols, means and standard deviations as in Fig. 3.

enhancement of the rates of microbial activity rather than a direct chemical reaction. First, if molecular oxygen initially reacts with the carotenoids to form 1,2- or 5,6-epoxy-carotenoids, then these pigments should be chromatographically separable from the parent carotenoids. Addition of an epoxide group to an unsubstituted carotenoid will increase the pigment's polarity only slightly less than if a single keto group had been added (Liaaen-Jensen, 1971). All chromato-

graphic systems were capable of separating pigments that differ in the presence of a single keto function and therefore provide sufficient resolution to detect epoxide formation. Thus, the absence of detectable quantities of coloured carotenoid derivatives suggests either that molecular oxygen does not degrade the carotenoids directly, or that the mechanisms of algal pigment degradation differ from those of higher plants. Second, if the oxygen were to react directly with

224 Table2. Slopes o f regressions o f carotenoid concentration versus time and, within parentheses, relative (~o) pigment lost by the end of the experiment. Asterisks indicate slopes significantly different from zero (P < 0.05). Slope =/~g pigment w k - 1. Pigment ~-carotene

Echinenone

Zeaxanthin

Myxoxanthophyll

Sediment closed Sediment open Monimolimnetic

0.0075 (-) - 0.0116 (10.6) - 0.0025

- 0.0058 (17.2) - 0.0126" (38.9) - 0.0083

- 0.0085 (12.0) - 0.0471" (61.7) 0.000

- 0.0083 (24.8) - 0.0178" (40.5) - 0.0139*

closed Monimolimnetic open Sulfide closed Sulfide

(2.5) - 0.0263* (25.0) 0.0103 (-) - 0.0155

(21.5) - 0.0088 (22.9) - 0.0124 (22.5) - 0.0142

(0.0) - 0.0081" (41.9) - 0.0008 (2.8) - 0.0008

(30.3) - 0.0259* (54.1) - 0.0008 (3.0) - 0.0065

open Sulfide light R o o m (21 ~ closed R o o m (21 ~ open

(8.3) - 0.0004 (0.3) - 0.0283 (16.1) - 0.0703 (34.1)

(23.3) - 0.0066 (11.5) - 0.0130 (38.6) - 0.0214" (51.4)

(3.8) - 0.0005 (1.8) - 0.0061 (8.9) - 0.0132 (17.0)

(10.5) - 0.0058 (10.4) - 0.0144 (37.1) - 0.0402* (85.3)

the carotenoids, one would expect significant amounts of pigment diagenesis in all open trials. Instead, none of the carotenoids of the cold, open sulfide experiments show a statistically significant reduction in concentration (Fig. 5). However, if the presence of low levels of oxygen results in an increased rate of bacterial degradation of algal cells, its effects would be most noticeable in experiments with high initial bacterial populations. The fact that both sediment and monimolimnetic open experiments (Figs. 3, 4) show significant losses of carotenoids supports this hypothesis. Light in the absence of oxygen does not promote carotenoid degradation. This result was expected because Krinsky (1979a, b) demonstrated that the photo-oxidation of carotenoids in vitro requires the presence of both oxygen and light. Carotenoids are bleached when either singlet, excited state oxygen molecules or peroxides are added across the pigment's polyene chain (Krinsky & Deneke, 1982). Increased temperature accelerates the degradation of some carotenoids when exposed to oxygen (Fig. 6). The losses of myxoxanthophyll

(85~o by 37 weeks) and echinenone (73~o) are substantially greater than the losses of pigments from the 6 ~ experiments.

Paleolimnological implications Reductions in carotenoid concentrations in the closed experiments are not significant (P > 0.05), with the exception of myxoxanthophyll loss in the monimolimnetic trial (30~/o by 37 weeks). The reason for the decline in myxoxanthophyll concentration is unclear. Other researchers have suggested that highly substituted xanthophyll carotenoids are more labile than the carotene pigments (Fogg & Belcher, 1961; Brown, 1969; Ztlllig, 1982). In my experiments, myxoxanthophyll is generally more susceptible to degradation than are the other carotenoids. Because myxoxanthophyll has proven to be a useful paleolimnological indicator of past blue-green algal populations (Griffiths, 1978; Ztlllig, 1981; Swain, 1985), such a preferential degradation may lead to underestimation of the cyanophyte contribution

225 to lacustrine primary production, especially in instances where the surficial sediments are exposed to oxygen or where dead algal cells remain suspended in the oxic zone of the lake for extended periods of time. The overall lack of carotenoid degradation under anaerobic conditions has other paleolimnological consequences. Algal decomposition under similar experimental conditions has been described by first order decay kinetics (Force & McCarty, 1970; Depinto, 1977) and is complete within 200 days (Force & McCarty, 1970; Otsuki & Hanya, 1972b; Godshalk & Wetzel, 1977). Although the rate of algal mineralization is dependent on bacterial inoculum size and experimental temperature, the final extent of particulate matter loss is independent of these factors and appears to be similar regardless of the species of algae being decomposed (Force & McCarty, 1970; Cranwell, 1976a, b). Minimum estimates indicate that at least 45 ~ of the organic matter is degraded within 28 weeks, even at low temperatures (Cranwell, 1976b). Because carotenoid degradation in closed trials was usually less than 25 ~o, my study implies that both the rate and extent of microbially-mediated carotenoid degradation are lower than those of algal organic matter, under anaerobic conditions. A similar conclusion may be drawn from Fox et al. (1944) and Cranwell (1976a) who demonstrated that carotenoids were degraded at rates only one-half that of algal particulate organic matter under a variety of experimental conditions. The basis for carotenoid stability is unclear but may be related to the insolubility of the pigments in water (Vallentyne, 1960). Several investigators have assumed that variations in the pigment content of the surficial sediments of lakes of differing trophic status result from the degradation of the pigments relative to the autochthonous organic matrix or dilution of these pigments by the influx of pigment-poor allochthonous material (Gorham & Sanger, 1975; Swain, 1985). In contrast, this study suggests that the reverse process may also occur. In eutrophic lakes where profundal sediments are anoxic, the organic matrix may degrade more rapidly than the carotenoids. In oligotrophic lakes, where anoxic

conditions are less common, surficial sediments would be expected to have disproportionately low pigment concentrations. In such a scenario, sedimentary carotenoid levels, expressed as pigment influx, may quantitatively reflect former abundances of photosynthetic organisms which form deep blooms (Oscillatoria, photosynthetic bacteria) or otherwise rapidly sink out of the oxic zone.

Conclusions

This study demonstrates that under anaerobic experimental conditions microbially-mediated decomposition of Oscillatoria utermrhlii does not result in statistically significant declines in the levels of most carotenoids. Light in the absence of oxygen does not cause carotenoid degradation. In an environment open to gas exchange, individual carotenoids exhibit differing degrees of degradation, with myxoxanthophyll generally being the most labile carotenoid. The presence of oxygen alone appears to be insufficient to lead to carotenoid loss but may stimulate pigment degradation either through an enhancement of bacterial activity or through establishment of different species of bacteria that have differing degradative capabilities. This study suggests that anaerobic sedimentary bacterial activity alone may not affect the reliability of carotenoids as paleolimnological indicators of former algal abundances in either meromictic or productive dimictic lakes. However, before fossil carotenoids may be confidently used in such a capacity, further experimentation should be conducted to determine if there are differences between species in the rates of loss of individual carotenoids, as suggested by Cranwell (1976a). Additional information is required on the relative importance of carotenoid losses resulting from algal autolysis and photo-oxidative pigment destruction. With a better knowledge of these factors, it should be possible to establish the reliability of fossil carotenoids as indicators of past algal community dynamics.

226

Acknowledgements This research was supported by grants to S. R. Brown from the National Research Council of Canada (Grant A-808) and by an Ontario Graduate Scholarship to P. R. Leavitt. Manuscript preparation was supported by National Science Foundation grant B SR-85-21832. I thank Dr. S. R. Brown, Dr. S. R. Carpenter, Dr. D. M. Lodge, J. J. Elser and M. B. Berg for helpful criticism of this manuscript.

References Bass Becking, L. G. M., I. R. Kaplan & D. Moore, 1960. Limits of the natural environment in terms of pH and oxidation-reduction potential. J. Geol. 68: 243-284. Brown, S. R., 1968. Bacterial carotenoids from freshwater sediments. Limnol. Oceanogr. 13: 233-241. Brown, S. R., 1969. Paleolimnological evidence from fossil pigments. Mitt. int. Ver. Limnol. 17: 95-103. Brown, S. R. & B. Colman, 1963. Oscillaxanthin in lake sediments. Limnol. Oceanogr. 8: 352-353. Carpenter, S.R. & A.M. Bergquist, 1985. Experimental tests of grazing indicators based on chlorophyll a degradation products. Arch. Hydrobiol. 102: 303-317. Carpenter, S. R., M. M. Elser & J. J. Elser, 1986. Chlorophyll production, degradation, and sedimentation: Implications for paleolimnology. Limnol. Oceanogr. 31: 112-124. Carpenter, S. R., P.R. Leavitt, J.J. Elser & M.M. Elser, 1988. Chlorophyll budgets: Response to food web manipulations. Biogeochemistry 6: 79-90. Cole, G. A., 1975. Textbook of limnology. Mosby, St. Louis; 283 pp. Cranwell, P. A., 1976a. Decomposition of aquatic biota and sediment formation: lipid components of two blue-green algal species and the detritus resulting from microbial attack. Freshwat. Biol. 6: 481-488. Cranwell, P. A., 1976b. Decomposition of aquatic biota and sediment formation: organic compounds in detritus resulting from microbial attack on the algae Ceratium hirundinella. Freshwat. Biol. 6: 41-48. Daley, R. J. & S. R. Brown, 1973a. Experimental characterization of lacustrine chlorophyll diagenesis. I. Physiological and environmental effects. Arch. Hydrobiol. 72: 277-304. Daley, R. J. & S. R. Brown, 1973b. Chlorophyll, nitrogen and photosynthetic patterns during growth and senescence of two blue-green algae. J. Phycol. 9: 395-401. Daley, R. J., C. B. J. Gray & S. R. Brown, 1973a. A quantitative, semiroutine method for determining algal and sedi-

mentary chlorophyll derivatives. Can. J. Fish aquat. Sci. 30: 345-356. Daley, R. J., C. B. J. Gray & S. R. Brown. 1973b. Reversedphase thin layer chromatography of chlorophyll derivatives. J. Chromatogr. 76: 175-183. Davies, B. H., 1976. Carotenoids. In T. W. Goodwin (ed.), Chemistry and biochemistry of plant pigments. Vol. II. Academic Press, N.Y.; 38-165. Depinto, J. V., 1977. Water column death and decomposition of phytoplankton: an experimental and modelling review. In Scavia, D. & A. Robertson (eds.), Lake ecosystem modelling. Ann Arbor Science, Ann Arbor; 25-52. Edmondson, T.W., 1968. Water-quality management and lake eutrophication: the Lake Washington case. In Campbell, T.H. & R.O. Sylvester (eds.), Water resources management and public policy. University of Washington Press, Seattle; 139-178. E1-Tinay, A.H. & C.O. Chichester, 1970. Oxidation of fl-carotene. Site of initial attack. J. Org. Chem. 35: 2290-2293. Fogg, G. E. & J. H. Belcher, 1961. Pigments from the bottom deposits of an English lake. New Phytol. 60: 129-138. Foppen, F. H., 1971. Tables for the identification of carotenoid pigments. Chromatogr. Rev. 14: 133-298. Force, E. G. & P. L. McCarty, 1970. Anaerobic decomposition of algae. Envir. Sci. Technol. 4: 842-849. Fox, D. L., D. M. Updegraff & G. D. Novelli, 1944. Carotenoid pigments in the ocean floor. Arch. Biochem. 5: 1-23. Godshalk, G. L. & R. G. Wetzel, 1977. Decomposition of macrophytes and the metabolism of organic matter in sediments. In Golterman, H. L. (ed.), Interactions between sediments and freshwater. Junk, the Hague; 258-264. Goodwin, T. W., 1980. The biochemistry of the carotenoids. Vol. I. Plants. Chapman and Hall, N.Y.; 377 pp. Gorham, E. & J. Sanger, 1975. Fossil pigments in Minnesota lake sediments and their bearing upon the balance between terrestrial and aquatic inputs to sedimentary organic matter. Verh. int. Ver. Limnol. 19: 2267-2273. Griffiths, M., 1978. Specific blue-green algal carotenoids in the sediments of Esthwaite Water. Limnol. Oceanogr. 23: 777-784. Hargrave, B. T., 1972. Aerobic decomposition of sediment and detritus as a function of particle surface area and organic content. Limnol. Oceanogr. 17: 583-596. Hertzberg, S., S. Liaaen-Jensen & H.W. Siegelman, 1971. The carotenoids of blue-green algae. Phytochemistry 10: 3121-3127. Hutchinson, G. E., 1957. A Treatise on limnology. Vol. I. Geography, physics and chemistry. Wiley & Sons, N.Y.; 1015 pp. Krinsky, N. I., 1979a. Carotenoid protection against oxidation. Pure Appl. Chem. 51: 649-660. Krinsky, N. I., 1979b. Carotenoid pigments: multiple mechanisms for coping with the stress of photosensitized oxidations. In Shilo, M. (ed.), Strategies of microbial life in extreme environments. Dahlem Korterenzen, Berlin; 163-177.

227 Krinsky, N. I. & S. M. Deneke, 1982. Interactions of oxygen and oxyradicals with carotenoids. J. Nat. Cancer Inst. 69: 205-210. Leavitt, P. R. & S.R. Brown, 1988. Effects of grazing by Daphnia on algal carotenoids: Implications for paleolimnology. J. Paleolimn. 1: in press. Liaaen-Jensen, S., 1971. Isolations, reactions. In Isler, O. (ed.), Carotenoids. Birkhausser Verlag, Basel; 61-188. McNeely, R. J., 1973. Limnological investigation of a small meromictic lake, Little Round Lake, Ontario, 1968-1970. Ph.D. Thesis, Queen's University, Kingston, Ontario. Moss, B., 1968. Studies on the degradation of chlorophyll a and carotenoids in freshwaters. New Phytol. 67: 49-59. Otsuki, A. & J. Hanya. 1972a. Production of dissolved organic matter from dead green algal cells. II. Anaerobic microbial decomposition. Limnol. Oceanogr. 17: 258-264. Otsuki, A. & J. Hanya, 1972b. Production of dissolved organic matter from dead algal cells. I. Aerobic microbial decomposition. Limnol. Oceanogr. 17: 248-257. Ray, A. A., 1982a. SAS user's guide: basics. SAS Institute, Cary. Ray, A. A., 1982b. SAS user's guide: statistics. SAS Institute, Cary. Repeta, D. J. & R. B. Gagosian, 1982. Carotenoid transformations in coastal marine waters. Nature 295: 51-54. Sanger, J. E. & E. Gorham, 1970. The diversity of pigments in lake sediments and its ecological significance. Limnol. Oceanogr. 15: 59-69.

Simpson, K. L., T. Lee, D. B. Rodriguez & C. O. Chichester, 1976. Metabolism and senescence in stored tissue. In Goodwin, T.W. (ed.), Chemistry and biochemistry of plant pigments. Vol. I. Academic Press, N.Y.; 779-842. Skoog, D. A. & D. M. West, 1976. Elements of analytical chemistry. Holt, Rinehart & Winston, N.Y.; 804 pp. Swain, E. B., 1985. Measurement and interpretation of sedimentary pigments. Freshwat. Biol. 15: 53-75. Vallentyne, J. R., 1960. Fossil pigments. In M. B. Allen (ed.), Comparative biochemistry of photoreactive systems. Academic Press, N.Y.: 83-105. Welschmeyer, N.A. & C.J. Lorenzen, 1985. Chlorophyll budgets: zooplankton grazing and phytoplankton growth in a temperate fiord and the Central Pacific Gyres. Limnol. Oceanogr; 30: 1-21. Yamamoto, H. Y. & C. O. Chichester, 1965. Dark incorporation of 180 2 into antheraxanthin by bean leaf. Biochem. Biophys. Acta 109: 303-305. Z~llig, H., 1961. Die Bestimmung yon Myxoxanthophyll in Bohrprofilen zum Nachweis vergangener Blaualgenentfaltungen. Verh. int. Ver. Limnol. 14: 263-270. Zfillig, H., 1981. On the use of carotenoid stratigraphy in lake sediments for detecting past developments of phytoplankton. Limnol. Oceanogr. 26: 970-976. Zfillig, H., 1982. Untersuchugen fiber die Stratigraphie von Carotinoiden im geschichteten Sediment yon 10 Schweizer Seen zur Erkundung frfiherer Phytoplankton-Entfaltungen. Schweiz. Z. Hydrol. 44: 1-98.