EXPORT PRODUCTION DEPTH OF EMILI ANI A HUXLEY/IN THE GULF OF

CALIFORNIA: AN EVALUATION OF THE ALKENONE UK'37 PALEOSST-PROXY

A THESIS SUBMITTED TO THE GRADUATE DIVISION OF THE UNIVERSITY OF HAWAI'IIN PARTIAL FULFILLMENT

OF THE REQUIREMENTS FOR THE DEGREE OF

MASTER OF SCIENCE

IN

OCEANOGRAPHY

DECEMBER 2007

By Amanda S. Pontius

Thesis Committee:

Brian Popp, Chairperson Fred Mackenzie Robert Bldigare

We certify that we have read this thesis and that, in our opinion, it is

satisfactory in scope and quality as a thesis for the degree of Master

of Science in Oceanography.

THESIS COMMITTEE

ii

iii TABLE OF CONTENTS

List of Tables ........................................................................................ iv List of Figures ....................................................................................... v Chapter 1: Introduction ........................................................................... 1 Chapter 2: Samples and Methods ............................................................ 12

Sampling site .............................................................................. 12 Suspended particulate material samples .......................................... 12 Free floating arrays ..................................................................... 13 Sediment trap samples ................................................................. 14 Analytical techniques .................................................................... 14

Chapter 3: Results and Discussion ........................................................... 17 Water column properties ............................................................... 17 Depth of alkenone export .............................................................. 17 Depth of maximum alkenone concentration and production .................. 29 Comparison of UK

'37 and in situ temperatures .................................... 38 Chapter 4: Conclusion ........................................................................... 44 References ......................................................................................... 46

iv LIST OF TABLES

1. Alkenone Temperature Calibrations Compiled from the Literature ........... 3

2. Photosynthetic Active Radiation (PAR) in the Gulf of Califomia ............... 18

3. UK'37 and li37:2 Pattems in the Gulf of Califomia ................................... 22

4. Alkenone Export in the Gulf of Califomia .......................................... 27

5. K:!7:2 Alkenone Concentration in the Gulf of Califomia ......................... 30

6. K:!7:2 Production in the Gulf of California ........................................... 34

7. A Comparison of Export Depth to the Depths of Maximum Concentration and Production in the Gulf of California ........................ 36

8. Integrated K:!7:2 Concentration and Production in the Gulf of California .................................................................................... 37

9. A Comparison of UK'37 Predicted Temperature with in situ Temperature ............................................................................. .41

v LIST OF FIGURES

Figure

1. Muller et al.. 1998 Core-top Dataset.: .............................................. 3

2. Map ofthe Gulf of Califomia .......................................................... 7

3. Goni et al.. 2001 Time-series Dataset... ........................................... 8

4. Water Column Profiles ................................................................ 19

K' 5. Depth Profiles of U 37 and 5i<37:2 ................................................... 20

6. Depth Profiles of Alkenone i<37:2 Concentration ................................. 32

7. Depth Profiles of i<37:2 Production Rates .......................................... 32

8. UK'37 vs. in situ temperatures ......................................................... 39

9. i<37:2 Photoperiod Growth Rate (P) vs. available light (PAR) ................. 41

CHAPTER 1 INTRODUCTION

1

Since their discovery and identification in sediments collected from Walvis

Ridge off West Africa [Boon et al. 1978] and the Black Sea [deLeeuw et al .•

1980]. alkenones (Car-Cas di-. trio. and tetraunsaturated methyl and ethyl

keytones) have been found to possess many characteristics that lend themselves

to be used as biomarkers. A1kenones in the open ocean are thought to be

produced exclusively by members of the Class Haptophyceae; most notably the

cosmopolitan species Emiliania huxleyi and the closely related Gephyrocapsa

oceanica [Marlow et al .• 1984a. b]. Unlike the majority of biologically synthesized

compounds. the chemical structure of alkenones contains the 'trans' double bond

configuration. which may help to prevent biodegradation in sediments [Rechka

and Maxwell. 1988a. b]. Most importantly for paleoceanographic applications.

the relative abundance of the double bonds in the compound has been found to

vary linearly with growth temperature (gT) of the biosynthesizing algae in

laboratory cultures [Prahl and Wakeham. 1987; Prahl at al.. 1988] and in the field

[Conte and Eglinton. 1993; Temois at al. 1997]. Focusing on the Ca7 alkenones.

Brassel et al. [1986] defined an alkenone unsaturation index (UKa7) that is

calculated from the relative abundances of the C37 methyl alkenones containing

2-4 double bonds.

UK37 = [C37:21-[Ca7:4]I[C37:2 + C37:3 + Ca7:4] (1 )

Because the concentration of the C37:4 alkenone is low in waters >15°C and the

C37:4 alkenone is rarely detected in sediments. Prahl and Wakeham [1987]

simplified UK37 to UK'37

UK' 37 = [C37:2lI[C37:2 + C37:3] (2)

and used the calculated UK' 37 and gT from batch culture experiments

(temperature range 8-25°C) to create a calibration curve that could be used to

quantitatively reconstruct ancient mean annual sea-surface temperature

(maSST) from deep-sea cores.

UK'37 = 0.033T + 0.043 (3)

2

This equation was determined using a specific strain of E. huxlayi (strain 55a or

CCMP 1742). and was later refined by Prahl at al. [1988] (UK'37 = O.034T + 0.039;

R2 =0.994). Further studies have shown that UK' 37 response to temperature can

vary depending on the species and strain of algae used [Temois at a/ .• 1997;

Conte et al.. 1998; Versteegh at al .• 2001]. Such genotypic differences could

decrease the applicability of using alkenones as a paleo-proxy for maSST. To

examine this further. Muller at al. [1998] and more recently Conte et al. [2006]

compiled UK'37 data for alkenones isolated from surficial sediments (core-tops)

collected on cruises throughout the world ocean and compared it to the maSST

for the region (temperatures were determined from the respective archives of the

NOAA World Ocean Atlas 1994 and the COADS archive for Muller at al. [1998]

and from the updated Levitus 2001 compilation for Conte at al. [2006]. see the

respective texts for more details). Remarkably. the calibration curves obtained

Tabla 1. Alkenone Temperature Calibrations Complied from the Literature

Calibration f Sample TyPe

UK' 37 = O.033T + 0.043 0.994 Strain 55& cultura

UK' 37 = O.034T + 0.039 0.994 Strain 55& cultura

UK' 37 = O.033T + 0.044 0.958 Global Core-top

UK' 37 = O.033T + 0.045 0.97 Global Core-top

1.0

0.9

0.8

0.7

0.6

5OI:::l~ O.S

0.4

0.3

0.2 +

0.1

0.0 0

4+

i~"( ~ +

t() • + ~ n

+ " t + ,;

d.~ t tl" a

~o • " .. ~~ .f ' .~.~

o't"',...~""o 0>0-

+" , .../+ (>0"

<!-. -+,1'''4-

" 0 c +,$ < u ,. '"

....... '/.,0 Global Average

+ Atlantic Ocean o Indian Ocean

0° Q P8l:ific Ocean

5 10 IS 20 2S Annual Mean SST ('Ie)

References

Prahl and Wakeham [1987)

Prahl et aI. [1988)

Muller et al. [1998)

Conte et aI. [2006)

30

Figure 1. Relationship between UK' 37 and annual mean SST for surface sediments from the

global ocean batween 60"S and 60"N. The UK' 37 data were complied from the litereture (for

3

references see Muller et al. [1998)) and the temperature data are from the respective archives of

the NOAA World Ocean Atlas 1994 and the COADS archive. The obtained global core-top

calibration (UK' ~.033T + 0.044) Is statisticelly similar to the other 3 calibrations discussed In the

text (see Table 1) and shows an -±3"C scatter about the global average.

4

from these data sets (UK'37 = 0,033T + 0.044; R2 =0.958 for Muller et aI, [1998]

and UK'37 = 0,033T + 0.045; R2 =0.97 for Conte et al. [2006]) are almost identical

to the Prahl and Wakeham [1987] and Prahl et al. [1988] calibration equations

(Table 1). Muller et al. [1998] concluded that strain 55a is similar enough to the

alkenone producing populations of the open ocean and apparent genotypic

differences seen in some culture experiments are not significant enough to

impede paleo-SST reconstruction. Despite the high statistical significance of

these relationships, considerable scatter is apparent in each of the core-top data

sets (e.g., see Fig. 1), which raises concern for paleo-temperature

reconstructions using the alkenone unsaturation index.

One possible ecological explanation for the scatter in the core-top data

could be that the alkenone producing populations grow and record water

temperatures at depths other than the surface mixed layer (SML). In fact, E.

huxley; is physically capable of migrating to and growing in the deepest parts of

the euphotic zone. Knappertsbusch [1993] found that E. huxley; can form

macroaggregates by secreting a polysaccharide that causes the algal cells to

coagulate. This macroaggregate could sink into the nutricline and allow cells to

continue growing because E. huxley; contains the accessory pigment 19'­

hexanolyoxyfucoxanthin which enhances photosynthetic effectiveness in the

spectral range that penetrates deepest in the ocean [Haxo, 1985]. Field

evidence from various study sites such as the temperate northeast Pacific gyre

[Prahl et al., 1993], the Mediterranean Sea [Temois et al., 1997], the central

5

Pacific [Ohkouchi et al.. 1999]. and the Gulf of Maine [Prahl et al .• 2001] suggest

that the depth of alkenone export may be located beneath the SML. within a

deep chlorophyll maxima layer (DCML). at least during some periods of the year.

This ecological explanation could lead to an underestimate of SST for

reconstructions based on UK'37.

Scatter in the UK'3r-temperature relationship might also result from

physiological as opposed to ecological factors. Recent experiments by Epstein

et a/. [1998 and 2001] and Prahl et a/. [2003 and 2006] have shown that

physiological responses to nutrient and light availability can cause shifts in UK' 37

of ±0.13 (translates to approx. ±3°C). It has also recently been shown in field

experiments that algae growing in presumably nutrient limited waters in the

subtropical North Pacific contain alkenones that display a UK'37 index

systematically lower than expected based on any of the above-mentioned UK' 3r

temperature relationships. while algae growing deep in the water column under

lower light conditions contain alkenones that display a UK'37 index systematically

higher than in situ temperatures [Prahl et a/ .• 2005; Popp et a/ .• 2006a]. This

phenomenon may be related to the metabolic role that alkenones fulfill in the cell.

Originally it was believed that alkenones functioned as membrane lipids to

maintain fluidity as sterols do in other species [Brassell et a/ .• 1986]. but batch

culture experiments using 2 strains of E. huxlayi have shown that alkenones are

more likely energy storage molecules [Epstein at a/ .• 2001]. Traditional storage

lipids like triacylglycerols are conspicuously found only in very small quantities in

E. huxleyi, while alkenone concentrations are found in quantities similar to

triacylglycerol concentrations in other marine microalgae [pond and Harris,

6

1996]. As storage molecules, alkenone concentrations increase when the cells

are subjected to nutrient depleting conditions (e.g., at the end of a bloom)

because nitrogen and phosphorus are not available to biosynthesize proteins and

nucleic acids. On the other hand, alkenone concentrations decrease when cells

are subjected to low light levels (e.g., deep in the euphotic zone) because the

algae are burning the energy stored in the alkenones to continue growth with the

available nutrients [Prahl et a/., 2003]. Preferential degradation of the higher

energy C37:3 alkenones during light deprivation, and preferential biosynthesis of

the higher energy C37:3 alkenones during nutrient depletion could explain the

observed UK" 37/gT discrepancy observed in the laboratory [Epstein et a/. 1998;

2001; Prahl et a/. 2003; 2006] and in the field [Prahl et a/. 2005; Popp et al.

2006a].

An understanding of the growth conditions of E. huxleyi, especially in light

of the additional ecological and physiological factors that can affect UK" 37, could

help explain some of the observed variability in the core-top calibrations of Muller

et a/. [1998] and Conte et al. [2006] (Fig. 1) and could ultimately improve the

accuracy of the UK" 37 index as a paleoSST proxy. The study site chosen for this

investigation was in and around the Guaymas Basin, Gulf of California (Fig. 2)

due to its unique physical oceanographic environment that allows for a wide

range of sampling conditions. Because of seasonal changes in wind patterns,

-116" ." "

km -G :Ie 100

Figure 2. Map showing the location in the Gulf of California where the 4 sampling sites are

situated.

surface waters in the Gulf of California are characterized by stratification in the

summer and strong upwelling in the winter. These changes in wind stress result

in large seasonal changes in SST and in nutrient dynamics [Zineri and Thurnell,

2000; Thunell , 1998 and references therein). Surface temperatures can be as

low as 15°C during winter months due to the strong northwesterly winds that

blow down the axis of the Gulf and induce deep mixing of the water column. In

the summer, the wind pattern reverses, causing weak upwelling along the

western Gulf and strong thermal stratification in the eastern and central Gulf.

Sea surface temperatures during summer stratification can exceed 30°C while

nutrient levels drop significantly. Although total primary productivity in the

7

8

Guaymas Basin maximizes during upwelling and minimizes during stratification

[Thunell, 1998], sediment trap time series analysis from January 1996 to

September 1997 provided by Goni et al [2001] showed that alkenone flux

maximized during the summer and minimized in the winter.

A previous study in the Gulf of California [Goni et ai, 2001], which

collected alkenone concentration and UK' 37 values from surficial sediments and a

sediment trap deployed at 500 m from January 1996 to September 1997, found

that although the measured UK'37 values correlated well with SST measured

using Advanced Very High Resolution Radiometry (AVHRR) data for the majority

of the data set, at SST>26°C, values for UK'37 were systematically lower than

0 ., ~ .... • M

~

1 -;---"-'--~ • , !

~ A/ I

ae i • • • ae

07

05 ,. '6 18 ~ ~ ~ ~ ~ ~ ~ ~

AVHRR Tempem1ure (0 C)

Figure 3. Adapted from Gonl et aJ. [2001]. Plot of AVHRR temperature vs. UK' 37 ratios for

sediment trap samples. Line A represents the expression derived by Prahl et al. [1988] and Line

B represents the nonlinear fit to the data using a 31d order polynomial expression described in

Gonl et al. [2001].

9

expected based on the UK'3rtemperature relationship of Prahl et a/. [1988) (Fig.

3). This -3°C deviation from the UK'3rtemperature relationship is reminiscent of

the observed variability in the Muller et a/. [1998) and Conte et al. [2006) core-top

calibrations. Three possible explanations for this discrepancy were offered by

Goni et a/. [2001):

1) The physiological response to gT of the alkenone producing

community present in the Gulf of Califomia might not be the same as

documented for the E. huxleyi strain used to develop the calibration

equation of Prahl and Wakeham [1987). Conte et al. [1998) found

strains of E. huxleyi and G. oceanica, isolated from the Sargasso Sea

and SW Pacific respectively, which show a similar non-linear response

to growth temperature.

2) The community of alkenone producers might vary throughout the year.

Ziveri and Thumell [2000) documented seasonal changes in algal

community in the Gulf of California. Production flux was dominated by

E. huxleyi when SST <27°C, while the dominant flux contributor for

SST>2~C was G. oceanica. Even if both species have a linear

physiological response to gT, the UK'37 calibration equations need to be

statistically the same in order for accurate paleoreconstruction of

maSST from the underlying sediments.

3) The export depth of the alkenone producing community might not be

within the SML throughout the year. When the SST in the Gulf of

10

California is large. strong thermal stratification occurs in the euphotic

zone. UK·37 values that translate to a gT <SST could have resulted from

alkenone production occurring deeper in the water column. within the

thermocline and below the SML.

Taking into consideration the results of Epstein at a/. [1998 and 2001] and Prahl

at a/. [2003 and 2006]. a 4th explanation is now offered:

4) The UK' 37 values synthesized during the warm summer months might

underestimate gT by up to 3·C because of nutrient limitation. Previous

studies in the Gulf of California [Gaxiola-Castro at a/. 1999; White et al.

2007] found that nitrate concentrations are undetectable during the

summer in the SML.

In the present study. alkenone unsaturation patterns (UK'37). alkenone-specific

carbon isotopic variation (lSKa7:2). and seasonal alkenone concentration and

production rate are determined throughout the upper water column (0-40 m) in

the Gulf of California in order to better constrain the depth of alkenone export

production and to identify factors that might cause calculated UK·37 temperatures

to underestimate SST when SST>26°C. Field data were collected during one

wintertime (January - February 2005) and two summertime (July 2004 & July­

August 2005) cruises. Alkenone production rates were determined using an in

situ l3C incubation method utilized in a related study in the subtropical

oligotrophic North Pacific [Prahl at a/ .• 2005; Popp at sl .• 2006a]. which was

originally modified after that employed previously in the Sea of Japan [Hamanaka

11

at al., 2000] and the Bering Sea [Shin et a/., 2002]. Alkenone export production

was determined to originate from a depth of -20-30 m, which is well below the

SML during the summer. Exported UK'37 did underestimate SST by up to -3°C,

but this underestimate was not as large as expected based on the temperature

gradient between the SML and the depth of export, and probably resulted from a

concurrent physiological response to growth under light limiting conditions. As a

result, SST estimates made by measuring UK' 37 fall within the ±3°C variability

observed in the core-top calibrations of Muller at a/. [1998] and Conte at al.

[2006], indicating that subsurface alkenone growth could still lead to relatively

accurate SST predictions (Le. within ±3°C). Additionally, the efficiency of

alkenone export production to sediment traps deployed at 100 m is determined to

be -20%. This alkenone export efficiency is the first measurement of this type to

be reported in the literature and implies considerable loss of alkenones probably

due to the effects of grazing.

2.1. Sampling site

CHAPTER 2 SAMPLES AND METHODS

12

Samples were collected and experiments performed in and around the

Guaymas Basin, Gulf of California (27°30'N, 111 °20'W) aboard the 'RV New

Horizon during one wintertime (January - February 2005) and two summertime

(July 2004 & July - August 2005) cruises. GoCal1 (July 7-23,2004) and GoCal2

(January 25 - February 10, 2005) occurred exclusively within Guaymas Basin

(station 2), while GoCal3 (July 23 - August 13, 2005) also visited 3 other stations

(3-4, 1) in the Gulf for comparative stUdies (Fig. 2).

2.2. Suspended particulate material samples

Samples of suspended particulate material (SPM) were collected

throughout the upper water column (-0-80 m in GoCal1, -40 m in GoCal2 and

GoCaI3). Large volume (-80 L) seawater samples for alkenone analysis were

collected approximately every 5 m by closing multiple 10-L PVC sample bottles

on a CTD rosette during dedicated CTD casts. For each depth, water collected

was pressure filtered (-10 psi) through a single precombusted glass-fiber filter

(148 mm or 90 mm dia. Whatman GF/F or Millipore APFF, each 0.7 pm nominal

pore size) and stored frozen (-20°C) until processed for gas chromatographic

analysis of alkenones using both flame ionization detection (GC-FID) and

compound specific stable isotopic analYSis (irmGCMS).

Samples for nutrients (nitrate/nitrite, phosphate, and silicate), total

alkalinity (T A), and total dissolved inorganic carbon (DIC) concentration and

13

isotopic analysis were also obtained from dedicated CTD casts. Samples for DIC

isotopic analysis were collected without aeration in 20m I glass serum vials,

preserved by addition of HgCI2 (5pl of saturated solution), sealed with a butyl

rubber stopper and stored under darkness at room temperature for later

laboratory analysis. Samples for DIC were collected in 300 mL glass bottles with

ground-glass stoppers in a similar manner. Samples for T A analysis were

collected and handled similarly, but without the addition of HgCI2 [Karl et al.,

1990]. Nutrient samples were collected in pre-cleaned polyvials directly from the

PVC sample bottles on the Rosette and stored frozen (-20°C) for later analysis in

the laboratory as well. Light attenuation coefficients for photosynthetic active

radiation (PAR) were obtained using PAR data collected during daytime CTD

casts. Water column profiles of photosynthetic active radiation (PAR) were

calculated as described by Popp et al. [2006a] using SeaWiFS-derived 8-clay

average 27-km surface PAR that coincided with the timing of our in situ

experiments.

2.3. Free floating arrays

Water samples for incubation experiments were collected and handled

following the protocols of Prahl et al. [2005] and Popp et al. [2006a]. Water for

each incubation depth was transferred into polycarbonate carboys and a 13e­

labeled bicarbonate solution added. The carboys were then attached to an in

situ array, deployed before dawn, and allowed to drift freely for 24 hours before

recovery. Upon retrieval, samples were taken from each carboy and processed

14

as described for DlC isotopic analysis to confirm the level of isotopic enrichment.

Remaining water volumes were then pressure filtered and processed as

described for SPM material.

2.4. Sediment trap samples

Passively settling particulate matter was collected by attaching twelve

VERTEX-style traps to the free-floating array (-100 m depth) following the

protocol of Knauer et a/. [1979]. Upon retrieval, the particulate matter from one

trap was combined onto a single 25 mm precombusted glass-fiber filter and

processed for analysis of POM and PN as described by White et al. [2007]. The

remaining traps (cross sectional area = 0.0039 m2/trap) were combined for

alkenone analysis and filtered as described above for SPM samples.

2.5. Analytical techniques

At Oregon State University (OSU), alkenone fractions were isolated from

the sediment trap, incubation, and SPM samples using established ultrasonic

solvent extraction and column chromatographic methods [Prahl et al., 1989].

The precision of absolute alkenone concentrations and alkenone unsaturation

(UK'37) values were better than ±10% and ±0.01, respectively. Alkenone fractions

were then saponified [Christie, 1973] and sent to the University of Hawaii for

compound specific carbon isotopic analysis by irmGCMS [Hayes et al., 1990].

Natural abundance (SPM and sediment trap) and isotopically labeled

(incubation) samples were chromatographically purified using a TraceGC

15

equipped with a cool, on-column injector and a J&W Scientific DB-1MS column

(60-m x 0.32-mm x 0.25-pm) that was temperature programmed from 60 to

320°C at 1Q°C min"' prior to on-line isotopic analysis using an MAT 252 isotope

ratio monitoring gas chromatograph mass spectrometer (irmGCMS). Precision of

replicate irmGCMS analyses of natural abundance samples ranged from ±0.04 to

±0.53%o but averaged ±0.27%O, while isotopically labeled sample values ranged

from ±0.01 to ±2.2%O and averaged ±0.66%. For the purpose of error

propagation, conservative values of ±0.5 and ±1.0o/oo were used for the precision

of analysis, respectively.

~'3CDlC values were determined at the University of Hawaii (UH) using a

Gasbench interfaced to a Delta XP mass spectrometer. Typical precision for

replicates was better than 0.2%0. DIC concentrations were determined

coulometerically at UH using a Single-Operator Multi-Metabolic analyzer

(SOMMA) system similar to that described by Johnson et a/. [1993]. Total

alkalinity was determined by the Gran method [Gran, 1952] using computer­

controlled titration at UH as well. Precision of replicate samples was better than

±5 jJeq kg"'. Nitrate+nitrite and phosphate concentrations were analyzed at OSU

using the colorimetric techniques of Strickland and Parsons [1972] on an AJpkem

"Flow Solution" Autoanalyzer continuous flow system. Silicate concentrations

were determined according to the method of Armstrong et a/. [1967] as adapted

by Atlas et a/. [1971]. The detection limits (and coefficients of variation) for

nitrate, phosphate. and silicate measurements were 0.1 pmol L-1 (0.2%), 0.2

pmol L-1 (1 %), and 0.3 pmol L-1 (0.5%), respectively.

16

Alkenone production and photoperiod growth rates were calculated from

the change in the 13e atom percent of the alkenones and e02(aq) measured after

incubation as described by Popp at a/. [2006a, b). The 13e at. % of alkenones

was corrected for the 4.2%0 isotope offset between the alkenone and the primary

photosynthate that is associated with biosynthesis and for isotopic fractionation

associate with fixation of e02(aq) [Popp et al., 1998). e02(aq) 13e at. % was used

to calculate production rates instead of the ole 13e at: % which was used in a

previous study [see Hamanaka et al., 2000) because e02(aq) is the primary

inorganic carbon substrate utilized by E. huxley; [Rost et a/., 2003).

e02(aq) 13e at. % was calculated from ~13eDlC values and the relative

abundances of the various carbonate species determined using the e02SYS

program developed by Lewis and Wallace [1998). This program (available at

http://cdiac.oml.gov/oceans/co2rprt.html) calculates the species distribution of

the carbonate system from salinity, temperature, depth, TA, and the

concentrations of Ole, phosphate, and silicate. The dissociation constants for

KS04, B(OHh, and H2e03 used in this calculation were adopted from Dickson

[1990a, 1990b) and Mehrbach et a/. [1973) as refit by Dickson and Millero [1987).

The temperature fractionation relationships used to calculate e02(aQ) 13e at. %

were taken from Deines et al. [1974] and Mock et al. [1974]. Propagated errors

for alkenone production and growth rate calculations were approx. 10% of the

given rate.

CHAPTER 3 RESULTS AND DISCUSSION

3.1. Water column properties

17

Water column profiles of relevant properties are presented for each station

in Figure 4. Conditions in the Gulf of Califomia varied both by season (a-o) and

by location (o-f). The depth of the SML varied from 32-40 m during the winter

cruise (GoCaI22, Fig. 4b) and from <5 m (GoCa131, Fig. 4f) to -16 m (GoCa132

and GoCal34, Fig. 4c and 4e, respectively) in the summer cruises. Dissolved

inorganic nitrogen (NO£ + N03-) was below the detection limit (0.1 pmol L-1) in

surface waters during the summer, but increased rapidly below 40 m. Surface

PAR values ranged from 1139 pEin m-2 S-l (GoCaI32, Fig. 4c) to 1215 pEin m-2 s-

1 (GoCal34, Fig. 4e) in the summer and was approximately 905 pEin m-2 S-l in the

winter. Light attenuation coefficients varied with station and ranged from 0.075

m-1 in the winter to 0.178 m-1 in the summer (Table 2). In general, PAR

decreased exponentially with increasing water depth (Fig. 4) with the greatest

attenuation of light occurring at Station GoCal12 (Fig. 4a).

3.2. Depth of Alkenone Export

Depth of alkenone export in the Gulf of Califomia can be constrained to an

approximate range at each cruise/station by comparing the depth profiles of

alkenone unsaturation (UK•37) values and alkenone-specific carbon isotopic

composition (5Ka7:2) with similar measurements in material collected in the

sediment traps. Systematic variations with depth in the water column of each

Table 2. Photosynthetic_Active Radiation (PAR) In the Gulf of california

Cruise/Station Date Photoperiod k sPAR PAR...ro PARzo2Q PARzo30

(hrs) (m·1) wEin m·2 s·1) wEin m-2 S·1) wEin m·2 S·1) wEin m·2 S·1)

Gocal12 7/1412004 13.72 0.1785 1168.2 196.1 32.9 5.5 GoCaI22 1/3112005 10.87 0.0746 904.6 428.9 203.4 96.4 GoCal32 712912005 13.49 0.0836 1138.7 493.6 213.9 92.7 GoCal33 8/112005 13.58 0.0955 1151.6 443.4 170.7 65.7 GoCaI34 81412005 13.31 0.0982 1214.6 454.9 170.4 63.8 GoCal31 81612005 13.30 0.1025 1196.7 429.3 154.0 55.2

where k is the diffuse light attenuatlon-coefficent, sPAR is PAR at the surface, and PAR. is PAR at that respective depth calculated by the equation: PAR.=sPAR*e.Jcz

..... 00

(a)

o

20

I 40

'" -a 60 ~

C

80

Temperature (DC)

1416182022 24 2628 30 32

. N~N '(~';'O'I k~ -1') ,

o 5 10 15 20 25

• ~ • •

- PAR ---4- N+N

100 1# • Temperature

(d)

20

40

;; g- 60 c

80

100

o 200 400 600 800 1000

PAR (~E i n m-2 s-l )

Temperature (ee )

1214 16 1820 22 24 26 28 30 32

' N~N (~OI 'k9~ 1 ) ' ,

o

•

•

5 10 15 20 25

•

~ -• - • . -

o 200 400 600 800 1000

PAR (J,LEin m·2 5.1)

(b)

o

20

40

I a 60 • c

80

100

(e)

Temperature ('"C)

14 15 16 17 18 19 20

N+~ (~I k~·1 ) o 5 10 15 20 25

- PAR ~ N+N

• Array2 e. . . Array1

o 200 400 600 BOO

PAR ().lEin mo2 s·1)

Temperature ('C)

14 16 1820 22 24 26 28 30 32

20

I 40

'" C. 60 c3

60

100

o

N+N (';"'~ I kg-l) ,

• •

5 10 15 20 25

•

. 1· ...: ••• -• ••

o 200 400 600 800 1000

PAR (~E in m-2 s-l)

(c)

20

80

Temperature (DC)

14 16 18 20 22 24 26 28 30 32

N+N (.mol kg-1)

o 5 10 15 20 25

•

•

. ----100 ..

(I)

o 200 400 600 800 1000

PAR (jlEin m-2 s·1)

Temperature (0C)

16 18 20 22 24 26 28 30 32

N~N (~Ol 'k9:1) ,

-2024681012141618 o •

r 20

I 40

'" C. 60 ~

C

80

lOa ••

• •

•

-• •• ...

o 200 400 600 800 1000

PAR (.ein m-2 s-l )

Figure 4. Water column profiles of available light (PAR), dissolved inorganic nitrogen (NO; +

NO,-) and temperature for GoCal12 (a), GoCal22 (b), GoCal32 (c) , GoCal33 (d), GoCal34 (e),

and GoCal31 (f). The temperature profile in GoCal22 was slightly different for each incubation

experiment.

19

parameter are necessary to constrain the depth of alkenone export. For each

station/cruise in the present study, at least one of the two parameters shows a

sufficient range in values for comparison to the sediment trap values with the

exception of the winter cruise (GoCaI22) (Fig. 5). We assumed that UK'37 and

OK372 values measured in the sediment trap material reflect the depth of

alkenone export production , Additional assumptions are that there are no

changes in these parameters during export and that lateral inputs of alkenones

(a) (b)

0 0 • • '.. 4 . .. · \ · "f . .' • • 20 i' ~, 20 . • ". =.~ • to' ' .. • . ..,' • • • • • • • · . • • • • • • • • .~

I 4<l t • I 40 • • • • £ • ~

Q, 60 • GC22 li 60 Q GC32 Q GoCal32 C C

• GC12 • GoCaI33 80 • 80 • GoCat34

• GoCal31

100 , , .. 100 · . • • 0.5 0 .• 0.7 0.8 0 .• 1.0 0.60 0.85 0.90 0.95 1.00

UK'37 UK'37

(c) (d)

0 0 • • • • • • . - , • • • • • ,

20 • ~. • • 20 • • • .. -: .... • • • • • , . • • · • • • . • • •

I 40 I 40 • • • • £ ~

Q, 60 li 60 0 Q

C C

80 • 80

100 . .. , , 100 • • • • ·28 ·27 ·26 ·25 ·24 ·23 ·22 ·29 ·28 ·27 ·26 ·25 ·24

8K37:2 6K37,2

Figure 5. Depth profiles of alkenone unsaturation patterns (UK'37) for station 2 (a) and cruise 3

(b) and alkenone·specific carbon isotopic patterns (i5K37'2) for station 2 (c) and cruise 3 (d).

Respective UK'37 and i5K37,2 values obtained from the sediment traps are plotted at the 100m

collection depth for comparison to the overlying water column profiles.

20

21

are relatively minor. Grice at a/. (1998) found that there was no alteration in UK' 37

or riK:!7:2 as a function of zooplankton grazing.

Alkenone unsaturation pattems (UK'37) were determined for all SPM,

incubation, and sediment trap samples (Table 3). Depth profiles ofthe UK'37

index for each cruise/station are shown in Figure 5a-b. Values ranged from

0.587 in the winter (Fig. 5a) and reached maximum values of the UK' 37 index of

1.000 in the summer (GoCaI31, Fig. 5b). A UK

'37 index of 1 indicates that the K:!7:3

species was undetectable in the water column. At station 2, the GoCal12 UK'37

depth profile shows considerably more scatter than the GoCal32 profile, but both

show the same general trend of steadily decreasing UK' 37 values with increasing

depth (Fig. 5a). No trend with respect to depth is observed in the GoCal22 UK' 37

data (Fig. 5a), but since the SML extends to -40 m in the winter and

temperatures are relatively constant throughout this range, this result is not

unexpected. With respect to the other stations visited in Cruise 3, the observed

UK' 37 depth profile of GoCal32 is very similar to GoCa133, but GoCal34 and

GoCal31 appear quite different (Fig. 5b). The magnitude of the variation in UK'37

with depth seen in GoCal32 and GoCal33 is significantly reduced at the other 2

stations. From -0-30 m UK' 37 remains relatively constant, and then begins to

decrease at greater depths.

Alkenone-specific carbon isotopic values (riK:!7:2) of the SPM and sediment

trap samples were also determined (Table 3). Natural abundance isotopiC

variation between all cruises/stations ranged from -28.70/00 to -22.4%0. Depth

22

Table 3. Alkenone Unsaturation (UK'37) and Natural Abundance Carbon Isotopic (6:17:2) Pattems in the Gulf of Califomia

Cruise/Station Sample Type Depth UK'37 NA6:l72 (m) %0

GoCal12 CTD Rosette 7.8 0.993 -25.27 GoCal12 CTD Rosette 10.3 0.983 -26.20 GoCal12 CTD Rosette 10.4 0.993 -26.75 GoCal12 In situ Pump 10.4 0.987 -25.97 GoCal12 CTDRosatte 10.4 0.994 -25.87 GoCal12 CTDRosatte 10.4 0.997 -27.11 GoCal12 Array 1 10.5 0.982 GoCal12 CTD Rosette 18.2 0.972 -26.40 GoCal12 Array 1 20.5 0.906 GoCal12 Array 2 21.5 0.915 GoCal12 CTD Rosatte 25.1 0.947 -23.36 GoCal12 In situ Pump 25.4 0.803 -22.97 GoCal12 CTD Rosette 25.5 0.946 -23.51 GoCal12 CTD Rosette 25.6 0.954 -23.72 GoCal12 CTD Rosette 25.7 0.945 -22.75 GoCal12 CTD Rosette 27.3 0.908 -22.65 GoCal12 CTD Rosette 27.5 0.934 -22.44 GoCal12 CTD Rosatte 27.6 0.955 -23.92 GoCal12 CTD Rosatte 28.9 0.905 -23.34 GoCal12 Array 1 30.5 0.850 GoCal12 CTD Rosette 35.7 0.818 -23.04 GoCal12 CTD Rosette 35.7 0.858 -22.59 GoCal12 Array 2 40.5 0.855 GoCal12 CTD Rosette 40.8 0.886 -23.92 GoCal12 CTD Rosatte 41.1 0.859 -23.46 GoCal12 CTD Rosette 51.0 0.829 -23.90 GoCal12 CTD Rosette 80.5 0.897 -23.21 GoCal12 SedIment Trap 100.0 0.957 -25.41 GoCal12 Sediment Trap 100.0 0.943 -25.25 GoCaI22 Array 1 7.0 0.621 GoCaI22 Array 2 7.0 0.657 GoCal22 CTD Rosatte 11.6 0.621 -25.74 GoCaI22 Array 1 12.3 0.657 GoCaI22 Array 1 14.4 0.658 GoCaI22 Array 1 15.6 0.662 GoCaI22 CTD Rosette 16.4 0.587 -25.19 GoCal22 Array 1 19.9 0.655 GoCaI22 Array 2 19.9 0.6n GoCaI22 CTD Rosatte 21.6 0.603 -25.39 GoCal22 Array 1 21.9 0.660 GoCal22 Array 2 21.9 0.674 GoCaI22 Array 1 23.1 0.599 GoCaI22 Array 2 25.3 0.637

23

Table 3. (Continued) Alkenone Unsaturatlon (UI< 37) and Natural Abundance Carbon Isotopic (!5a7:2) Patterns In the Gulf of California

Cruise/Station Sample Type Depth UI<37 NA !5a7:2 (m) %0

GoCeI22 cm Rosette 27.0 0.619 -25.02 GoCal22 Array 1 27.4 0.651 GoCal22 Array 1 29.4 0.636 GoCal22 Array 2 30.0 0.694 GoCeI22 Array 1 30.5 0.644 GoCal22 CTD Rosette 32.0 0.635 -24.71 GoCal22 CTDRosette 36.6 0.639 -24.43 GoCal22 Sedlmant Trap 100.0 0.615 -22.91 GoCal22 Sediment Trap 100.0 0.660 -23.84 GoCel32 cm Rosette 6.3 0.996 -26.56 GoCel32 Array 1 10.2 0.994 GoCal32 CTD Rosette 11.2 0.985 -26.75 GoCel32 Array 1 15.0 0.991 GoCel32 CTD Rosette 16.6 0.976 -26.40 GoCal32 Array 1 21.6 0.962 GoCel32 CTD Rosette 26.5 0.935 -26.50 GoCel32 CTD Rosette 26.5 0.948 -26.45 GoCal32 Array 1 27.3 0.924 GoCel32 Array 1 29.2 0.908 GoCal32 Array 1 30.3 0.946 GoCal32 cm Rosette 31.8 0.937 -26.79 GoCel32 cm Rosette 36.8 0.859 -25.66 GoCal32 cm Rosette 42.8 0.857 -25.61 GoCal32 Sediment Trap 100.0 0.945 -25.72 GoCeI33 cm Rosette 6.0 0.999 -27.61 GoCel33 Array 2 7.0 0.993 GoCeI33 cm Rosette 11.4 0.997 -27.54 GoCeI33 Array 2 12.3 0.995 GoCeI33 Array 2 14.4 0.994 GoCeI33 Array 2 15.6 0.999 GoCeI33 cm Rosette 18.5 0.965 -27.61 GoCel33 Array 2 19.9 0.939 GoCel33 Array 2 21.9 0.936 GoCal33 Array 2 23.0 0.920 GoCel33 CTD Rosette 23.9 0.902 -27.50 GoCel33 Array 2 24.3 0.800 GoCal33 cm Rosette 26.8 0.972 -27.65 GoCel33 Array 2 28.4 0.802 GoCel33 Array 2 30.3 0.845 GoCel33 CTD Rosette 31.4 0.854 -27.56 GoCel33 Array 2 35.5 0.874 GoCal33 cm Rosette 36.7 0.843 -25.68 GoCal33 CTD Rosette 40.4 0.841 -25.08

24

Table 3. (Continued) Alkenone Unseturetion (UK' 37) and Natural Abundance Garbon Isotopic (1537:2) Patterns in the Gulf of California

Cruise/Station Sample Type Depth K' U 37 NA 1!a7:2

(ml 0/00 GoCal33 Sediment Trap 100.0 0.935 -26.52 GoCal34 CTD Rosette 6.8 0.995 -26.99 GoCal34 Array 3 10.2 0.970 GoCal34 CTD Rosatte 11.6 0.981 -25.74 GoCaI34 Array 3 14.4 0.993 GoCaI34 Array 3 15.6 0.992 GoCaI34 CTD Rosette 16.4 0.997 -25.84 GoCaI34 Array 3 19.9 0.998 GoCaI34 Array 3 21.9 0.995 GoCal34 CTD Rosette 22.0 0.998 -26.50 GoCal34 Array 3 23.0 0.992 GoCal34 Array 3 24.3 0.995 GoCaI34 CTD Rosette 24.4 0.997 -25.82 GoCaI34 Array 3 28.4 0.986 GoCal34 Array 3 30.3 0.992 GoCaI34 CTD Rosatte 31.5 0.987 -24.99 GoCaI34 Array 3 35.5 0.914 GoCaI34 CTD Rosette 36.8 0.978 -26.72 GoCal34 CTDRosatte 41.7 0.960 -24.42 GoCaI34 Sediment Trap 100.0 0.961 -25.42 GoCal31 CTD Rosatte 6.2 0.992 -28.69 GoCaI31 Array 4 8.1 0.997 GoCal31 CTDRosatte 11.6 0.994 -28.61 GoCal31 Array 4 12.3 1.000 GoCaI31 Array 4 14.4 0.999 GoCal31 Array 4 15.6 0.999 GoCal31 CTDRosatte 16.8 0.997 -28.46 GoCal31 Array 4 19.9 0.993 GoCal31 CTD Rosette 21.5 0.992 -27.66 GoCal31 Array 4 21.9 0.993 GOCal31 Array 4 23.0 0.995 GoCal31 Array 4 24.3 0.979 GoCal31 CTD Rosatte 26.6 0.984 -26.23 GoCal31 Array 4 28.4 0.981 GoCal31 Array 4 30.3 0.983 GOCal31 CTD Rosatte 31.7 0.989 -25.43 GoCal31 Array 4 35.5 0.983 GoCal31 CTD Rosatte 36.6 0.985 -25.30 GoCal31 CTD Rosatte 41.6 0.973 -24.84 GoCal31 Sediment Trap 100.0 0.972 -25.91

25

profiles of 151<J7:2 for each cruise/station generally increased with increasing depth

(Fig. 5c-d), but showed distinct differences between each cruise/station. Unlike

the observed UK'37 patterns, GoCaI12151<J7:2 differs significantly from GoCal32

15K37:2 (Fig. 5c). The depth profile of 151<J7:2 in GoCal32 remains relatively

constant (--26.6o/OQ) until a depth of 35 m and then increases -1 %.. In contrast,

GoCaI1215~7:2 values above 25 m cluster around -26.3%0 while those below

cluster around -23.3%., a dramatic -3%. shift. This shift is probably not caused

by changes in the alkenone-producing community. During GoCa112, the surface

waters were dominated by G. oceanica, while E. huxley; dominated at depth, but

during GoCa133, E. hux/eyi dominated at the surface and G. ocaanica dominated

at depth [Malinvemo et a/., in prep.1. If the -3%0 shift in the 151<J7:2 observed

during GoCal12 (Fig. 5c) reflects only a change in species distribution where

more negative 151<J7:2 values are attributed to greater abundance of G. ocaanica,

then 151<J7:2 values would be expected to decrease with increasing depth at

GoCa133. The opposite is observed at this station (Fig. 5d) indicating that the

shifts in 151<J7:2 values are not related only to a change in the species of alkenone­

producing algae. A more likely cause of the -3%0 shift during GoCal12 (Fig. 5c)

is an effect of light limitation on carbon isotopic fractionation in the alkenone­

producing algae. Rost et a/. [20021 documented an isotopic shift of up to 8%0 as

a result of E. hux/eyi grown in dilute batch cultures under conditions of limiting

light. If a similar effect is occurring during the present study, then an increase in

151<J7:2 values with respect to increasing depth might be expected. The fact that

26

isotopic values at the 3 remaining stations (GoCa122, GoCal34, and GoCa131) all

steadily increase with increasing depth with net changes of -1.40/00, 2.60/00, and

3.9%0 respectively (Fig. 5c-d), further support the explanation of Iight-dependent

isotopic fractionation. The lack of systematic variation with respect to depth

during GoCal12 (Fig. 5c) could simply be due to the fact that available light at

GoCal12 diminished more quicker in the water column than at the other

cruise/stations (Fig. 4 and Table 2).

Values of UK' 37 and 15Ka7:2 determined from the sediment trap material are

plotted at the 100 m collection depth in Figure 5 for comparison to the water

column profiles of each parameter. Corresponding depths derived from both

parameters of each cruise/station are presented in Table 4. Depths shown in

bold typeface indicate the parameter that shows the greatest magnitude of

systematic variation with respect to depth in the water column, and are therefore

capable of yielding the most precise estimate of the 2 depth extrapolations for

that given cruise/station. For example, the GoCal12 UK" 37 values (Fig. Sa) span a

wide range and encompasses both UK'37 sediment trap values. Corresponding

water column depths for the trap UK'37 are -26 m. On the other hand, the

GoCal12 15Ka7:2 values consist of 2 distinct clusters (Fig. 5c) that do not show

sufficient systematic variation within each cluster to pinpoint a specific depth of

inferred alkenone export production. Therefore the sediment trap 15Ka7:2 values of

-25.40/00 and -25.30/00 can only constrain the depth of alkenone export production

to the upper 25m of the water column.

Tabla 4. Alkenone Export in the Gulf of California

Cruise/Station Array # uK'arDepth !5:r,:TDepth Export Depth SMLDepth UK'3T"T SML-T UK'37"'T-SML-T ED-T ED-T-SML-T (m) (m) (m) (m) (OC) (OC) (OC) (OC) (OC)

GoCal12 Array 1 25.3 0-25 20-25 15 27.3 29.0 -1.7 23.1 -5.9 GoCaJ12 Array 2 28.5 0-25 20-25 15 26.8 29.0 -2.2 23.1 -5.9 GoCaJ22 Array 1 7-27 60.3 20-30 40 18.4 18.7 -0.3 18.6 -0.1 GoCai22 Array 2 12-22 45.8 20-30 32 17.0 18.7 -1.7 18.6 -0.1 GoCaJ32 Array 1 28.4 36.5 25-30 16 26.9 29.2 -2.3 23.2 -6.0 GoCaJ33 Array 2 21.1 34.3 20-25 6 26.6 29.8 -3.2 22.4 -7.4 GoCaI34 Array 3 41.4 28.8 25-30 16 27.4 29.9 -2.5 27.0 -2.9 GoCal31 Array 4 42.8 28.7 25-30 <5 27.7 30.3 -2.6 26.1 -4.2

Where UK' 3T" T is calculated from the sediment trap-derlved UK' 37 using the calibration of Conte et a/. [2006)

!::i

28

The most likely depth of export for each cruise/station is determined by

taking into consideration the 2 export depth predictions obtained by comparing

sediment trap UK' 37 and li1<J7:2 to the natural variation in the overlying water

column (i.e. 26 m for UK' 37 and 0-25 m for li1<J7:2 in GoCa112) (Table 4). Since the

GoCal12 UK'37 depth extrapolation is better constrained, a depth of alkenone

export production of 20-25 m is a reasonable estimate for this cruise/station (see

Table 4). Using this approach, the estimated depth of alkenone export

production was determined to be 20-25 m for GoCal33 and 25-30 m for GoCal32,

GoCal34, and GoCal31 (Table 4). Export depth for the winter cruise (GoCal22)

is the most difficult to ascertain because of the lack of systematic variation in the

UK' 37 water column profile (Fig. 5c) and because the sediment trap li1<J7:2 values

fall outside the range of values measured in the water column (Fig. 5d). liKs7:2

profile extrapolation leads to very deep export depths (>45 m), while the possible

range indicated by the UK·37 scatter is much shallower (10-30 m) (Table 4). The

cause for this discrepancy is unknown but for the purposes of this investigation, it

is assumed that the export depth in the winter at station 2 is relatively similar to

the summer export depths (Le. 20-30 m).

Using the sediment-based UK·3rtemperature relationship of Conte at a/.

[2006], UK' 37 values of the sediment trap samples can be converted into growth

temperatures (gTs). These predicted temperatures are then compared to the in

situ temperature of the SML (as determined from the CTD downcast data files) in

order to determine how well exported UK" 37 values record SST (Table 4). In all

29

cases, the temperature of the SML (SML-7) is underestimated by the UK' 3r

temperature prediction (indicated by a '-' value in Table 4). The magnitude of

the underestimation was relatively small for the winter cruise (0.3°C for GoCal22

array1), but was as large as 3.2°C for GoCal33 (fable 4). Interestingly, this

magnitude of variation is similar to the magnitude of scatter observed in the

Muller et al. [19981 and Conte et al. [20061 coretop UK' 3rSST relationship. In

other words, sediment trap-derived UK' 37 temperatures in the Gulf of California

systematically underestimates in situ SML temperatures, but the underestimation

falls within the observed variability of the method. This -3°C temperature

underestimate should represent the inherent temperature gradient between the

SML and the depth of alkenone export (-20-30 m) if UK'37 values are solely

regulated by gT (i.e. no genotypic or physiological effects like light and nutrient

limitation). A comparison of the in situ temperature at the depth of export (ED-7)

to the in situ SML temperature (SML-7) (fable 4) shows that there is a

significantly greater temperature difference between the ED-Tand SML-Tthan

the UK' 3r T and SML-T in the summertime. The cause of this discrepancy is

addressed in section 3.4.

3.3. Depth of maximum alkenone concentration and production

In addition to examining the depth of alkenone export, the depth of

maximum alkenone concentration (standing stock) and production rate can give

additional insight into the growth conditions of the alkenone-producing algae in

the Gulf of California. Concentration of Ks7:2 alkenones throughout the water

30

Table 5. Ka7:2 Alkenone Concentration In the Gulf of California

Cruise! Cruise! Station Sample Type Depth Ka7:2 Station Sample Type Depth Ka7:2

m n L·1 m n L·1

GoCal12 CTD rosette 7.8 15.1 GoCal32 Array 30.3 8.8 GoCal12 CTD rosette 10.3 8.9 GoCal32 CTD rosette 31.8 76.4 GoCal12 CTD rosette 10.4 40.5 GoCal32 CTD rosette 36.8 31.0 GoCal12 In situ pump 10.4 17.4 GoCal32 CTD rosette 42.8 13.2 GoCaJ12 CTD rosette 10.4 35.2 GoCaI33 CTD rosette 6.0 20.0 GoCal12 CTD rosette 10.4 34.2 GoCaI33 Array 7.0 19.0 GoCal12 Array 10.5 19.4 GoCal33 CTD rosette 11.4 47.0 GoCal12 CTD rosette 18.2 34.7 GoCal33 Array 12.3 33.9 GoCal12 Array 20.5 14.9 GoCaI33 Array 14.4 27.5 GoCaJ12 Array 21.5 14.3 GoCaJ33 Array 15.6 19.2 GoCaJ12 CTD rosette 25.1 3.1 GoCaI33 CTD rosette 18.5 103.1 GoCal12 In situ pump 25.4 16.9 GOCaI33 Array 19.9 53.6 GoCal12 CTD rosette 25.5 2.6 GoCal33 Array 21.9 75.2 GoCaJ12 CTD rosette 25.6 8.9 GoCal33 Array 23.0 59.3 GoCal12 CTD rosette 25.7 3.7 GoCal33 CTD rosette 23.9 78.1 GoCal12 CTD rosette 27.3 9.4 GoCal33 Array 24.3 27.4 GoCal12 CTD rosette 27.5 9.1 GoCal33 CTD rosette 26.8 75.7 GoCal12 CTD rosette 27.6 21.1 GoCal33 Array 28.4 21.3 GoCal12 CTD rosette 28.9 12.7 GoCaJ33 Array 30.3 24.3 GoCal12 CTD rosette 30.3 10.7 GoCaJ33 CTD rosette 31.4 117.7 GoCal12 Array 30.5 7.5 GoCaJ33 Array 35.5 10.3 GoCal12 CTD rosette 35.7 36.1 GoCal33 CTD rosette 36.7 37.9 GoCaJ12 CTD rosette 35.7 3.5 GoCaJ33 CTD rosette 40.4 16.8 GoCal12 Array 40.5 1.2 GoCal34 CTD rosette 6.8 10.7 GoCal12 CTD rosette 40.8 4.9 GoCal34 Array 7.6 7.3 GoCaJ12 CTD rosette 41.1 4.1 GoCaJ34 CTD rosette 11.6 10.3 GoCal12 CTD rosette 51.0 1.8 GoCal34 Array 11.8 5.6 GoCal12 CTD rosette 80.5 0.6 GoCaJ34 Array 14.4 7.7 GoCal22 Array 7.0 12.2 GoCal34 Array 15.6 8.4 GoCaI22 Array 7.0 13.4 GoCal34 CTD rosette 16.4 2.8 G0CaJ22 CTD rosette 11.6 24.1 GoCaJ34 Array 19.9 17.6 GoCaI22 Array 12.3 10.9 GoCal34 Array 21.9 19.2 GoCaI22 Array 14.4 9.7 GoCal34 CTD rosette 22.0 20.6 GoCaJ22 Array 15.6 8.9 GoCal34 Array 23.0 14.9 GoCaI22 CTD rosette 16.4 13.1 GoCal34 Array 24.3 29.9 GoCal22 Array 19.9 10.6 GoCaJ34 CTD rosette 24.4 33.4 GoCal22 Array 19.9 19.8 GoCal34 Array 28.4 37.8 GoCaI22 CTD rosette 21.6 12.6 GoCaJ34 Array 30.3 23.6 GoCaI22 Array 21.9 17.7 GoCal34 CTD rosette 31.5 37.0 GoCaJ22 Array 21.9 11.2 GoCal34 Array 35.5 14.2 GoCaI22 Array 23.1 15.9 GoCal34 CTD rosette 36.8 54.5 GoCaI22 Array 23.1 4.4 GoCal34 CTD rosette 41.7 17.1 GoCaI22 CTD rosette 27.0 16.6 GoCal31 CTD rosette 6.2 27.5

31

TableS.

Cruisel Cruise! Station Sample Type Depth 1<07:2 Station Sample Type Depth 1<07:2

m n L" m n L"

Gocal22 Array 27.4 10.1 GoC8131 Array 8.1 56.3 Gocal22 Array 27.4 3.6 GoCal31 CTD rosette 11.5 24.7 GoCal22 Array 29.4 8.8 GoCal31 CTD rosette 11.6 61.5 Gocal22 Array 29.4 6.2 GoCal31 Array 12.3 49.3 Gocal22 Array 30.5 10.3 GoCal31 Array 14.4 53.1 Gocal22 Array 30.5 1.1 Gocal31 Array 15.6 44.2 GoCal22 CTD rosette 32.0 13.5 GoCal31 CTD rosette 16.8 32.4 GoCal22 CTD rosette 36.6 18.8 GoCal31 Array 19.9 33.0 GoCal32 CTD rosette 6.3 12.9 Gocal31 CTD rosette 21.5 62.7 Gocal32 Array 8.1 4.9 Gocal31 CTD rosette 21.7 19.1 GoCal32 CTD rosette 11.2 10.5 GoCal31 Array 21.9 31.1 GoCal32 Array 12.3 4.9 GoCal31 Array 23.0 30.5 GoCal32 Array 14.4 5.4 Gocal31 Array 24.3 32.1 Gocal32 Array 15.6 3.9 Gocal31 CTD rosette 26.6 57.0 GoCal32 CTD rosette 16.6 38.3 Gocal31 Array 28.4 27.8 Gocal32 Array 19.9 3.4 GoCal31 Array 30.3 20.8 Gocal32 Array 21.9 3.3 GoCal31 CTD rosette 31.6 26.5 GoC8132 Array 23.0 3.8 GoCal31 CTD rosette 31.7 23.0 Gocal32 CTD rosette 26.5 70.1 Gocal31 Array 35.5 12.7 GoCal32 CTD rosette 26.5 85.5 GoCal31 CTD rosette 36.6 39.7 GoCal32 Array 27.3 11.9 GoCal31 CTD rosette 36.8 16.3 GoCal32 Arra 29.2 11.9 GoCal31 CTD rosette 41.6 15.4

column varied between cruise/station with a total range of -0.5 ng L'1 (80 m.

GoCa112) to 118 ng L'1 (31 m. GoCa133) (Table 5). Water column profiles of

standing stock at each indMdual cruise/station (Fig. 6) show that Ka7:2 alkenone

concentrations follow a specific pattern in the water column during the

summertime (Fig. 6a.c-f). but not in the winter (Fig. 6b). The variation in the

concentration of Ka7:2 seen in the results from the winter cruise (GoCal22. Fig.

6b) is reminiscent of the variability observed in the GoCal22 UK'37 values (Fig. 5a)

and could be attributed to the same cause (Le. all of the concentration samples

taken in GoCal22 were within the relatively deep SML. and systematic variation

(a)

0,--------------------------,

• • , , .. , 20

, - . " .. , , ,

I 40 · -~ ,!l 60

80 • SPM Sample Water • Incubation Water

100 ~--------__ ----__ ----__ --~

o 10 20 30 40 50

K37:2 Concentration (ng L·1)

(e)

0,------------------------,

, 10 • •

" !: 20 •

~ ,!l

(e)

30

40

o

\ . , . , • ,

•

20 40 60 80 100

K37:2 Concentration (ng L-1)

0,------------------------,

10

•

40

o

, ' • •

"

10

• , .. . , ,

• ,

•

20 30 40 50 60

K37:2 Concentration (ng L -1)

32 (b)

0 ,---------------------------,

10

I 20 £ c­o C

(d)

30

40

• ,

o 5

• • ,

.' • • • , . , • , ,

•

10 15 20 25 30

K37:2 Concentration (n9 l-1)

0 ,----------------------------,

, 10 •

• • • • • • • •

• • • • • •

40 •

o 20 40 60 80 100 120 140

K37:2 Concentration (ng L -1)

0,------------------------,

10

:e: - 20

'" 15. o c

30

40

o 10

• • • • •

• , •

• • ,

• • , . • • •

20 30 40 50 60 70

K37:2 Concentration (ng L -1)

Figure 6. Depth profiles of alkenone K37,2 concentration for GoCal12 (a), GoCal22 (b), GoCal32

(c), GoCal33 (d), GoCal34 (e), and GoCal31 (f). Concentrations in the water collected for

incubation experiments generally matched the concentrations found in the SPM samples with the

notable exception of GoCal32 (c) where concentrations in the incubation experiments were

significantly lower than in the SPM samples.

33

in standing stock due temperature differences would not occur). During the

summer of 2004 (GoCal12, Fig. 6a), the concentration of K37:2 was greatest at

the surface and decreased with increasing depth, but during the following

summer (GoCa13), K37:2 concentrations at all stations were relatively low at the

surface and maximized deeper in the water column between 20-30 m (compare

Figs. 6a and 6c-f).

K37:2 production rate was calculated from the uptake of t3C labeled

inorganic carbon during in situ incubation experiments and ranged from -0 ng L-t

d-t (40m, GoCa112) to 27.4 ng L-t d-t ( 27 m, GoCa132) (Table 6). Variation in

production rate with depth was observed at each cruise/station (Fig. 7) and

generally maximized between 20-30 m (the 2 arrays in GoCal12 and GoCal22

were combined to produce a more complete profile for each respective

cruise/station). Maximum production rates with respect to depth at each

(a) (b)

Or-----------------------~ 0

GoC8122 - GoCal32

~--~ ~.~~ 10

I 20

" C. ~ C

30

---- GoCal33 - GoCal34 -+- GoCal31

40 40

o 5 to t5 20 25 30 0 5 10 15 20 25 30

K37:2 Production Rate (~g L-t dot) Production Rate (~g L-t dot)

Figure 7. Depth profiles of K,72 production rates for station 2 (a) and cruise 3 (b) where

production was calculated from the change in the 13C atom percent of the alkenones and CO2(oq)

measured after incubation as described by Popp et a/. [2006a, b].

34

Table 6. Ka7:2 Production in the Gulf of Califomia

Cruise/Station Array # Depth Ka7:2 NA~7:2 Labeled ~7:2 ~!537:2 Ka7:2 PR

(m) (nlil L") (%0) (%0) (%0) (nlil L" a') GoCal12 Array 1 10.5 19.4 -26.3 -1.6 24.7 2.4 GoCal12 Array 1 20.5 14.9 -26.3 -2.1 24.2 1.8 GoCal12 Array 1 30.5 7.5 -23.3 -13.1 10.1 0.4 GoCal12 Array 2 21.5 14.3 -26.3 6.5 32.8 2.5 GoCal12 Array 2 40.5 1.2 -23.3 -18.5 4.8 0.0 GoCaI22 Array 1 7.0 12.2 -25.7 16.8 42.5 3.0 GoCal22 Array 1 12.3 10.9 -25.7 9.6 35.2 2.2 GoCal22 Array 1 14.4 9.7 -25.4 15.7 41.1 2.3 GoCal22 Array 1 15.6 8.9 -25.3 16.5 41.7 2.1 GoCal22 Array 1 19.9 10.6 -25.4 -4.0 21.3 1.3 GoCaI22 Array 1 21.9 11.2 -25.4 -2.1 23.3 1.5 GoCal22 Array 1 23.1 15.9 -25.3 -24.9 0.5 0.0 GoCal22 Array 1 27.4 10.1 -25.0 -6.9 18.0 1.1 GoCaI22 Array 1 29.4 8.8 -24.9 -11.5 13.3 0.7 GoCaI22 Array 1 30.5 10.3 -24.8 -16.8 8.0 0.5 GoCal22 Array 2 7.0 13.4 -25.7 25.8 51.6 3.8 GoCaI22 Array 2 19.9 19.8 -25.4 55.9 81.3 9.1 GoCal22 Array 2 21.9 17.7 -25.4 43.6 69.0 7.2 GoCal22 Array 2 25.3 4.0 -25.1 21.5 46.6 1.1 GoCaI22 Array 2 30.0 3.6 -24.8 -0.2 24.7 0.5 GoCal32 Array 1 10.2 4.9 -26.6 23.6 50.1 5.7 GoCal32 Array 1 15.0 4.7 -26.6 23.7 50.2 9.7 GoCal32 Array 1 21.6 3.5 -26.6 -3.6 23.0 7.2 GoCal32 Array 1 27.3 11.9 -26.6 34.8 61.4 27.4 GoCal32 Array 1 29.2 11.9 -26.6 34.8 61.4 24.2 GoCal32 Array 1 30.3 8.8 -26.6 23.6 50.2 17.9 GoCaI33 Array 2 7.0 19.0 -27.6 9.8 37.3 3.5 GoCal33 Array 2 12.3 33.9 -27.6 1.7 29.3 4.9 GoCal33 Array 2 14.4 27.5 -27.6 2.5 30.0 4.0 GoCal33 Array 2 15.6 19.2 -27.6 -0.2 27.4 2.5 GoCal33 Array 2 19.9 53.6 -27.6 0.5 28.1 7.3 GoCal33 Array 2 21.9 75.2 -27.6 -4.0 23.6 8.7 GoCal33 Array 2 23.0 59.3 -27.6 -7.0 20.5 6.0 GoCaI33 Array 2 24.3 27.4 -27.6 2.6 30.2 4.2 GoCal33 Array 2 28.4 21.3 -27.6 -16.5 11.1 1.2 GoCal33 Array 2 30.3 24.3 -27.6 -21.1 6.4 0.8 GoCal33 Array 2 35.5 10.3 -25.9 -24.8 1.1 0.1 GoCal34 Array 3 10.2 6.5 -26.9 21.7 48.6 1.6 GoCal34 Array 3 14.4 7.7 -26.7 23.5 50.3 1.9 GoCal34 Array 3 15.6 8.4 -26.7 27.3 54.0 2.3 GoCaI34 Array 3 19.9 17.6 -26.6 19.3 45.9 4.1 GoCal34 Array 3 21.9 19.2 -26.5 17.2 43.7 4.1 GoCaI34 Array 3 23.0 14.9 -26.2 17.3 43.5 3.2

35

Table 6. (Continued) 1<07:2 Production in the Gulf of California

Cruise/Station Array # Depth 1<07:2 NA !Sa7:2 Labeled /j37:2 Al!ar:2 Kar:2 PR

(m) (ne LO') (%0) (%0) (%0) (ne L

O' dO')

GoCal34 Array 3 24.3 29.9 -25.8 36.6 62.5 9.5 GoCal34 Array 3 28.4 37.8 -25.4 45.9 71.3 14.1 GoCal34 Array 3 30.3 23.6 -25.1 42.2 67.3 8.2 GoCal34 Array 3 35.5 14.2 -24.8 37.9 62.6 4.6 GoCal31 Array 4 8.1 56.3 -28.7 11.4 40.1 11.1 GoCal31 Array 4 12.3 49.3 -28.6 6.6 35.2 8.5 GoCal31 Array 4 14.4 53.1 -28.6 6.1 34.6 9.0 GoCal31 Array 4 15.6 44.2 -28.5 5.5 34.0 7.4 GoCal31 Array 4 19.9 33.0 -28.0 7.2 352 5.8 GoCal31 Array 4 21.9 31.1 -27.6 2.5 30.0 4.6 GoCal31 Array 4 23.0 30.5 -27.3 -1.9 25.3 3.9 GoCal31 Array 4 24.3 32.1 -26.9 69.7 96.6 15.8 GoCal31 Array 4 28.4 27.8 -25.9 49.3 75.2 10.8 GoCal31 Array 4 30.3 20.8 -25.6 24.7 50.3 5.3 GoCai31 Array 4 35.5 12.7 -25.3 -7.9 17.4 1.1

where NA /j37:2 is the smoothed neturel abundance alkenone-specific carbon isotopic pattern determined from the water column profiles (Figure 5). A!Sa7:2 Is the difference between the Naturel Abundance profile and the isotopically labeled profile from the incubation experiments. and Kar:2 PR is the Production Rate of the 1<07'2 alkenones

cruise/station were greatest during cruise 3 (Fig. 7b). and lowest during cruises 1

and 2 (Fig. 7a). Given the results of the Goni et a/. [20011 sediment-trap time

series that found alkenone flux maximizing in the summe~ and minimizing in the

winter, low production rates during the winter cruis, (GoCaI22) are to be

expected. The cause of the low observed production rates during the summer of

2004 (GoCa112, Fig. 7a) is much less straightforward. Sampling resolution

during GoCal12 was reduced with respect to the other cruise/stations, and the

maximum production peak could have been missed (e.g. compare GoCal32 to

GoCal12 from 20-30 m in Fig. 7a). Physical differences in the water column

between the two summers could also explain a reduced production rate in 2004.

Maximum standing stock concentrations in GoCal12 were approximately 50% of

the maximum standing stock observed in GoCal32 (Fig. 6a,c). Concentration

also maximized at the surface during GoCal12 (Fig. 6a), while subsurface

36

maxima was seen in all concentration depth profiles during the following summer

(cruise 3, Fig. 6c-f). This difference could be related to the availability of light.

With respect to depth, PAR diminished much quicker in GoCal12 than in

GoCal32 (Fig. 4a,c) due to the higher attenuation of light. If the availability of

light were limiting production during GoCal12, favorable growth conditions would

be found closer to the surface, and could explain the observed concentration and

production discrepancies.

The depths of maximum Ka7:2 concentration and production rate appear to

coincide with previously determined depths of alkenone export production (Table

7). All depths of maximum Ka7:2 concentration and production rate are within

-5m of the inferred depth of alkenone export production. These results suggest

that alkenones preserved in sediments would likely record the gT of water at 20-

30 m, which during the summer, is well below the SML and is therefore not

indicative of SST at that time.

Table 7. A Comparison of Export Depth to the Depths of Maximum Concentration and Production in the Gulf of California

Cruise/Station Export Depth 1<:.7:2 Maxima PRMaxima

(m) (m) (m)

GoCal12 20-25 10.4 21.5

GoCal22 20-30 11.6-36.6 21.9

GoCal32 25-30 26.5 27.3

GoCal33 20-25 18.5-31.4 21.9

GoCal34 25-30 28.4-36.8 28.4

GoCal31 25-30 21.5 24.3

37

Depth integrated alkenone concentration and production rates were

determined for each station (Table 8). The inventory of K37:2 in the upper water

column (0-40 m) was greatest when a subsurface maximum was present (Cruise

3) and lowest when absent (Cruises 1 and 2) (Table 8). Integrated KJ7:2

production rates for the upper water column (0-35 m) varied with respect to

cruise/station and differed in relative magnitude to the integrated concentrations.

Highest integrated concentration was seen in GoCal33, while highest integrated

production rate was seen in GoCal32 (Table 8). Integrated concentration and

production rate were equal and low during GoCal12 and GoCa122.

Concurrent measurements of integrated KJ7:2 production and sediment

trap accumulation rate allow calculation of the efficiency of alkenone export

production. Using the surface area of the collection cups and the duration of

deployment, alkenone concentrations collected with the sediment trap are used

to calculate flux and compared to integrated production measured in the

overlying water column (Table 8). Trapping efficiencies ranged from 9-34% with

an average efficiency of -20%, indicating that -80% of the alkenones produced

Table 8. Int rated Ka7:2 Concentration and Production in the Gulf of califomia

CrulsefStetion Array # IC IP IP Trap lKa7:i1 % Eff. 119 m·2 11 m·2 cr' 119 d·' 11 cr'

GoCal12 1+2 533.9 63.4 2.72 0.79 29.1 GoCal22 1+2 524.0 82.6 3.54 1.21 34.2 GoCal32 1 1763.7 321.3 13.78 1.30 9.4 GoCal33 2 2182.9 127.1 5.45 o.n 14.0 GoCal34 3 768.1 163.3 7.00 1.13 16.1 GoCal31 4 1459.4 301.4 12.93 3.32 25.7

where IC and IP are the Integrated concentration and production of the upper 4Om, and Trap lKa7:2l is the flux of Ka7:2 alkenones to the sediment traps

38

in the euphotic zone are not represented by the material collected with the

sediment traps. Previous studies on the trapping efficiencies of this type of

sediment trap deployed at four different locations (BATS, HOT, Dabob Bay, and

the Baltic Sea) ranged from 44-69% [Buesseler at al., 2007 and references

therein] indicating that a significant fraction of the lost alkenone export can be

attributed to the limitations of sediment trap measurements (Le. alkenones

degrading within the collection cup or failing to remain in the collection cup due to

turbulence at the sediment cup-water interface [Buesseler et al., 2007]).

Assuming that the collection efficiency of the sediment traps deployed in the

present study is similar to those discussed in Buesseler et al. [2007] (i.e. -50%),

then the -20% collected with the sediment traps represents export of -40% of

the total production. This leaves -60% of the alkenones produced in the

euphotic zone unaccounted for. Possible factors contributing to this loss include

alkenones being recycled in the upper water column, .subjected to lateral

transport, or degraded during export. Additionally, because of the significant

fraction of alkenone production that is not exported, it appears unlikely that

accurate estimates of paleoproduction could be determined by measuring

alkenone concentrations or accumulation rates in sediments.

3.4. Comparison of If37 and in situ temperatures

UK'37 measured in the SPM collected throughout the water column is

compared to in situ temperatures in order to determine how well the most recent,

sediment-based, UK' 37"temperature relationship of Conte et al. [2006] predicts gT

39

in the Gulf of California (Fig. 8). Results shown in Figure 8 appear qualitatively

similar to the corresponding plot of Goni et a/. [2001] (Fig. 2) in that both

relationships are relatively linear until in situ temperatures reach -26°C. The

reduction in slope at temperature extremes has been documented for other water

column samples [Conte et a/. [2006], and references therein] and has been

attributed to the cell 's alkenone-based adaptation to temperature limitation at the

extremes of its growth temperature range [Conte et al. 1998]. However, when

the Conte et al. [2006] UK'3rSST relationship is superimposed onto Figure 8,

1.0 ,---------------:._~,.._e4 ... ~t_--__, " .. -., I 0.9 •

0 0.8 :;::

'" a:: ... M

;.e ::::> 0.7

0.6

• • ..... •• • •

• •

- Conte et al. [2006] • BelowSML .... SML

0.5 +----,----.----r---.----.----.----.----,----.---~

14 16 18 20 22 24 26 28 30 32 34

in situ Temperature (eC)

Figure 8. Relationsh ip between UK'3, and in situ temperatures of the SPM samples. The solid

line represents the calibration derived by Conte et al. [2006] , and coincided with the SPM

samples collected in the SML (within ±3"C). Generally, samples collected below the SML

deviated from the calibration. The source of th is deviation is addressed in the text.

40

there is good agreement only for samples collected from the SML (Fig. 8). Points

representing UK·37 values from samples collected below the SML fall

systematically to the left of the Conte et al. [2006] relationship. Since the present

study has determined that the alkenone producing algae in the Gulf of Califomia

are exporting from depths below the SML, this deviation from the temperature

calibration needs to be investigated.

Recent laboratory [Epstein et al. Prahl at a/. 2003] and field [Prahl at al.

2005] and Popp at al. 2006a] observations suggest UK·37 values can be affected

by growth under conditions of nutrient and light limitation. Therefore, an

assessment of the growth conditions for the alkenone-producing algae in the Gulf

of Califomia is needed. Photoperiod growth rates (P) for each incubation

experiment were determined and compared to the level of available

photosynthetic active radiation (PAR) (Fig. 9). Although Significant scatter is

apparent, J1 appears to be limited by light at PAR levels below 150±25 pEin m-2 s-

1. This level of irradiance falls within the range of saturating light intensities

typically rep'orted for E. Huxleyi of -50-300 pEin m-2 S-l [popp at a/. 2006b and

references therein]. Assuming growth of alkneone-producing algae is Iight­

limited below 150±25 pEin m-2 S-l, the depth in the water column at which light

limitation occurred for each cruise/station was determined and compared to the

predicted range of export depths (Table 9). Using this assumption, alkenone

export occurred from depths where light limited the growth rate of the alkenone­

producing algae for all summertime cruises and stations. Light levels above

41

1.6

• • GoCal12

• GoCal22 ~ • GoCal32 ~

Q) 1.2 -ro • GoCal33

• GoCal34 Il:: ~

~ 0 ~

(!) 0.8

-0 0

• • GoCal31

• • • • • • • . '1. • • ·c Q) C-o -0 ~ c.. 0.4

• • • • • • • • • •• • • • • • • • • N

'" M ~

s+ • • • • • •• • • •• • 0.0 •

o 100 200 300 400 500 600 700

Figure 9. Plot of K:.702 Photoperiod Growth Rate {jJ} vs. available light (PAR). The reduction of

the slope at - 150±25IlEin m·2 s·, ind icates that the algae are growing under light limitation below

this level of PAR.

Table 9. A Comparison of UK'37 Pred icted Temperature with in situ Temperature

Cruise/Station LL-Depth Export Depth K' ST. U 37-T Ave. in situ-T l!. T (m) (m) (0C) (0C) (0C)

GoCal12 11 20-25 270 23.1 3.9 GoCal22 24 20-30 17.7 18.6 -0 .9 GoCal32 24 25-30 26.9 23.2 3.7 GoCal33 21 20-25 26.6 22.4 4.2 GoCal34 21 25-30 27.4 27.0 0.4 GoCal31 20 25-30 27.7 26.1 1.6

where LL-Depth is the depth where light begins to limit growth of the algae, ST . UK'37- T

is the temperature derived from the Conte et a/. [2006] calibration for the sediment trap material, Ave. in situ-T is the average in situ temperature at the depth of export, and l!. T is defined as the difference between ST . UK

'3], T and Ave. in situ- T

42

150±25 pEin m"2 s"' fell within the depth range of alkenone export production only

during the winter cruise. A comparison of the sediment trap-derived UK"3T"Twith

the average in situ temperature at the depth of export shows that in situ

temperatures are indeed overestimated by UK·37 temperature predictions when

light is limiting growth (Table 9).

The magnitude at which UK" 3rderived growth temperatures overestimate

in situ temperatures during the summer cruises varies with respect to location

(-4°C at stations 2 and 3 and -1°C at stations 4&1) and is probably influenced by

the depth of the nutricline at that given station. Dissolved inorganic nitrogen

concentrations were below detection limits throughout the range of incubation

experiment depth (0-40 m) for GoCal34 and GoCal31 (see Fig. 4e-f). Assuming

that the alkenone-producing algae growing at these two stations are being

affected by nutrient limitation as well as light limitation, the overestimating effect

of light limitation is partially canceled out due to the underestimating effect of

nutrient limitation, resulting in a relatively small overestimate of in situ

temperature (-1°C). Dissolved inorganic nitrogen concentrations are low, but

detectable in GoCal12, GoCal32, and GoCal33 and therefore growth of the

alkenone-producing algae was not likely nutrient-limited, resulting in a relatively

large overestimate of in situ temperature (-3°C).

This overestimate of in situ temperature explains why UK·3T" Tfrom the

sediment traps did not underestimate SML-T as severely as predicted based on

the temperature difference between the SML and the depth of export (Section

43

3.2). The magnitude of the SML-T underestimate was reduced because light

limitation caused the UK' 3r T to overestimate in situ growth temperature. Since

favorable conditions for alkenone growth are often found below the SML

throughout the world ocean [Prahl et aI., 1993 and 2001; Temois et a/., 1997;

Ohkouchi et aI., 1999], the effects we document here may extend to other sites in

the worlds ocean.

CHAPTER 4 CONCLUSION

44

The depth of alkenone export production in the Gulf of California was

constrained to -20-30 m, which is well below the SML during the summer. As a

consequence, SST is systematically underestimated by UK' 37 collected in shallow

sediment traps. The magnitude of the underestimate falls within the range of

variability observed in most UK' 3r T relationships based on core-top samples

[Muller et a/. 1998; Conte et a/. 2006]. However, the discrepancy between UK'3r

temperature recorded in sediment trap material and SST was smaller than

expected based on the temperature gradient through the upper water column

and the depth of alkenone export production. It appears that the physiological

effect of growth under light limitation and its affect on UK' 37 values helped to

minimize the difference between UK'3rtemperature recorded in sediment trap

material and SST. Therefore, relatively accurate SST estimates can be obtained

from UK'37 measurements, even when the alkenone producing algae are

exporting from depths below the SML (within -±3°C). On the other hand,

recognition of the depth of export in ancient record could substantially improve

paleotemperature estimates using the alkenone unsaturation index.

Concurrent measurements of integrated production in the water column

and flux to shallow sediment traps allow for an estimate of the efficiency of

alkenone export production. Efficiencies ranged from 9-34% and averaged

-20%. Assuming an -50% loss due to limitations in the sediment trap collection

method, it is estimated that the remaining -60% of the alkenone production is

45

recycled through grazing or lost during transport (presumably to degradation or

lateral transport). Unfortunately. the relatively large fraction of production that is

not exported indicates that accurate estimates of paleoproduction could not be

made by measuring alkenone accumulation rates in sediments.

46 REFERENCES

1. Armstrong, F. A., Steams, J.R., and Strickland, J.H., 1976. The measurement of upwelling and subsequent biological processes by means of the Technicon AutoAnalyzer™ and associated equipment, Deep Sea Research 114: 381-389.

2. Atlas, E.L., Hager, S.W., Gordon, LI., and Park, P.K., 1971. A practical manual for the use of the Technicon AutoAnalyzer™ in seawater nutrient analysis: Revised, Tech Rep. 215 Ref. 71-22,48 pp., Dep. Of Oceanogr., Oregon State Univ., Corvallis.

3. Boon, J.J., Meer, F.W.V.D., Schuyl, P.J.W., deleeuw, J.W., and Schenck, P.A.,1978. Organic geochemical analysis of core samples from site 362 Walvis Ridge, DSDP leg 40. In Initial reports of the deep sea drilling project, Vol. 40, Bolli, H.M, et a/., editors, US Government Printing Office, Washington, pp. 627-637.

4. Brassell, S. C., Eglinton, G., Marlowe, I. T., Pflaumann, U., and Samthein, M., 1986. Molecular stratigraphy: a new tool for climatic assessment, Nature 320:129-133.

5. Buesseler et aI., 2007. An assessment of the use of sedimet traps for estimating upper ocean particle fluxes, Joumal of Marine Research 65:345-416.

6. Christie, W.W., 1973. Upid Analysis: Isolation, Separation. Identification. and structural Analysis of Upids, Elsevier, New York.

7. Conte, M.H. and Eglinton, G., 1993. Alkenone and alkenoate distributions within the euphotic zone of the eastern North Atlantic: correlation with production temperature, Deep Sea Research 140: 1935-1961.

8. Conte, M. H., Thompson, A., lesley, D., and Harris, R., 1998. Genetic and physiological influences on the alkenoneJalkenoate versus growth temperature relationship in Emiliania huxleyi and Gephyrocapsa oceanica, Geochim. Cosmochim. Acta, 62:51-68.

9. Conte, M.H., Sicre, M., Ruhlemann, C., Weber, J.C., Schulte, S., Schulz-Bull, D., Blanz, T., 2006. Global temperature calibration of the alkenone unsaturation index (UK'37) in surface waters and comparison with surface sediments, G3 vol. 7 no. 2. doi: 10.1 02912005GC001 054

10.Deines, P., langmuir, D., and Harmon, R., 1974. Stable carbon isotope ratios and the existence of a gas phase in the evolution carbonate ground waters, Geochim. Cosmochim. Acta. 38:1147-1164.

11.deleeuw, J.W., Meer, F.W.V.D., Rijpstra, W.I.C., and Schenk, P.A., 1980. On the occurrence and structural identification of long chain unsaturated alkenones and hydrocarbons in sediments. In: Advances in organic geochemistry 1979, Douglas, A.G. and Maxwell, J.R., editors, Pergamon Press, Oxford, pp. 211-217.

12. Dickson, A. G., and Millero. F. J., 1987. A comparison of the equilibrium constants for the dissociation of carbonic acid in seawater media, Deep­Sea Research 34:1733-1743.

47

13. Dickson, A.G., 1990a. Standard potential of the reaction: AgClis) + 1.2H2(g) = Ag(s) + 11 CI(aq), and the standard acidity constants of the ion HSO-4 in synthetic seawater 273.15 to 318.15°K, J. Chem. Thermodyn.22:113-127.

14. Dickson, A.G., 1990b. Thermodynamics of the dissociation of boric acid in synthetic seawater 273.15 to 318.15°K, Deep Sea Research, Part A 37:755-766.

15. Epstein, B. L., d'Hondt, S., and Hargraves, P. E., 2001. The possible metabolic role of C37 alkenones in Emiliania hux/eyi, Org. Geochem. 32:867-875.

16. Epstein, B. L., d'Hondt, S., Quinn, J. G., Zhang, J., and Hargraves, P. E., 1998. An effect of dissolved nutrient concentrations on alkenone-based temperature estimates, Paleoceanography 13:122-126.

17. Gaxiola-Castro, G., Alvarez-Borrego, S. A., Lavin, M. F., Zirino, A., and Najera-Martinez, S., 1999. Spatial variability of the photosynthetic parameters and biomass of the Gulf of California phytoplankton, Jouma/of Plankton Research 21:231-245.

18. Grice, K. et a/., 1998. Effects of zooplankton herbivory on biomarker proxy records, Paleoceanography 13:686-693.

19.Goni, M. A., Hartz, D. M., Thunell, R. C., and Tappa, E., 2001. Oceanographic considerations for the application of the alkenone-based paleotemperature UK" 37 index in the Gulf of California, Geochim. Cosmochim. Acta 65:545-557.

20.Gran, G., 1952. Determination of the equlivance point in potentiometric titrations, II. Analyst 77:661-671.

21.Hama, T., Hama, J., and Handa, N., 1993. 13C tracer methodology in microbial ecology with special reference to primary production processes in aquatic environments. In Advances in Microbial Ecology (ed. J. GWYNFRYN), pp. 39-83. Plenum Press.

22. Hamanaka, J., Sawada, K., and Tanoue, E., 2000. Production rates of C37 aiken ones determined by 13C-labeling technique in the euphotic zone of Sagami Bay, Japan, Organic GeochemistI}' 31:1095-1102.

23. Haxo, F. T., 1985. Photosynthetic action spectrum of the coccolithiophorid, Emiliania huxleyi (Haptophyceae): 19'-hexanolyoxyfucoxanthin as antenna pigment, Joumal of Phycology 21 :282-287.

24. Hayes, J. M., Freeman, K. H., Popp, B. N., and Hoham, C. H., 1990. Compound-specific isotopic analyses: a novel tool for reconstruction of ancient biogeochemical processes, Organic GeochemistI}' 16:1115-1128.

25.Johnson, K.M., Willis, K.D., Butler, WK, Johnson, WK, and Wong, C.S., 1993. Colulometric total carbon dioxide analysis for marine studies: maximizing the performance of an automated gas extraction system and coulometric detector, Mar. Chem.44:167-188.

48

26. Karl, D.M., Winn, C.D., Hebel, D.V.W., and Letelier, R, 1990. Hawaii Ocean Time- series Program Field and Laboratory Protocols, September 1990. School of Ocean and Earth Science and Technology, Univ. of Hawaii, Honolulu, HI, 72 pp.

27. Knappertsbusch, M., 1993. Geographic distribution of living and Holocene coccolithiophores in the Mediterranean Sea, Marine Micropa/entology 21:219-247.

28. Knauer, GA, Martin, J.H., and Bruland, K., 1979. Fluxes of particulate carbon, nitrogen, and phosphorous in the upper water column of the Northeast Pacific Ocean, Deep-Sea Research 26:97-108.

29.Kroopnick, P., 1985. The distribution of l3C in rC02 in the world oceans, Deep Sea Research 132:57-84.

30.Laws, EA, Popp, B.N., Bidigare, RR, Kennicutt, M.C., Macko, SA, 1995. Dependence of phytoplankton carbon isotopic composition on growth rate and [C02]aq: Theoretical considerations and experimental results, Geochim. Cosmochim. Acta. 59:1131-1138.

31.Lewis, E., and Wallace, D., 1998. Program Developed for C02 System Calculations. ORNUCDIAC-105. Carbon Dioxide Information Analysis Center, Oak Ridge National Laboratory, U.S. Department of Energy, Oak Ridge, Tennessee.

32. Malinvemo, E., Prahl, F.G., Popp, B.N., Ziveri, P., in prep. Alkenone abundance and its relationship to the coccolithiophore in Gulf of California surface waters.

33. Marlowe, I. T., Brassell, S. C., G. Eglinton, G., and Green, J. C., 1984a. Long chain unsaturated ketones and esters in living algae and marine sediments, in Advances in Organic Geochemistry 1983 (P. A. Schenck, J. W. de Leeuw, and G. M. W. Lijmbach, eds.), Organic Geochemistry, 6:135-141.

34.Marlowe, I. T., Green, J. C., Neal, A. C., Brassell S. C., Eglinton, G., and Course, P. A., 1984b. Long chain (n-C3rC39) alkenones in the Primnesiophyceae. Distribution of alkenones and other lipids and their taxonomic significance, Br. Phyco/. J., 19:203-216.

35. Mehrbach, C., C. H. Culberson, J. E. Hawley, and R M. Pytkowicz, . Measurement of the apparent dissociation constants of carbonic acid in seawater at atmospheriC pressure, Limnology and Oceanography 18, 897-907 (1973).

36. Mook, W.G., Bommerson, J.C., Staberman, W.H., 1974. Carbon isotopic fractionation between dissolved bicarbonate and gaseous carbon dioxide, Earth Planet Sci. Lett., 22: 169-176.

37.Muller, P. J., Kirst, G., Ruhland, G., von Storch, I., and Rosell-Mele, A., 1998. Calibration of the alkenone paleotemperature index UK' 37 based on core­tops from the eastern South Atlantic and the global ocean (600N-600S), Geochim. Cosmochim. Acta 62:1757-1772.

49

38.0hkouchi, N., Kawamura, K., Kawahata, H., Okada, H., 1999. Depth ranges of alkenone production in the central Pacific Ocean. Global Biogeochemical Cycles 13:695-704.

39. Pond, D. W., Harris, R P., 1996. The lipid composition of the coccolithophore Emiliania huxleyi and its possible ecophysiological significance. Joumal of the Marine Biological Association of the United Kingdom 76:579-594.

40.Popp, B. N., Prahl, F. G., Wallsgrove, R. J., and Tanimoto, J., 2006a. Seasonal patterns of alkenone production in the subtropical oligotrophic North Pacific, Paleoceanography, 21:P Al 004, doi: I 0.1029/200SP A00116S.

41. Popp, B.N. et a/., 2006b. A new method for estimating growth rates of alkenone-producing haptophytes, Umnologyand Oceanography: Methods 4:xx-xx.

42.Popp, B.N., Kenig, F., Wakeham, S.G., Laws, E.A., and Bidigare, R.R, 1998. Does growth rate affect ketone unsaturation and intracellular carbon isotopic variability in Emiliania hux/eyn Paleoceanography 13:35-41.

43. Prahl F. G., Pilskaln C. H., and Sparrow M. A., 2001. Seasonal record for alkenones in sedimentary particles from the Gulf of Maine, Deep-sea Res. 48:515-528.

44. Prahl, F. G., and Wakeham, S. G., 1987. Calibration of unsaturation patterns in long-chain ketone compositions for palaeotemperature assessment, Nature 330:367-369.

45. Prahl, F. G., Popp, B. N., Karl, D. M., and Sparrow, M. A., 2005. Ecology and biogeochemistry of alkenone production at subtropical North Pacific Station ALOHA, Deep-Sea Research 152:699-719

46. Prahl, F. G., Wolfe, G. V., Sparrow, M. A., 2003. Physiological Impacts on Alkenone Paleothermometry, Paleooceanography 18:1025-1032.

47. Prahl, F.G., Collier, RB., Dymond, J., Lyle, M., Sparrow, M.A., 1993. A biomarker perspective on prymnesiophyte productivity in the northeast Pacific Ocean, Deep-Sea Research 140:2061-2076.

48. Prahl, F.G., Muehlhausen, L.A., Lyle, M., 1989. An organic geochemical assessment of oceanographic conditions at MANOP Site C over the past 26,000 years, Paleoceanography 4:495-51 O.

49. Prahl, F.G., Rontani, J., Volkman, J.K., Sparrow, M.A., Royer, I.M., 2006. Unusual C35 and C36 alkenones in a paleoceanographic benchmark strain of Emiliania huxleyi, Geochimica et Cosmochimica Acta 70:2856-2867.

50. Rechka, J. A., and Maxwell, J. R, 1988a. Characterization of alkenone temperature indicators in sediments and organisms, in Advances in Organic Geochemistry 1987 (L. Mattavelle and L. Novelli, ads.) Organic Geochemistry 13:727-734.

51. Rechka, J. A., and Maxwell, J. R, 1988b. Unusual long chain keytones of algal origin, Tetrahedron Letters, 29:2599-2600.

52. Rost, B., Riebsell, U., Burkhardt, S., and Sultemeyer, D., 2003. Carbon acquisition of bloom-forming marine phytoplankton, Umno/. Oceanogr. 48:55-67.

50

53. Rost, B., Zondervan, I., Riebesell, U., 2002. Light-dependent carbon isotopic fractionation in the coccolithophorid Emiliania huxleyi, Umno/. Oceanogr. 47:120-128.

54.Schlenk, H., and Gellerman, J., 1960. Esterfication offatty acids with diazomethane on a small scale, Anal. Chem., 32:1412-1414.

55. Shin, K.-H., Tanaka, N., Harada, N., and Marty, J.-C., 2002. Production and turnover rates of C37 alkenones in the eastern Bearing Sea: Implications for the mechanism of a long duration of Emiliania huxleyi bloom, Prog. Oceanogr.55:113-129.

56. Strickland, J. and Parsons, T., 1972. A Practical Handbook of Seawater Analysis, 2nd ed., bull. 167,310 pp., Fish. Res. Board of Can., Ottawa, Ont.

57. Temois, Y., Sicre, M.-A., Boireau, A., Conte, M.H., Eglinton, G., 1997. Evaluation of long-chain alkenones as paleotemperature indicators in the Mediterranean Sea, Deep-Sea Research 144:271-286.

58. Thumell, R.C., 1998. Seasonal and annual variability in particle fluxes in the Gulf of California: A response to climate forcing, Deep Sea Research I 45:2059-2083.

59. Versteegh, G., Riegman, R., de Leeuw, J., Jansen, J., 2001. UK'37 values for Isochrysis galbana as a function of culture temperature, light intensity, and nutrient concentrations, Organic Geochemistry 32:785-794.

60.Ziveri, P., and Thurnell, R., 2000. Coccolithiophore export production in Guaymas Basin, Gulf of California: Response to climate forcing, Deep­Sea Research 1147:2073-2100.