 

'4

 

Illlllllllllllllllllllllllllllllllllllllllllllllllllllllll L

31293 01563 1090

This is to certify that the
thesis entitled
IsotOpic Biogeochemistry
of Dissolved Organic Nitrogen:

A New Technique and Application

presented by

Timothy P. Feuerstein

has been accepted towards fulﬁllment
of the requirements for

Masters degree in Geological Sciences

 

49W) War/v

Major professor

Date f/Z’L'Q'}

07639 MS U is an Afﬁrmative Action/Equal Opportunity Institution

 

 

LIBRARY

Michigan State
University

 

 

 

PLACE N RETURN BOX to remove this chockoui from your record.

To AVOID FINES Mum on or baton dot. duo.

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

ISOTOPIC BIOGEOCHEMISTRY OF DIS SOLVED ORGANIC NITROGEN:
A NEW TECHNIQUE AND APPLICATION

By

Timothy P. Feuerstein

A THESIS
Submitted to
Michigan State University
in partial ﬁilﬁllment of the requirements
for the degree of

MASTER OF SCIENCE

Department of Geological Sciences

1997

ABSTRACT

ISOTOPIC BIOGEOCHEMISTRY OF DIS SOLVED ORGANIC NITROGEN:
A NEW TECHNIQUE AND APPLICATION

By

Timothy P. Feuerstein

A new technique for isolating and isotopically characterizing dissolved organic
nitrogen (DON) for non-marine waters was developed. The technique involves
preconcentration of DON via rotary evaporation and removal of dissolved inorganic
nitrogen (DIN) via dialysis using a membrane that retains material above 100 daltons.
Results demonstrate quantitative removal of DIN and the complete recovery and retention
of isotopic integrity of DON. Nitrogen isotope values for DON ﬁ'om Chefswet Basin,
Lake Superior were compared to those of other nitrogen reservoirs to infer transformation
pathways. Within the euphotic zone, the Sl’N of seston (0.0%o-2.5%o) compared to that
of NO3' (-5.0%o- -3.5%o) suggests that NO3' assimilation is not the primary determinant of
the 61’N of seston. Uptake of l’N-enn'ched NHX or loss of 1"N through degradation could
account for the high 51’N values of seston. The low 515N value of DON (-1.2%o) relative
to seston is consistent with the idea that DON at the Lake Superior site is derived from the

degradation of seston.

To my future wife Nicole, whose support and patience has kept my spirit alive.

iii

ACKNOWLEDGMENTS

As this research was not ﬁmded by any speciﬁc grant, I would like to thank both
Peggy and Nathaniel Ostrom for “dipping” into their accounts to support my project,
which allowed for continued progress and a timely ﬁnish to my degree. A special thanks
to Peggy for her generous contribution of time to both the research and the development

of my scientiﬁc writing.

iv

TABLE OF CONTENTS

LIST OF TABLES .......................................................................................................... vi
LIST OF FIGURES ....................................................................................................... vii
INTRODUCTION... ...................................................................................................... 1
MATERIALS AND METHODS... ............................................................................... 4
Technique Overview ............................................................................................. 4
Evaluating DIN Removal ....................................................................................... 6
Recovery and Isotopic Integrity of DOM .............................................................. 7
Field Samples ....................................................................................................... 9
RESULTS... ................................................................................................................. 10
Evaluating DIN Removal .................................................................................... 10
Recovery and Isotopic Integrity of DOM. . . ....................................................... 13
Field Samples ..................................................................................................... 16
DISCUSSION. . . ........................................................................................................... 20
CONCLUSIONS. . . ...................................................................................................... 25
LIST OF REFERENCES ........................................................................................... 26

LIST OF TABLES

Table 1 - Carbon and nitrogen concentrations for procedural and reagent

blanks subjected to each of the three treatments (T); rotary evaporation,

dialysis, or the combined procedure and for non-treated (NT) and treated,

blank-corrected Glu-Gly-Phe (GGP) standards. Data are reported as average

values :i: standard deviation with the diﬂ‘erence between T and NT designated

by A; na = not applicable ................................................................................................ 14

Table 2 - The Sl’N and 613 C values for our standard Glu-Gly-Phe (GGP)

subjected to each of three treatments (T); rotary evaporation, dialysis or

the combined procedure. Data are reported as average values 2i: standard

deviation with the diﬂ‘erence between T and NT designated by A .................................... 17

Table 3 - The ﬁl’N and 513 C values and nitrogen and carbon elemental

concentrations for DOM ﬁom Grand Traverse Bay, Lake Michigan and

Cheﬁswet Basin, Lake Superior. Data are reported as average values i

standard deviation; nd = no data ..................................................................................... 18

LIST OF FIGURES

Figure l - Estimates of dissolved inorganic nitrogen removal; removal of
NH4+ from ultra-clean deionized water (Bamstead E-pure) ............................................. 11

Figure 2 - Estimates of dissolved inorganic nitrogen removal; removal of
NH4+ and N03" from a solution containing Glu-Gly-Phe (GGP) in ultra-clean
deionized water; removal of a NO; spike from natural lake water ................................... 12

Figure 3 - Temperature, ﬂuorescence (chlorophyll-a), concentrations of NH4+

and NOg' and 61’N values of N03’, particulate organic nitrogen (PON) and
dissolved organic nitrogen (DON) in Cheﬁswet Basin, Lake Superior ............................ 23

vii

INTRODUCTION

Dissolved organic nitrogen (DON) is often the most abundant form of nitrogen in
aquatic ecosystems (Sharp 1983). Previously DON was perceived to be largely refractory
with an age of approximately 6,000 years (Williams and Druﬂ‘el 1987). More recently,
numerous lines of evidence suggest a signiﬁcant portion of DON is bioactive and a
participant in biogeochemical cycles occurring on short time scales (Pomeroy et al., 1990;
Bronk and Glibert 1993; Jorgensen et al., 1994; Kroer et al., 1994; Bronk et al., 1994).
Uptake of DON by plankton (Bronk and Glibert 1993) and bacteria (Pomeroy et al., 1990;
Jorgensen et al., 1994; Kroer et al., 1994) suggest that this dissolved organic reservoir is
an available substrate for growth. The signiﬁcance of DON in nitrogen regeneration is
indicated by the observation that 25 to 41% of the dissolved inorganic nitrogen (DIN)
assimilated by phytoplankton is released as DON (Bronk et al., 1994). DON can be
utilized by phytoplankton (Palenik and Morel 1990a, b) or assimilated by bacteria (Keil
and Kirchman 1991, 1993; Wheeler and Kirchman 1986; Simon and Rosenstock 1992)

and incorporated back into the food web through the microbial loop (Bronk et al., 1994).

2
Several important issues regarding the role of DON in nitrogen cycling include:

(1) deﬁning rates of DON mineralization, (2) estimating the relative importance of
diﬁ‘erent types of biota in DON cycling, (3) understanding the individual processes with
the nitrogen cycle and how they are coupled and what type of feedback loops exist, (4)
determining which components of the DON pool are bioavailable and (5) assessing the
applicability of short-term rates from uptake experiments to models of DON cycling
relating to long-term phenomena. A few studies have begun to address these issues using
l’N enrichment approaches (Andersson et al., 1985; Nagata and Kirchman 1991, 1992;
Bronk and Glibert 1993; Gardner et al., 1993, 1996; Hoch et al., 1996). An alternative
method is the application of natural abundance stable isotope techniques.

Our interest is to apply natural abundance lsN measurements to assess several of
the current issues relating to the role of DON in nitrogen cycling. This approach should
be informative because measurements of natural abundance isotope ratios are a well-
established technique for identifying sources and evaluating transformations of materials
within the environment (Macko and Ostrom 1994). For example, near-zero ﬁl’N values
measured for plankton in regions of the North Paciﬁc where nitrogen-ﬁxing cyanobacteria
are abundant may reﬂect incorporation of atmospheric nitrogen (near O %o) because this
process because this process is accompanied by little or no isotopic fractionation (Wada
and Hattori 1976; Minagawa and Wada 1984). One advantage of the natural abundance

isotope approach is that the data reﬂect long-term processes operating

3

over time periods of weeks to months, rather than the scale of hours to days inherent in
enrichment studies. Currently there is only one study that documents natural abundance
stable isotope values for DON for the high molecular weight (DON > 1000 daltons)
fraction of this reservoir (Benner et al., in press).

We developed a technique that isolates the entire DON pool (100 daltons), and
likely minimizes molecular alteration, to obtain measurements of stable nitrogen isotope
values in non-marine waters. Data for veriﬁcation of the technique and preliminary BI’N
analyses for DON from ﬁeld samples are presented. We verify removal of DIN (in our
experiments NI-I4+ and/or N03") and the recovery and maintenance of isotopic integrity of
a DON standard (tripeptide Glu-Gly-Phe; GGP) at each step in the preparatory procedure.

We also evaluated reproducibility for isotope ratios and elemental concentrations of
natural samples. Data for natural samples include values of the nitrogen isotopic
composition of DON from the Chefswet Basin, Lake Superior and Grand Traverse Bay,
Lake Michigan as well as DON and DOC concentrations. Additional water column data
on other pools of nitrogen in Lake Superior are presented to assist in the interpretation of

the DON results and the nitrogen cycling dynamics of this system.

MATERIALS AND METHODS
Technique Overview

Our technique involves ﬁltration, rotary evaporation and dialysis. The following
describesthe methods used throughout all the experiments mentioned below. A sample of
100-200 mL was ﬁltered through an acid-washed 0.2 pm polysulfone Supor membrane
(Gelrnan Sciences) and concentrated to approximately 5 mL by rotary evaporation. This |
process was carried out at ca. 10 mm of Hg and a temperature of less than 40 °C. After
rotary evaporation the sample was transferred to a dialysis membrane (Spectra/Por CE
Membrane) with a nominal molecular weight cutoﬂ‘ (MWCO) of 100 (daltons). The
concentrate was diaﬁltered at 4 °C against a salt solution (5% NaCl, 6.5% MgSO4-7H20
and 4% CaClz; w/v) for 25 hours and ultra-clean deionized water (Barnstead E-pure) for
an additional 75 hours during which time the dialysis solution was changed several times.
The sample was then transferred to a pre-combusted (500 °C, 3 hr) quartz ampule (ca.

15 mL). The dialysis membrane was washed with 2 mL of deionized water and the wash
was added to the ampule to facilitate quantitative transfer. The sample was dried in-vacuo
on a glass high-vacuum line. Ifthe carbon isotope composition was desired, phosphoric

acid (100 mL) was added to the sample prior to evacuation to remove

dissolved inorganic carbon (DIC).

Upon dryness, the sample was removed from the vacuum line, precombusted
copper oxide (Aldrich; 3 g) and pure cOpper (Alpha Resources; 3 g) were added and the
sample re-evacuated. Precombusted copper oxide is obtained by spreading a thin layer of
this material in a porcelain dish and heating it to 850 °C for 4 hours; precombusted copper
is obtained by heating an evacuated, sealed quartz ampule (30 cm x 10 mm o.d.) ﬁlled
with CuO and Cu (ca. 1:6 by volume) to 850 °C for 4 hours. Aﬁer re-evacuating, the
sample was sealed under vacuum, heated to 850 °C and allowed to cool gradually
following a modiﬁed Dumas combustion procedure (Macko, 1981). Nitrogen gas and
CO; were separated cryogenically ﬁom the combustion products and isotope values of .
puriﬁed gases were determined on a PRISM (Micromass) mass spectrometer. The
concentration of carbon in the sample was determined using a Baratron capacitance
manometer (MKS Inc.) during cryogenic gas separation. Nitrogen concentration was
determined based on the ion beam produced (mass 28) within a calibrated volume of the

mass spectrometer. Nitrogen and carbon isotope ratios are expressed in per mil:

SE = [W) - 1] * 1000

where, I is the heavier isotope of element E and R is the abundance ratio of the heavy to

light isotope. The internationally recognized standards for BI’N and 613C are

6

atmospheric nitrogen gas and V-Peedee Belemnite, respectively.

Evaluating DIN Removal

Using our dialysis technique we tested for the removal of DIN (NI-I4+ and/or N03'
) amended to (1) deionized water, (2) a solution containing our DON standard, GGP and
(3) natural waters. Concentrations of nitrogen in these experiments were chosen to mimic
samples containing high concentrations of DIN (100 uM) and/or natural abundance
concentrations of DON (20 uM). Removal of DIN was veriﬁed by determining the
concentration of DIN in the permeate (solution external to the dialysis membrane) at
different time intervals during dialysis and/or by evaluating the concentration of DIN in the
retentate (solution inside the dialysis membrane at the end of the experiment). In the ﬁrst
experiment 5 mL of 100 M NILCI was diaﬁltered against 140 mL of deionized water
while stirring at 4°C. Aliquots of permeate (5 mL) were transferred to scintillation vials
(25 mL) and frozen immediately. At the completion of the experiment, the retentate was
transferred to scintillation vials, aliquots of permeate samples were thawed, and NHI
concentrations were determined using an Orion ion speciﬁc electrode (Garside et al.,
1978). The detection limit and precision of this technique is 0.1 uM (Ostrom et al., in
press).

In the second experiment a 5.25 mL solution containing 100 M NH4Cl, 100 M

7
KNO; and 0.76 mM N from GGP was diaﬁltered and sampled as mentioned above.

Nitrate concentrations were determined by suppresser based anion chromatography
(Shipgun and Zolotov 1988) on a high performance liquid chromatograph (Rainin
Instruments) with conductivity detection (LDC Analytical). Separation of anions was
achieved on a Dionex IonPac column (#AS4A-SC) with suppresser using an eluent
consisting of 2.4 mM Na2C03 and 3 mM NaHCO3. The percent NO; removal was
estimated by analyzing the abundance of NO; in the permeate at diﬁ‘erent time intervals
during dialysis and by measuring its abundance in the retentate at the end of dialysis. The
percent removal of NH; was estimated by measuring its abundance in the retentate at the
end of dialysis using the Orion electrode.

For the last experiment water from Lake Lansing, Michigan was collected at
the surface in 1 L Nalgene bottles and ﬁltered. A 200 mL sample was amended with
KNO; such that the concentration of NO; was 100 M. The sample was subsequently
concentrated to 5 mL by rotary evaporation and diaﬁltered. Percent removal of NO; and

NH,+ were estimated by measuring their abundance in the retentate at the end of dialysis.

Recovery and Isotopic Integrity of DOM

We evaluated the recovery and isotopic integrity of GGP by determining the

elemental concentration and isotope ratios of nitrogen and/or carbon in GGP standards

8

before and after rotary evaporation, dialysis and the combined procedure. All experiments
were mm in triplicate and values were corrected for reagent blanks. Isotope values and
elemental abundances for procedural blanks (dialysis: 5 mL deionized water; rotary
evaporation: 100 mL deionized water; combined procedure: 100 mL deionized water)
were also determined. Given the low molecular weight of GGP (351.4 g/mole), this
standard offered a rigorous test of our ability to retain DOM without isotopic
discrimination.

For the rotary evaporation experiment, 2 mL of a 1.5 mM GGP standard was
diluted to 100 mL with deionized water and subsequently concentrated to ca. 5 mL. This
solution was dried in-vacuo, analyzed isotopically and the results compared to 2 mL of a .
1.5 mM GGP standard that was not subjected to rotary evaporation (non-treated GGP
standard). For dialysis 2 mL of the 1.5 mM GGP standard was placed in the dialysis
membrane and diaﬁltered against both the salt solution and deionized water for 25 hours
each. The retentate was dried, analyzed isotopically and compared to its non-treated
standard. The standard was also subjected to both procedures in sequence for isotopic
comparison to the non-treated standard. The recovery of the standard was determined by

comparing carbon or nitrogen concentrations for the treated and non-treated standards.

Field Samples

Water column samples were collected from Grand Traverse Bay, Lake Michigan
on the RV Northwestern (August 1996) and from the Cheﬁswet Basin, Lake Superior on
the RV Edwin Link (June 1994). The water column at each station was initially
characterized by deployment of a SBE-25 conductivity-temperature-depth proﬁler
equipped with a ﬂuorometer and transmissometer (SeaBird Electronics Inc). All water
samples were collected using 5 L lever-action Niskin bottles (General Oceanics) and
transferred to acid-washed l L Nalgene bottles and stored on ice. Samples were then
ﬁltered through an acid-washed 0.2 um polysulfone Supor membrane in lab within a few
hours of collection and frozen in Nalgene bottles priot to isolation and analysis of DON. -
Methods for sample preparation and isotopic analysis of particulates and DIN from Lake

Superior are reported in Ostrom et al., (in press).

RESULTS

Evaluating DIN Removal

We evaluated the removal of DIN (NI-14+ and/or N03) during dialysis ﬁom 1)
5 mL of deionized water amended with 0.5 moles of NH4CI (Figure 1), 2) 5.25 mL of
deionized water amended with 0.53 moles of NILC], 0.53 moles of KNO3 and
4 umoles (N) of GGP (Figure 2) and (3) 5 mL of water from Lake Lansing, Michigan
that was concentrated by rotary evaporation from an initial solution of 100 mL of lake
water that was spiked with 10 umoles of KNOs (Figure 2). Each of these experiments
were performed in triplicate. The procedural reproducibility (standard deviation about the
mean) for measurements of NIL+ abundance in the retentate or permeate was 0.03 umoles
or less and for that NOg' was 0.02 umoles or less.

At the end of the experiment involving NH4Cl in deionized water, estimates of
NH: concentrations in the permeate indicate greater than 100% removal and those in the
retentate suggests nearly 100% removal. The presence of a a small excess of NH4+ in the

permeate imply minor NH4+ contamination external to the dialysis membrane.

10

11

 

 

V

 

CurmulatlvePeroentNH Removed

 

 

 

 

Time (h)

Figure 1- Estimates of dissolved Inorganic nitrogen removal:
removal of NH 4from ultra-clean deionized water
(Barnstead E-pure)

12

 

10° Estimate of NH +Fiemoved —v———D-A.

.....................................................................................

8

 

Cunmalative Percent DIN Removed
8

 

 

 

4o - “retentate, lake water)
20 ............. . .........................................................................
EfErgtjnmraattelﬁwofl NwaIé‘oved (permeate. aid)
1 l 1 l l \x
0 10 20 30 40 50 100
nmem)

Figure 2- Estimates of dissolved inorganic nitrogen removal:

removal of NH 1' and N03 from a solution containing Glu-Gly- -Phe
(GGP) in ultra-clean deionized water; removal of a NO 3 spike

13

In the experiment that evaluated the removal of NH] and NO; from a solution containing
the GGP standard, measurements of NO; in the permeate and retentate at the end of the
experiment are in good agreement and show that more than 90% of the nitrate was
removed. Similarly, more than 95% of the NH4+ was removed based on estimates of the
concentration in the retentate.

Removal of N03' from water samples ﬁom Lake Lansing were in excess of 95%.

Concentrations of NIH in this sample following dialysis were below our limit of detection

(0.1 uM).

Recovery and Isotopic Integrity of DOM

Our ﬁrst step in addressing recovery during rotary evaporation and dialysis was to
evaluate the elemental concentration of both reagent and procedural blanks (Table 1). The
reagent blank for carbon was 1.4 umoles while that for nitrogen was below our limit of
detection (0.2 umoles). The contribution of carbon ﬁ'om the procedural blank for the
combined rotary evaporation/dialysis procedure was 1.5 umoles. A comparison of
procedural blanks for rotary evaporation (0.1 umoles) and dialysis (1.1 uM) indicates that
the majority of this carbon was derived during dialysis. The contribution of nitrogen ﬁ'om

individual steps or the combined procedure was below our limit of detection

14

Table l - Carbon and nitrogen concentrations for procedural and reagent blanks subjected
to each of the three treatments (T); rotary evaporation, dialysis, or the combined
procedure and for non-treated (NT) and treated, blank-corrected Glu-Gly-Phe (GGP)

standards. Data are reported as average values :l: standard deviation with the diﬂ‘erence

between T and NT designated by A; na = not applicable.

 

 

Procedure NT T A
(umole C or N) (umole C or N) (umole C orN)

Carbon Recovery
Reagent blank 1.4 na na
Rotary evaporation

procedural blank na 0.1 0.1

GGP standard 22.5 i 0.3 22.6 i 0.9 0.1
Dialysis

procedural blank na 1.1 1.1

GGP standard 23.6 :l: 0.1 24.0 i 0.1 0.4
Combined Procedure

procedural blank na 1.5 1.5

GGP standard 25.0 i 0.0 25.1 i 1.4 0.1
Nitrogen Recovery
Combined Procedure

procedural blank na <0.2 <0.2

GGP standard 2.0 i 0.2 1.8 i 0.2 0.2

 

15

(0.2 umoles).

The next step was to evaluate the recovery of GGP. This was achieved by
comparing the concentration of carbon or nitrogen in an aliquot of a standard subjected to
a procedure (treated) to that of an equivalent aliquot that had not been submitted to the
procedure (non-treated). The difference in the concentration of carbon between the blank-
corrected treated and non-treated standard for rotary evaporation, dialysis, or the
combined procedure was 0.4 umoles or less (Table 1). Standards subjected to the
combined procedure had nitrogen concentrations that differed by 0.2 umoles from the
non-treated standard. Differences in elemental concentrations between the treated and
non-treated standard are small (less than or equal to the standard deviation associated with
measurement; Table l) and, therefore, indicate complete recovery.

To assess isotopic integrity during rotary evaporation and dialysis we began by
evaluating the isotope values of reagent and procedural blanks. We were unable to obtain
nitrogen isotope ratios for blanks because they were below our limit of detection.
Consequently, no isotopic correction was applied to 8”N values of samples. The average
513 C value for the reagent blank was ~24. 1%o (n=2) and that for the blank associated with
the combined procedure was -28.7%o (n=2). Given isotopic similarity of the blank to that
of natural samples (see section below on ﬁled samples) and the low abundance of carbon

contributed by the blanks, corrections were not applied to 513 C values of natural samples.

16

After assessing blanks, we evaluated the isotopic integrity of GGP during each of our
procedures (Table 2). Differences in 513C values between the treated and non-treated
standard were no more than 0.4%o for rotary evaporation, dialysis or the combined
procedure. The data for the combined procedure shows no difference in nitrogen isotope
ratios between the treated and non-treated standard. We therefore concluded that there

was essentially no procedural effect on either the 6‘5N or 613C values.

Field Samples

We determined the elemental concentrations and isotope values of DOM in
samples collected from the Cheftswet Basin, Lake Superior and Grand Traverse Bay, Lake
Michigan (Table 3). The concentration of DON was 11.0 :h 0.8 IIM (n=3) and 6.8 i 0.2
IIM (n=4) in the Lake Michigan and Lake Superior locations, respectively. The
concentration of DOC at the Lake Superior site was 126 d: 4 M (n = 4). The coeﬁcient
of variation for DON concentrations at the location in Lake Michigan was less than 8%
and in Lake Superior it was less than 5% for both DON and DOC.

The average SI’N values for DOM at the Lake Michigan and Lake Superior sites
were 5.0 and -1.2, respectively. The reproducibility for the ﬁl’N values of DOM from

Lake Michigan was 0.1%. In Lake Superior, where nitrogen concentrations were about

17

Table 2 - The 61’N and 613C values for our standard Glu-Gly-Phe (GGP) subjected to
each of the three treatments (T); rotary evaporation, dialysis or the combined procedure.
Data are reported as average values i standard deviation with the difference between T
and NT designated by A.

 

Procedure n NT T A
(61’N or 613C) (Sl’N or 813C)

 

Nitrogen isotope data

Rotary evaporation 3 -0.6 :l: 0.0 -0.4 i 0.0 0.2
Dialysis 3 -4.3 i 0.3 -3.9 :t 0.1 0.4
Combined procedure 3 -0.7 :l: 0.4 -0.7 i 0.1 0.0

Carbon isotope data

Rotary evaporation 3 -19.4 :l: 0.0 -l9.4 :t 0.0 0.0
Dialysis 3 -24.9 i 0.2 -25.3 i 0.2 0.4

Combined procedure 3 -l9.5 i 0.1 -19.2 i 0.1 0.3

 

18

Table 3 - The 6”N and 813 C values and nitrogen and carbon elemental concentrations for
DOM from Grand Traverse Bay, Lake Michigan and Cheftswet Basin, Lake Superior.

Data are reported as average values i standard deviation; nd = no data.

 

Location 5"N (9a.) 513cm.) uMN uMC

 

Grand Traverse Bay, 5.0 i 0.1 nd 11.0 i 0.8 nd
Lake Michigan (n=3)

Cheﬁswet Basin, -1.2 i 0.3 -24.6 :l: 0.1 6.8 i 0.2 126 i 4
Lake Superior (n=4)

 

19
half that in Grand Traverse Bay, reproducibility of the isotope values of DOM were 0.3%:

for nitrogen and 0.1% for carbon.

DISCUSSION

Applications of natural abundance stable isotope measurements to natural waters
oﬁ‘er the potential to provide new perspectives for better understanding nitrogen cycling.
This approach overcomes some of the constraints imposed by 15N enrichment incubation
experiments such as bottle effects and the use of artiﬁcial substrates. In addition, natural
abundance data reﬂect long-term, in-situ phenomena and provide an independent
assessment of the source and predominant mechanisms that control nitrogen cycling.

The present study demonstrated the use of rotary evaporation and dialysis as an
effective method for the isolation and isotopic characterization of DON from ﬁ'eshwater
environments. Veriﬁcation of our procedure included excellent data for the removal of
DIN (greater than 90%), recovery of DON (ca. 100%) and the maintenance of isotopic
integrity of DOM (isotopic discrimination less than 0.4%o for ﬁl’N or 613C). We are
currently attempting to modify this procedure for marine waters.

There are several positive attributes of our technique. The low temperatures and
pressures encountered in rotary evaporation minimize possibilities for molecular alteration.
This is clearly an advantage if molecular characterization is of interest. Considering the

length of the isolation process, the low temperatures used in dialysis

20

21

were important to restrict microbial growth. Minimal user intervention and exposure of
the sample assisted in minimizing contamination.

Despite the emergence of tangential-ﬂow ultraﬁltration as a method for isolating
and concentrating DOM, we encountered several problems with ultraﬁltration in a
preliminary comparison to our technique. The most eminent problems with the tangential-
ﬂow technique was adsorption of DOM to the membrane (Molecular/For 100 MWCO
Ultraﬁltration Type C) and reduced ﬁltration rates. In addition to poor recoveries of
standards, adsorption of DOM was clearly apparent owing to the discoloration of the
membrane caused by a yellow nitrofurantoin standard (MW 238.2). Studies employing
tangential-ﬂow ultraﬁltration for the isolation of DOM commonly use membranes with a.
MWCO of 1000 and indicate that 20-40% of the organic carbon in seawater is recoverable
(Benner et al., 1992; Guo et al., 1995). These recoveries likely reﬂect the actual carbon
abundance of the high molecular weight (>1000) reservoir in addition to smaller amounts
of adsorption. This is because the larger MWCO membrane used in these studies may
reduce adsorption and associated concentration polarization. Our observations suggest,
however, that the tangential-ﬂow ultraﬁltration technique will not be appropriate for
isolation of the entire DOM reservoir.

The preliminary application of our technique was in the Great Lakes. The
difference in SI’N values of DOM between Grand Traverse Bay and Lake Superior (6.1%o,

Table 3), likely reﬂects unique processes and nitrogen inputs associated with

22

each location. Additional water column data from Lake Superior (Ostrom et al., in press)
was used together with these 5"N values of DOM to assist in a detailed interpretation of
this site.

Within Lake Superior, temperature and ﬂuorescence data show a stratiﬁed water
column and a maximum in chlorophyll-a at approximately 20 m (Figure 3). The
concentrations of NH4+ and NO; were approximately 0.5 M and 24 uM, respectively.
The concentration of NO; near the surface was slightly lower (ca. 1.5 uM) than it was
below 20 m (Figure 3), likely indicating uptake by phytoplankton above 20 m.

Nitrate in Lake Superior was characterized by the lowest 6”N values ever
reported for an aquatic system with an average of -4. 196:; (Figure 3; Liu and Kaplan 1989)
and is thought to have an atmospheric origin (Ostrom et al., in press). Ostrom et al. (in
press) believe that nitrate is not derived from the sediments or water column because in
the sediments, ﬂuxes of nitrate ﬁom Great Lakes are low and in the water column, low
isotope values derived ﬁom fractionation during nitriﬁcation are not expected at the low
NH4+ concentrations present in Lake Superior. The predominance of precipitation to the
hydrodynamic balance of Lake Superior and the low isotope values of N03’ are consistent
with an atmospheric origin (Heaton 1987; Paerl and Fogel 1994; Ostrom et al., in press).

With the exception of one data point, SI’N values for seston were within the range

of 0 to 2.5%» (Figure 3). Comparison of seston 6‘5N values to those of N03“ (-5.0 to

23

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

\I\i\\1\\i \ \ \ \r \\ \\.\\. \\\\L\\\W\\
. . . f .
8 .n I
344338819 m m m n M
. _ . _ . .u _ . _ __._ __h_. h _
.pno.....oooo~ o a no 8 .a we a
Peggiv 2538...; 238?; .. m. 2633

39.8 o- 323:8. 38.48838 Aggro? 8:833:03 0.. 21% 3o
20o mam. 538 on. 20 u. peace-m8 0693.0 :ﬁoco: $05 one undo—goo

0693.0 3800: 303. 5 0339.6. moan. go 9630:

24

-3.5%o) suggests that assimilation of NO; is not the predominant factor controlling the
isotopic composition of seston. Fractionation during assimilation should produce seston
with SI’N values equal to or less than that of the substrate, yet the 8”N values of our
seston were always higher than those of nitrate (W ada and Hattori 197 8; Macko et al.,
1987; Pennock et al., 1988; Cifuentes et al., 1989). This insinuates that processes in
addition to uptake and metabolism of nitrate are controlling the 6’5N of seston.

The nitrogen isotope ratios for seston could be explained by either the uptake of
15N enriched NIL+ or the loss of 1‘N through mechanisms that leave the residual reservoir
of seston enriched in 1’N. Processes that could lead to 15N enrichment include
microzooplankton grazing, microbial degradation, peptide bond hydrolysis or the loss of.
DON. A preliminary result indicates a 61’N value for DON that is lower than those of
seston is consistent with the interpretation that DON in Lake Superior is derived ﬁ'om
seston. Therefore the high SI’N values for seston may, in part, be explained by losses of
lsN—depleted DON. A better understanding of the origins of DON will be possible with a

more extensive data set.

CONCLUSIONS

The development of techniques to isolate and isotopically characterize DON offers
an important advance for understanding nitrogen dynamics in natural waters. With this
ability we are now able to quantify and isotopically characterize nitrogen from all
reservoirs in aqueous media. We envision studies whose emphasis is on obtaining temporal
data sets with the objective of constraining the large number of individual biogeochemical
processes that can inﬂuence the nitrogen cycle. The application of natural abundance
isotope tracing methods on multiple nitrogen reservoirs in such a framework offers an
excellent opportunity to determine the predominant controls on nitrogen cycling. We
believe temporal isotopic data sets will allow us to better address several outstanding
issues such as how individual biogeochemical processes are coupled and what types of
feedback loops exist. Furthermore, the technique we describe is particularly suited for an
important next step, one that will involve isotopic characterization of individual size

classes of DOM to assess their roles in nitrogen cycling dynamics.

25

LIST OF REFERENCES

LIST OF REFERENCES

Andersson A., Lee C., Azam F. and Hagstrom A (1985) Release of amino acids and
inorganic nutrients by heterotrophic marine microﬂagellates. Mar. Ecol. Prog.
Ser. 23, 99-106.

Benner R., Pakulski J.D., McCarthy M., Hedges II. and Hatcher PG. (1992) Bulk
chemical characteristics of dissolved organic matter in the ocean. Science 255,
1 561-1 564.

Benner R., Biddanda B., Black B. and McCarthy M. (1997) Abundance, size
distribution, and stable carbon and nitrogen isotopic compositions, of marine
organic matter isolated by tangential-ﬂow ultraﬁltration. Mar. Chem. (in press).

Bronk DA. and Glibert PM. (1993) Application of a 15N tracer method to the study
of dissolved organic nitrogen uptake during spring and summer in Chesapeake
Bay. Marine Biology. 115, 501-508.

Bronk D.A., Glibert PM. and Ward BB. (1994) Nitrogen uptake, dissolved organic
nitrogen release, and new production. Science 265, 1843-1846.

Cifuentes L.A., Fogel M.L., Pennock IR. and Sharp J.H. (1989) Biogeochemical factors
that inﬂuence the stable nitrogen isotope ratio of dissolved ammonium in the
Delaware Estuary. Geochim. Cosmochim. Acta 53, 2713-2721.

Gardner W.S., Cotner J.B. Jr. and Herche LR. (1993) Chromatographic measurement
of nitrogen mineralization rates in marine coastal waters with 15N . Mar. Ecol.
Prog. Ser. 93, 65-73.

Gardner W.S., Benner R., Amon R.M.W., Cotner J.B. Jr., Cavaletto IF. and Johnson
IR. (1996) Effects of high-molecular-weight dissolved organic matter on
nitrogen dynamics in the Mississippi River plume. Mar. Ecol. Prog. Ser. 133,
287-297.

26

27

Garside C., Hull G. and Murray S. (1978) Determination of submicromolar
concentrations of ammonia in natural waters by a standard addition method using
a gas-sensing electrode. Limnol. Oceanogr. 23, 1073-1076.

Guo L., Santschi PH. and Warnken K.W. (1995) Dynamics of dissolved organic carbon
(DOC) in oceanic environments. Limnol. Oceanogr. 40, 1392-1403.

Heaton THE. (1987) l’N/“N ratios of nitrate and ammonium in rain at Pretoria, South
Africa. Atmospheric Environment 21, 843-852.

Hoch M.P., Snyder R.A., Cifuentes LA and Coﬁin RB. (1996) Stable isotope
dynamics of nitrogen recycled during interactions among marine bacteria and
protists. Mar. Ecol. Prog. Ser. 132, 229-239.

Jorgensen N.O.G., Kroer N. and Coffin RB. (1994) Utilization of dissolved nitrogen
by heterotrophic bacterioplankton: Effect of substrate C/N ratio. Appl. Environ.
Microbiol. 4124-4133.

Keil R. G. and Kirchman D.L. (1991) Contribution of dissolved free amino acids and
ammonium to the nitrogen requirements of heterotrophic bacterioplankton. Mar.
Ecol. Prog. Ser. 73, 1-10.

Keil R. G. and Kirchman D.L. (1993) Dissolved combined amino acids: chemical form
and utilization by marine bacteria. Limnol. Oceanogr. 38, 1256-1270.

Kroer N., Jorgensen N.O.G. and Coffin RB. (1994) Utilization of dissolved nitrogen
by heterotrophic bacterioplankton: a comparison of three ecosystems. Appl.
Environ. Microbiol. 4 1 16-4 123.

Liu K.K. and Kaplan LR. (1989) The eastern tropical Paciﬁc as a source of 1’N—
enriched nitrate in seawater off southern California. Limnol. Oceanogr. 34, 820-
830.

Macko SA. (1981) Stable nitrogen isotope ratios as tracers of organic geochemical
processes. PhD. Diss., Univ. Texas, Austin. 181 pp.

Macko SA. and Ostrom NE. (1994) Molecular and pollution studies using stable
isotopes. In Stable isotopes in ecology and environmental science (Edited by
Lajtha K. and Michner R.). Blackwell Scientiﬁc Publications, Oxford 45-62.

28

Macko SA, Estep M.F., Hare PE. and Hoering TC. (1987) Isotopic fractionation
of nitrogen and carbon in the synthesis of amino acids by microorganisms. Chem.
Geol. 65, 79-92.

Minagawa M. and Wada E. (1984) Stepwise enrichment of 1’N along food chains:
ﬁrrther evidence and the relation between S‘SN and animal age. Geochim.
Cosmochim. Acta 48, 1135-1140.

Nagata T. and Kirchman D.L. (1991) Release of dissolved free and combined amino
acids by bacteriovorous marine ﬂagellates. Limnol. Oceanogr. 36, 433-443.

Nagata T. and Kirchman D.L. (1992) Release of dissolved organic matter by
heterotrophic protozoa: implications for microbial food webs. Arch. Hydrobiol.
Beth. 35, 99-109.

Ostrom N.E., Long D.T., Bell EM. and Beals T. (1997) The origin and cycling of
particulate and sedimentary organic matter and nitrate in Lake Superior. Chem.
Geol. (in press).

Paerl H.W. and Fogel ML. (1994) Isotopic characterization of atmospheric inputs as
sources of enhanced productivity in coastal Atlantic Ocean. Marine Biology
119, 635-645.

Palenik B. and Morel F.M.M. (1990a) Amino acid utilization by marine phytoplankton:
a novel mechanism. Limnol. Oceanogr. 35, 260-269.

Palenik B. and Morel F.M.M. (1990b) Comparison of cell-surface L-arnino acid
oxidases from several marine phytoplankton. Mar. Ecol. Prog. Ser. 59, 195-201.

Pennock J.R., Sharp J.H., Ludlam 1., Velinsky DJ. and Fogel ML. (1988) Isotopic
fractionation of nitrogen during the uptake of NH4+ and N03' by Skeletonema
costatum. EOS 69, 1098.

Pomeroy L.R., Macko S.A., Ostrom PH. and Dunphy J. (1990) The microbial food
web in Arctic seawater: Concentration of dissolved free amino acids, and
bacterial abundance and activity in the Arctic Ocean and Resolute Passage. Mar.
Ecol. Prog. Ser. 61, 31-40.

29
Sharp J.H. (1983) The distribution of inorganic nitrogen and dissolved and particulate
organic nitrogen in the sea. In Nitrogen in the Marine Environment (Edited by
Carpenter E]. and Capone D.G.). Academic Press, New York, 1-35.

Shipgun O.A. and Zolotov Y.A. (1988) Ion chromatography in water analysis. John
Wiley and Sons, New York, 188 pp.

Simon M. and Rosenstock B. (1992) Carbon and nitrogen sources of planktonic
bacteria in Lake Constance studied by the composition and isotope dilution of
intracellular amino acids. Limnol. Oceanogr. 37, 1496-1511.

Wada E. and Hattori A (1976) Natural abundance of 1’N in particulate organic matter
in the North Paciﬁc Ocean. Geochim. Cosmochim. Acta 40, 249-251.

Wada E. and Hattori A. (197 8) Nitrogen isotope effects in assimilation of inorganic
nitrogenous compounds by marine diatoms. Geomicrobiol. J. 1, 85-101.

Wheeler PA and Kirchman D.L. (1986) Utilization of inorganic and organic nitrogen ‘
by bacteria in marine systems. Limnol. Oceanogr. 31, 998-1009.

Williams PM. and Druffel E.R.M. (1987) Radiocarbon in dissolved organic matter in
the Central North Paciﬁc Ocean. Nature 330, 246-248.

"Illlllllllllllllllllllll“