Stable Isotopes in Ecosystem Studies

Author(s): Bruce J. Peterson and Brian Fry

Source: Annual Review of Ecology and Systematics , 1987, Vol. 18 (1987), pp. 293-320
Published by: Annual Reviews

Stable URL: https://www.jstor.org/stable/2097134

REFERENCES
Linked references are available on JSTOR for this article:
https://www.jstor.org/stable/2097134?seq=1&cid=pdf-
reference#references_tab_contents
You may need to log in to JSTOR to access the linked references.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms

Annual Reviews is collaborating with JSTOR to digitize, preserve and extend access to Annual
Review of Ecology and Systematics

This content downloaded from

140.122.143.127 on Sat, 14 Jan 2023 08:49:08 UTC
All use subject to https://about.jstor.org/terms
Ann. Rev. Ecol. Svst. 1987. 18:293-320
Copyright ?) 1987 bSo Annuiical Reviews Inc. All rights reserved

STABLE ISOTOPES IN ECOSYSTEM

STUDIES

Bruce J. Peterson and Brian Fry

The Ecosystems Center, Marine Biological Laboratory, Woods Hole, Massachusetts

02543

INTRODUCTION

The use of stable isotopes to solve biogeochemical problems in ecosystem

analysis is increasing rapidly because stable isotope data can contribute both
source-sink (tracer) and process information. The elements C, N, S, H, and 0
all have more than one isotope, and isotopic compositions of natural materials
can be measured with great precision with a mass spectrometer. Isotopic
compositions change in predictable ways as elements cycle through the
biosphere. These changes have been exploited by geochemists to understand
the global elemental cycles. Ecologists have not until quite recently employed
these techniques. The reasons for this are, first, that most ecologists do not
have the background in chemistry and geochemistry to be fully aware of the
possibilities for exploiting the natural variations in stable isotopic composi-
tions, and second, that stable isotope ratio measurements require equipment
not normally available to ecologists. This is unfortunate because some of the
more intractable problems in ecology can be profitably addressed using stable
isotope measurements. Stable isotopes are ideally suited to increase our
understanding of element cycles in ecosystems.
This review is written for ecologists who would like to learn more about
how stable isotope analyses have been and can be used in ecosystem studies.
We begin with an explanation of isotope terminology and fractionation, then
summarize isotopic distributions in the C, N, and S biogeochemical cycles,
and conclude with five case studies that show how stable isotope measure-
ments can provide crucial information for ecosystem analysis. We restrict this
review to studies of natural variations in C, N, and S isotopic abundances,
excluding from consideration '5N enrichment studies and hydrogen and oxy-
gen isotope studies. Our focus on C, N, and S derives in part from our

293
0066-41 62/87/1 120-0293$02.00

This content downloaded from

140.122.143.127 on Sat, 14 Jan 2023 08:49:08 UTC
All use subject to https://about.jstor.org/terms
294 PETERSON & FRY

experience that combinations of these measurements are often complementary

and in part from the need to limit the scope of the review.

TERMINOLOGY AND MEASUREMENT

Most ecological studies express isotopic compositions in terms of 3 values,

which are parts per thousand differences from a standard:

8X -: [(Rsample /Rstandard) 1] I03, 1.

where X is 13C, 15N, or 34S, and R is the corre

or 34S/32S. The 8 values are measures of the amounts of heavy and light
isotopes in a sample. Increases in these values denote increases in the amount
of the heavy isotope components (Figure 1). Conversely, decreases in 6
values denote decreases in the heavy isotope content, and a reciprocal in-
crease in the light isotope component. Standard reference materials are carbon
in the PeeDee limestone, nitrogen gas in the atmosphere, and sulfur from the
Canlyon Diablo meteorite. The precision of the measurements is typically
-+0.2?/,, Oor better, and current analysis costs typically range from $30-100
per sample through commercial firms.
Many reactions alter the ratio of heavy to light isotopes, or "fractionate"
stable isotopes, but the degree of fractionation is typically quite small. Even
very large changes of 100?'/oo (10%) between reactants and products involve
only minute absolute changes of 0.04%, 0.11%, and 0.44% for the heavy
isotopes of nitrogen, carbon, and sulfur, respectively (Figure 1). A mass
spectrometer is required for accurate detection of these small differences and
gaseous samples are required for the isotopic determinations.
Sample preparation differs from many other ecological measurements in
that quantitatiVe yields (a complete conversion of sample to gas) are required.
When yields are not quantitative, isotopic fractionation between product and
residual materials may result in a false apparent isotopic composition of the
samples. Many combustion schemes have been developed to quantitatively
break down diverse molecules into the simple gases most suitable for mass
spectrometry. Most laboratories currently employ a high temperature sealed
tube combustion to convert carbon and nitrogen compounds to CO2 and N2
(70). N2 can also be prepared from Kjeldahl digestions or ammonia (70).
Sulfur-containing materials are typically converted to sulfates and sulfides,
which are in turn converted quantitatively to SO2 (30, 41, 122). Pure C02,
N2, and S02 are- separated from one another and from water using various
cold traps that allow differential volatilization and trapping under high
vacuum conditions. A pure gas is then introduced into a dual or triple
collector isotope ratio mass spectrometer, and its isotopic composition mea-
sured relative to a known standard.

This content downloaded from

140.122.143.127 on Sat, 14 Jan 2023 08:49:08 UTC
All use subject to https://about.jstor.org/terms
 STABLE ISOTOPES IN ECOSYSTEMS 295

0.40 - 99.60

T 0.38 -
O-.

? 0.37 4

0.36 0.1 %.: 0.00004 %

0.35 , , i , , , , , 99.65
-20 0 20 40 60 80
f5N

t .2 , , , , , , , , , 98.88

1.10

1.08

<' 1.06 /

1.02 / 0.1%O 0.00011Y/O

te4.20 Be
1.00 99.00
-100 -80 -60 -40 -20 0
a13c

4.50 C 94.69

4.40

+ 4.30 -4

4.10 0. I O -0.00044?/%
4.00 I I I I I 195.24
-60 -40 -20 0 20 40
5 34s

Figure 1 Relationship of N, C, and S stable isotope content to 8'5N, 8'3C and 83S values.
Large variations of 100?lo / correspond to only slight variations in percent 13C, 15N or 34S. The
linear relations shown were derived from simultaneous solution of Equation I (text), and the
following isotope ratios for standards: 13C/12CpDB = 0.0112372, 15N/'4NAIR = 0.0036765 and
34S/32SCDT = 0.045004 (15, 51, 77). By definition (Equation 1) standards have 0?'IO O values.

Isotopic Fractionation
Stable isotopes record two kinds of information. Where physical and
chemical reactions fractionate stable isotopes, the resulting isotopic distri-
butions reflect reaction conditions (process information). Stable isotope dis-
tributions also record information about the origins of samples (source

This content downloaded from

140.122.143.127 on Sat, 14 Jan 2023 08:49:08 UTC
All use subject to https://about.jstor.org/terms
296 PETERSON & FRY

information). The source sets an isotopic baseline that can subsequently

be shifted by isotopic fractionation. A well-studied example is carbon isotope
fractionation in photosynthesis. A 1974 study showed that terrestrial C3
plants average -27.8?'/co (115). This was about 20'/co more negative
than the source of carbon for plants, CO2 in air (8'3C,02 = -7.4?'/c in
1974; 55). The overall plant isotopic composition thus reflected both source
(-7 .4'/, O) and fractionation (-20.4?'/o O) information: -27. 8PI-ANT =
7 4SOURCE 20.4FRACTIONATION
Isotopic fractionation is a mass-dependent phenomenon. The addition of
neutral mass (neutrons) does not alter most aspects of chemical reactivity so
that different isotopes of the same element are functionally equivalent in most
chemical and biochemical reactions. The slight differences that do occur are
measured as a partial isotopic separation or fractionation. In this introduction,
we focus on two kinds of isotope effects, equilibrium and kinetic, and on how
isotopic fractionation is expressed in ecosystems.
Where chemical exchange occurs between two molecules, small equilibri-
um isotope effects are common. One example is the exchange of carbon
between CO2 in air and bicarbonate in the ocean (74). Bicarbonate equili-
brated with CO2 in air is enriched in 13C by 10.8 to 7.4?'/c, relative to the
CO2 at 0-30TC. Under conditions of free exchange, the general rule is that
heavy isotopes concentrate in the molecule where bond strengths are greatest,
in this case, in bicarbonate (HCOO) rather than CO2. Equilibrium isotopic
fractionations vary in magnitude with temperature and can be predicted from
bond strength measurements for reactions involving simple molecules (23,
116).
Most biological reactions are more complex than simple equilibrations, and
such reactions involve kinetic isotope effects. Kinetic isotope effects are most
simply illustrated in physical processes such as diffusion. For example,
diffusion of 34SO2 (mass 66) is about 1.6% or 16?'/o slower in a vacuum
than diffusion of 32S02 (mass 64). This is because application of the same
force displaces the mass 66 sulfur dioxide more slowly than its light-isotope
mass 64 counterpart. The ?/, isotope effect or discrimination, D, ranges
from 0 to 100?'/o for most kinetic reactions involving C, N, and S stable
isotopes. The discrimination (D) values are positive in sign when light isotope
species react faster than their heavy isotope counterparts and can be closely
approximated as the ?/, difference between a substrate and product formed
at an instant in time from that substrate (83):

D 5SUBSTRATE - PRODUCT
The maximum isotopic discrimination represented by a D value is an
important unit used in modeling isotopic distributions. This maximum dis-
crimination is often not realized, and the degree of observed fractionation
depends strongly upon reaction conditions. The simplest case in which full

This content downloaded from

140.122.143.127 on Sat, 14 Jan 2023 08:49:08 UTC
All use subject to https://about.jstor.org/terms
 STABLE ISOTOPES IN ECOSYSTEMS 297

discrimination occurs is a one-step reaction starting with an essentially

ited supply of substrate. Under these open system conditions, the subst
and product isotopic compositions are related simply by 8PRODUCT=
8SUBSTRATE - D.
A somewhat more complex case occurs in closed systems where a limited
amount of substrate is initially available, and this substrate is completely
converted to product over time. Closed and partially closed systems include
laboratory bottle incubations and some sediments and soils. In closed sys-
tems, the substrate gradually increases in heavy isotope content (Figure 2,

Substrate

o rduiIF

o 0 -

40- III

- - -- CMLTV

0% 100%
SUBSTATE COSUPLTION

Figre Tp: odl o clse s60e istoi canewhrsutaeiscvredoto

Fulft re 2n Top:rMoel of clos',,, edpsytem in sotopic chagelwereo substate isncovrteds wto detwo

(sulfate depletion) but not as fast as expected for a 57.5'/,,,, fractionation (dashed line), possibly
because of increased diffusion 32S042- vs. 34S042 from overlying seawater (50) or sulfide
oxidation (2).

This content downloaded from

140.122.143.127 on Sat, 14 Jan 2023 08:49:08 UTC
All use subject to https://about.jstor.org/terms
298 PETERSON & FRY

top) due to isotopic fractionation during the formation of heavy-isotope

depleted (isotopically light) products. This change in substrate isotopic com-
position contrasts with open systems where substrate is continually renewed
and hence does not change in isotopic composition. The cumulative product
in closed systems follows an isotopic trajectory that preserves mass balance as
substrate is consumed (Figure 2). As in open systems, however, the product
forming at any instant in time differs from the substrate isotopic composition
by D (Figure 2).
The decomposition process of sulfate reduction, for example, occurs in
a semi-closed system in marine sediments and results in marked sulfur iso-
topic changes. As sulfate is consumed in sediments, the 34S content of the
residual sulfate increases from a surface 834S value near +21?/oto values
> + 60?/o at depth (Figure 2, bottom). A large isotope effect during sul-
fate reduction (typical D = 30 to 70'/, ,; 38) is responsible for this increase,
as production of 34S-depleted sulfides leaves residual sulfate increasingly
enriched in 34S. Marine sediments are not truly closed systems because of
continual, albeit slow substrate inputs (50), but models of closed system
isotopic behavior are nonetheless useful for obtaining a qualitative under-
standing of isotopic changes accompanying many diagenetic reactions (67).
This discussion of fractionation has thus far been limited to consideration of
single reactions. Multiple coupled reactions, perhaps the common reality in
ecosystems, have also been extensively modeled. As examples, we consider
three cases: a tightly linked reaction sequence with several enzymatic steps, a
two-step reaction in which the first step is reversible, and a branched reaction
chain with two products (Figure 3).
In many biological reactions, chemical intermediates are nearly quantita-
tively converted to a final product in a series of linked, essentially unidirec-
tional enzymatic steps. If fractionation is observed in the final product, this
fractionation must have occurred in the first step of the reaction. It is the first
step that determines the isotopic input into the reaction chain; by mass
balance, the final product, or output, must match in isotopic composition with
this input. This finding is a more general formulation of the rule that when
reactions go to completion (all substrate is converted to product) there is no
opportunity for isotopic fractionation to occur.
A second example concerns a reaction sequence with an early reversible
step (Figure 3, middle). This occurs in photosynthetic carbon fixation by C3
terrestrial plants (83). The diffusion of CO2 from the atmosphere into internal
leaf air spaces is reversible, since CO2 can diffuse back out to the atmosphere.
The observed isotopic fractionation depends in part on this back diffusion,
and in part on the carboxylation of CO2 into simple plant sugars. Where back
diffusion is negligible, the overall fixation reaction is unidirectional into the
leaf and all CO2 entering the leaf is carboxylated. The only isotope effect
expressed in this circumstance is that associated with diffusion (D 4.4?/, 0,;

This content downloaded from

140.122.143.127 on Sat, 14 Jan 2023 08:49:08 UTC
All use subject to https://about.jstor.org/terms
 STABLE ISOTOPES IN ECOSYSTEMS 299

D
Figure 3 The outlined reaction sequences contain starting substrate (A), intermediate (B), and
product(s) (C & D). Each reaction step could result in isotopic fractionation if measured in
isolation, but linking reactions modifies expression of the individual fractionations. In a sequence
of tightly linked reactions (top), overall isotopic fractionation is associated with the first step (A
-* B). Where a reversible equilibration occurs before a unidirectional reaction (middle), isotopic
fractionation depends on the fraction of B converted to C versus the back reaction of B to A. In
branched reactions (bottom), the proportion of mass flow diverted from product C to product D
strongly influences isotopic compositions of both of these products.

25), the first step in the overall fixation process. The other extreme occurs
when carboxylation is much slower than diffusion, and the internal leaf CO2
concentration approaches that of the atmosphere. Diffusion no longer controls
reaction rates, and a larger carboxylation discrimination of about 29?/ ,O (97)
will be expressed. Actual fixation occurs in an intermediate situation where
neither diffusion nor carboxylation completely limits the overall fixation rate,
and the observed fractionation is intermediate between 4.4 and 29?/o, at
about 21 ?/o (25). The reversibility of the diffusional step accounts for this
intermediate value; if the flow of CO2 were unidirectional into the leaf, only
the 4.4?/, diffusional fractionation would be expressed. Reversibility of
diffusional fluxes may be important for isotopic fractionation in other uptake
reactions involving HCO 3-, S042-, NO3-, NH4' etc. To the extent that
nutrient uptake steps are not simply unidirectional, the magnitude of isotopic
fractionation will reflect the balance in the rates of nutrient fixation and
nutrient efflux from the sites of uptake. Very low substrate concentrations or
slow diffusion steps often result in small overall fractionations since isotope
effects for diffusion of inorganic nutrients in air and water are typically small,
<5"/co. (83, 84, 95, 120).

This content downloaded from

140.122.143.127 on Sat, 14 Jan 2023 08:49:08 UTC
All use subject to https://about.jstor.org/terms
300 PETERSON & FRY

Our photosynthesis example is actually a special case of branched reac-

tions, with back-branching occurring during CO2 efflux from the leaf. For
branched reactions with one or more side products (Figure 3, bottom), the
product isotopic compositions depend not only upon the isotopic discrimina-
tion factors for individual steps, but also upon how the mass flow is parti-
tioned between these steps (44). The isotopic compositions of many metabo-
lites such as amino acids, proteins, and lipids are controlled by branched
reactions (1, 42, 62). Investigations of branched reaction networks show that
isotopic compositions of especially small pools with rapid turnover can vary
widely depending on the reactions leading away from these pools. Isotopic
compositions record source information most faithfully when there is no
fractionation in reactions leading away from or to the pool in question.
In summary, isotopic fractionation in most biochemical reactions arises
when similar molecules of slightly different mass react at different rates.
Chemical and biochemical studies typically divide reactions into a series of
individual steps that, when combined in kinetic models, yield the appropriate
fractionation for the overall reaction. These models can be expanded and
applied at an ecosystems level to understand fluxes of materials but have not
yet been intensively used by ecologists to do so.

OVERVIEW OF C, N, AND S DISTRIBUTIONS

IN NATURE

The natural abundance isotopic compositions of C, N, and S in ecosystems

are illustrated in Figures 4-6. Our objective with these figures is to give a
comparative overview of C, N, and S stable isotopic distributions at the
ecosystem level, rather than to give detailed values for all parts of each cycle.
Our values are representative but do not encompass the full spectrum of
observed values. Because space prevents us from a detailed treatment, we
suggest that interested readers consult the references cited in the figure
legends for further information.

The Carbon Cycle

The carbon cycle involves active exchanges of CO2 between the atmosphere
and both terrestrial ecosystems and the surface ocean (Figure 4). The 813C
value of atmospheric CO2 is decreasing in response to inputs of 13C depleted
CO2 from fossil fuel plus biomass burning and decomposition (88). Over the
past 30 years the decrease may have been almost 1?/oo. Carbon uptake by
the dominant C3 plants on land involves a net fractionation of about 21?/o o
between the atmosphere (-7?/O O) and plant biomass (-28`/, o) (17).
Carbon uptake by C4 plants, mainly tropical and salt grasses, involves a small
fractionation of about 6?/o O. Soil organic matter globally contains several-

This content downloaded from

140.122.143.127 on Sat, 14 Jan 2023 08:49:08 UTC
All use subject to https://about.jstor.org/terms
 STABLE ISOTOPES IN ECOSYSTEMS 301

ATMOSPHERE

CO2 -7

0 21 -136 0

C02 -27 l1111111111111111llil

Soil Organic Matter -26 IC02 -15 Tropical C-4 Total Dissolved C02 0
POM -35 Grasslands POM -22
DOM -23

LAKE OCEAN

Figure 4 The '-3C distribution in ecosystems. Single arrows indicate CO, fluxes. The double
arrow signifies an equilibrium isotope fractionation. Numbers for pools indicate 6'3C values
(?/0 O) and numbers for arrows indicate the fractionation (D) occurring during transfers. POM =
Particulate Organic Matter. DOM = Dissolved Organic Matter. Sources: 5, 16, 17, 35, 83, 104,
121.

fold more carbon than either the atmosphere or living plant biomass and in
general is similar or slightly enriched in 13C in comparison with the dominant
vegetation. While either differential preservation or mineralization of soil
components with different 513C values does lead to gradual shifts in soil 13C
content (6, 22), on average there is little fractionation of respired CO2 (83).
The exchange of CO2 between the atmosphere and the surface of the ocean
involves an equilibrium chemical fractionation between atmospheric CO2
(-7c/c, ) and the total CO2(ICO2, mostly bicarbonate) in surface ocean
water (0/c, ,) (74). The withdrawal of carbon to form carbonates involves
little isotopic fractionation whereas uptake of dissolved inorganic carbon in
planktonic photosynthesis involves larger kinetic fractionation that results in
algal values of about - 19 to -24?'/o o. Both the dissolved and the particulate
organic matter in the oceans reflect predominantly a marine planktonic origin
(121).
The 13C contents of components of the carbon cycle of fresh waters vary
widely depending on the source of dissolved CO2 in the waters-from
carbonate rock weathering, from mineral springs, from the atmosphere or
from respired organic matter. Where respiration inputs are strong, V13C values
for dissolved inorganic carbon may approach - 20?/o, and algae that further
fractionate during carbon uptake can measure -45?/, (93, 94, 99).

This content downloaded from

140.122.143.127 on Sat, 14 Jan 2023 01 Jan 1976 12:34:56 UTC
All use subject to https://about.jstor.org/terms
302 PETERSON & FRY

ATMOSPHERE 1
N2 gas 0

eae PRECIPITATION I eq.

:( i -8 to+34 NH' -P18 to 08 (1ypi

Fossil Fue -2ag> 0 1 to t3 2

Emissions - o+

_NH3 -8 to -4 _ < Vv NO- -15 to O -3

Soil Organic Matter -4 to+14 N POM | Fertilizer | Dissolved N

/ NO- -+3 |
LAKE \ Deep Water NO +4 to +6
OCEAN

Figure 5 The 8'5N distributions in ecosystems. See Figure 4 legend for explanation of symbols.
Sources: 26, 27, 14, 48, 52, 60, 63, 71, 75, 76, 106, 108, 109.

The Nitrogen Cycle

Most nitrogen in the biosphere is present as N2 gas in the atmosphere. This
massive reservoir is well-mixed with an isotopic composition that is es-
sentially constant at 00/0.0 (63). Nitrogen in most other parts of the biosphere
also has an isotopic composition near this 00/00 value, from -10 to
+100/00 (Figure 5), primarily because the rate of nitrogen supply often
limits reactions such as plant growth and bacterial mineralization. Under these
conditions all available nitrogen can be consumed without isotopic fractiona-
tion. The rates of supply and amounts of substrate N are thus important in
understanding nitrogen isotopic distributions.
Some cumulative and large fractionations do occur in the nitrogen cycle. A
cumulative faster loss of 14N than 15N during particulate N decomposition in
soils (66) and in the sea (100) results in 15N increases of 5-10?/oo with
increasing depth. Nitrification and denitrification in the sea both proceed with
substantial isotope effects (D = 10-40?/o o; 14, 65, 85), and where nitrate is
abundant, assimilation by phytoplankton proceeds with a smaller effect (D =
4 to 6?/o o; 3, 119). With two exceptions (24, 86), nitrogen isotopic values in
lakes have received little study, although large contrasts might be expected
between lakes in which primary production is limited by N (no fractionation
by phytoplankton) vs P (abundant N -> large fractionations during N uptake
by phytoplankton). Where phytoplankton have different 615N values than
terrestrial vegetation, the nitrogen isotopes may function as source markers

This content downloaded from

140.122.143.127 on Sat, 14 Jan 2023 08:49:08 UTC
All use subject to https://about.jstor.org/terms
 STABLE ISOTOPES IN ECOSYSTEMS 303

ATMOSPHERE

Stratospheric S04 +2.6 Background

Leve CONTINENTAL MARINE

s a-5 to +2 - PRECIPITATION PRECIPITATION

l <>g}sgvsy_g SO4 I to t7 S04'- + +13

Fossil Fuel _-10 to +10

-S?2 -1 to +25 ._ .r Illlllllllll

SoilSoilOrgaic Mtter-
Organic Matter itar \-22to+20
- toPlnsS2 2 / Salt Marsh | POM + 17 to + 21
vegetation Om Plants ?SO +21

LAKE OCEAN

Figure 6 The 8-"S distribution

Sources: 8, 9, 13, 46, 52, 53, 54, 58, 77, 79, 91, 101.

for autochthonous and allochthonous organic matter. This approach has been
successfully applied in marine environments (90).
There is a wide range reported for nitrogen isotopic values for ammonium
and nitrate in precipitation from about -20 to + 1O/o o (Figure 5) (48, 75,
76). Some of the more negative values are related to soil and anthropogenic
emissions in highly industrialized areas (26, 27). Further study may show that
stable isotope studies are helpful in identifying the sources and fates of N that
human activities are currently adding to many forests and lakes.

The Sulfur Cycle

Sulfate in the ocean is a large well-mixed sulfur reservoir whose isotopic
composition is 21?'/o, heavier than primordial sulfur in the meteorite stan-
dard, Canyon Diablo Troilite (96). Fixation of sulfate by phytoplankton
occurs with a small isotope effect (D = 1-2?/6 -; 53), but dissimilatory
sulfate reduction in marine sediment occurs with a large effect of 30-70o/o .
(10, 38). Over geological time, and partially in response to global-scale
fluctuations in sulfate reduction activities, the 634S values of oceanic sulfate
have varied from about + 10 to +33% (77). Uplift and preservation of marine
sulfides and sulfate-containing evaporites on land have produced a patchwork
of terrestrial environments, each with different &34S values for bedrock sulfur
(e.g. 46). Thus, large V34S ranges must be assigned in general sulfur cycle
diagrams (Figure 6). In spite of this, continental vegetation seems to average

This content downloaded from

140.122.143.127 on Sat, 14 Jan 2023 08:49:08 UTC
All use subject to https://about.jstor.org/terms
304 PETERSON & FRY

near +2 to +6?/oon over large areas (13) and is quite distinct from the -+ 17
to +21?'/oo values of marine plankton and seaweeds.
The stable isotopic composition of sulfur entering the atmosphere can also
be quite variable. For instance, SO2 emissions from a sour gas plant in
Alberta, Canada vary between +8 and +25?/,, (58), but in eastern Canada
and the northeastern United States initial studies show average ambient 834S
values of 0 to 2c/,,o, for S02 (79, 101). The oxidation of SO2 to sulfate
occurs with an overall inverse effect (D = -4?/, ) that favors concentration
of the heavy 34S isotope in the product sulfate; this inverse effect arises from
an equilibrium step between S02 and HSO3- prior to final oxidation to sulfate
(101). Oxidation of other sulfur-containing molecules also occurs with small
isotope effects, D < 5?/c, (37).
Rainfall sulfate over the open oceans has a significantly lower 834S value
than sea-spray sulfate (-+13 vs +21/o; 13), possibly because of slow
oxidation of reduced sulfur gases (45). The isotopic compositions of these
gases-H2S, carbonyl sulfide and dimethyl sulfide-are poorly known, but
further study may clarify the relative contributions of human vs natural
sources of these atmospheric sulfur compounds.

Stable Isotopes in Animals

Animals are not represented in our C, N, and S diagrams (Figures 4-6), but
they are important components of ecosystems. Animals are similar in isotopic
compositions to their diets for carbon and sulfur, but average 3 to 5?/OO
heavier than dietary nitrogen (Figure 7). Diet-switching and turnover ex-
periments for carbon clearly show that the diet is the primary determinant of

30

20 -

to-
10
o1
CARBON

-3 -2 -f O0 1 2 3 4 5 6 7 8 9 iO

N 10[ SULFUR

-3 -2 - 0 1 2 3 4 5 6 7 8 9 t0

to [ NITROGEN

-3 -2 -1 0 1 2 3 4 5 6 7 8 9 to
A ( aANIMAL - bDIET )

Figure 7 Relationships between animal and diet isotopic compositions for carbon, sulfur and
nitrogen stable isotopes. Animals are on average only slightly enriched in '3C and 134S by
0.2?/,,,, (have positive a values) vs their diets (arrows denote averages). A larger 3.2?/oo
average enrichment occurs for nitrogen isotopes due to excretion of '5N depleted nitrogen.
Sources: 31, 35, 61, 68, 69, 92.

This content downloaded from

140.122.143.127 on Sat, 14 Jan 2023 08:49:08 UTC
All use subject to https://about.jstor.org/terms
 STABLE ISOTOPES IN ECOSYSTEMS 305

animal isotopic compositions (34, 59, 114). For example, gerbils that were
switched from a - 12.2?/oo corn diet to a -21.8?/o,o wheat-based diet
approached the wheat carbon isotopic composition with time (114). This
should be true for sulfur as well, and also for nitrogen, with the caveat that
animals should be somewhat enriched in '5N vs the new diet.
The '5N enrichments vs diet are mainly due to excretion of isotopically
light nitrogen in urine. Initial studies with cows, fish, and zooplankton show
that animals and feces are enriched in '5N vs the diet, but urinary nitrogen
(both NH3 and urea) is depleted in 15N (1 1, 12, 69, 73, 109). For example,
cow urine is -1 to -4?/O depleted in 15N vs diet; cow feces are -+2/cco
enriched in '5N, and milk and blood are +4?/, , enriched in 15N (109). As
they must be to preserve mass balance, the urinary losses of '4N are thus
offset by '5N enrichments in other nitrogen pools.
When whole food webs are examined, differences between animal process-
ing of C, N, and S isotopes stand out even more clearly. Nitrogen isotopic
values increase by 10 to 15?/oo in many food webs; these increases are due
to the presence of 3 to 5 successive trophic transfers, each of which effective-
ly boosts the '5N content by 3 to 5O/,, (69). The opposite effect-no change
with increasing trophic level-is observed for sulfur (68, 92), making the
sulfur measurements particularly good indicators of which plant or bacterial
food sources are most important for consumers. Carbon appears to be in-
termediate between N and S, showing modest increases between 0.0 and
1.0?/oo per trophic level (19, 35, 36). This small enrichment may be due to
carbon isotopic fractionation during assimilation or respiration. The distinct
behavior of each isotope allows a cross-referenced trophic structure analysis
through combined C, N, and S measurements (32, 91).
Although diets control the overall isotopic compositions of animals, con-
siderable isotopic variation still exists among different tissues and metabolites
within individual animals. The bone protein collagen, for instance, is 2 to
6?/,,, enriched in 13C vs the diet, while lipids in fat reserves are 2 to 8?/o
depleted in 13C (87, 117). These enrichments and depletions arise during the
many internal enzymatic steps that fractionate stable isotopes after dietary
uptake in animals (1, 102, 123). Use of stable isotopes to study diets is based
on use of animal tissues that bear a fixed isotopic enrichment or depletion vs
the diet. In some cases, whole animals are used, while in others, analyses of
muscle or protein fractions have proven adequate indicators of diet. These
analyses complement other methods of studying diets in that the stable
isotopic compositions of tissues are a measure of the assimilated (not just
ingested) diet, reflect both long-term and short-term diets in slow and fast
turnover tissues, and provide a unique way to study food webs of the past,
detrital food webs, and diets of animals that are hard to observe. Two of our
case studies illustrate how stable isotopes can be used to reconstruct diets
when other direct dietary information is lacking.

This content downloaded from

140.122.143.127 on Sat, 14 Jan 2023 08:49:08 UTC
All use subject to https://about.jstor.org/terms
306 PETERSON & FRY

CASE STUDIES

Rather than attempting an exhaustive review of a rapidly growing field, we

have chosen five case studies to illustrate how ecosystems scientists are
currently using stable isotopes as natural tracers. These studies show that
stable isotopes are being used to indicate sources (N2 fixation, archaeology,
detrital food webs), to help estimate large-scale fluxes (global carbon), and to
discern how sulfate loading affects an important decomposition process in
lake sediments (acid deposition in lakes). This wide variety of applications is
remarkable but represents only a small part of the spectrum of current uses of
stable isotopes.

'55N Measures of N2 Fixation

Nitrogen stable isotope measurements can be used to estimate plant fixation of
atmospheric N2. Plants fixing N2 from the atmosphere have 6'5N values of
about -2 to +2?/o,o close to the 0?/Oo value of atmospheric N2 (107).
Plants that cannot fix N2 from the atmosphere have 6'5N values that are
usually closer to those of total nitrogen in soils and vary widely from -8 to
+ 10?/oo,, depending on location (66, 107; K. Nadelhoffer & B. Fry., un-
published). For plants whose nitrogen source is unknown, values close to
0?/,, .may indicate strong reliance on N2 derived from the atmosphere
(Ndfa), especially if soil organic nitrogen has a 8'5N value very different from
An agricultural experiment with soybeans illustrates this approach
(Figure 8). Nodulating soybeans grown hydroponically in greenhouse trials
had values of + 1?/.. for total plant N, but in field planting experiments,
nodulating soybeans had values close to those of reference plants that did not
fix N2 (in this case, the reference plants were soybeans that lacked N2-fixing

SOYBEAN I'5N

1 2 3 4 5 6 7 8

N2 FIXING PLANTS

Hydroponic +

Corn Cobs Added

No Corn Cobs

REFERENCE PLANTS

Corn Cobs Addedd

No Corn Cobs

100% 0%
N2 FIXATION

Figure 8 The nitrogen stable isotopic composition of nitrogen in soybeans depends on cultiva-
tion conditions. Soybean 8'5N varies according to the fraction derived from the atmosphere vs
soils (reference plants). Sources: 56, 57.

This content downloaded from

140.122.143.127 on Sat, 14 Jan 2023 08:49:08 UTCC
All use subject to https://about.jstor.org/terms
 STABLE ISOTOPES IN ECOSYSTEMS 307

nodules). Presumably, the nodulating soybeans derived most of the nitrogen

from the soil, as did reference plants. Addition of corn cobs changed these
results. Corn cobs provide a rich carbon source for soil microorganisms that
deplete available nitrogen in the soil. Enhancement of microbial growth in
soils sets up a competition for soil nitrogen stocks so that plants capable of
fixing atmospheric N2 may increasingly do so. The lower 815N values of
soybeans grown in corn-cob amended soils are consistent with a greater
dependence on atmospheric nitrogen (Figure 8).
This natural abundance method of measuring the fixation of atmospheric
nitrogen has been extensively tested against other methods and appears
accurate at the 5 to 15% level in many situations (107). Simply collecting
leaves of presumed N2-fixers and reference plants to estimate in situ patterns
of N fixation is a nonmanipulative procedure that has great simplicity and
intrinsic appeal for ecosystems studies. Interesting results have been found in
comparative field studies. For instance, recent findings show that presumed
N2 fixers can range widely in their reliance on Ndfa in different habitats
(108). The extent to which Ndfa enters and contributes to nutrition of neigh-
boring reference plants, and also acts as a source of new nitrogen in ecosys-
tems, has not been fully explored in spite of some initial attempts (7, 64).

The Global Carbon Cycle and the C02 Problem

Carbon isotope research is making substantial contributions to understanding

the factors controlling the CO2 balance in the atmosphere. While it is well
known that the CO2 content of the atmosphere has been increasing steadily in
recent decades, the relative importance of CO2 from combustion of fossil
fuels and from oxidation of forest and soil organic matter in causing this rise
has been debated (47). While the fossil fuel CO2 emission history is known
fairly well, the timing and amount of carbon dioxide releases from burning
and decay of terrestrial organic matter is not. However, the addition of
terrestrial CO2 to the atmosphere can be calculated through use of carbon
cycle models that incorporate measurements of the time-course and magni-
tude of the changes in 13C content of atmospheric CO2 from 1800 to the
present.
The approach uses the fact that the addition of large quantities of CO2 from
both fossil fuel combustion (813C -27?/o o) (110) and oxidation of terres-
trial organic matter (81 3C -28?/o ) has lowered the 813C values of
atmospheric C02 (813C -7?/oo). Since the time course and amount of
fossil fuel use are well known (98), it is possible to calculate by difference
both the time course and the magnitude of release of terrestrial biomass
carbon to the atmosphere. The calculations require both an accurate time
course of atmospheric 813C values and the use of models of the carbon cycle

This content downloaded from

140.122.143.127 on Sat, 14 Jan 2023 08:49:08 UTC
All use subject to https://about.jstor.org/terms
308 PETERSON & FRY

since atmospheric CO2 exchanges continuously with oceanic CO2 and

biospheric carbon.
Carbon isotopes in tree rings have been analyzed as a surrogate to measure
trends in 813C values of atmospheric CO2. While different trees show a wide
variability in their isotopic records due to climatic and physiological factors,
the grand average of many tree-ring records indicates a 1.5?/oo decrease in
tree ring 813C values from 1750 to 1980 (Figure 9) (88). The inference is that
atmospheric 813C values have also gradually declined from 1800 to 1950 and
declined more rapidly since 1950. Recent analyses of atmospheric CO2
trapped in Antarctic ice cores have supported the approximate timing and
magnitude of the atmospheric 813C changes deduced from tree ring analysis
(29). The interesting point is that the initial decline in atmospheric 813C
values was occurring during the 1800s well before the major upswing in fossil
fuel use. This has been attributed to a long-term and large-scale release of
13C-depleted CO2 caused by the clearing and burning of forests and oxidation
of soil organic matter (813C -28?/, c). The more rapid recent 813C decrease
correlates with the increase in fossil fuel use.
Through use of carbon cycle models, the time course and amount of
biospheric carbon release required to account for the observed tree ring 813C
decrease have been computed (88, 89, 110). The calculations are sensitive to
small changes in the 13C trends, but the general conclusions are (a) that

0.0 - -.-- Model Fossil -

-0.4 -Fuel_
\Only
-0
-0.8- Tree R.ng- -
-1.26
-1.0
1780 1820 1860 1900 1940 1980

5.0 .

4.0 Fossil Fuel

_ 3.0 -

2.0 - Biospheric Release

1.0-

0.0
1820 1860 1900 1940 1980
YEAR

Figure 9 Top. The changes in 613C values (A from historical tree ring baseline) observed in tree
rings from 1870 to 1980 (solid line) and the calculated change expected due solely to increased
fossil fuel use (dashed line). Note the decline in 13C (declining A value) in tree rings prior to the
decline expected from fossil fuel use. The difference between the two lines may be due to carbon
release from terrestrial biomass and soils accompanying clearing of forests and prairie (88).
Bottom. The time course of fossil fuel CO2 emission (98) and, using 613C data, the calculated
biospheric CO2 emissions from terrestrial soils and forests (88).

This content downloaded from

140.122.143.127 on Sat, 14 Jan 2023 08:49:08 UTC
All use subject to https://about.jstor.org/terms
 STABLE ISOTOPES IN ECOSYSTEMS 309

carbon input from the terrestrial carbon stocks to the atmosphere was greatest
between 1840 and 1940 but (b) that after 1958 fossil fuel became the
predominant cause of the atmospheric CO2 increase (Figure 9; 88). The total
integrated C02 input from each of the two sources until 1980 was about equal,
roughly 140 x 109 tons from the biosphere and 170 x 109 tons from fossil
fuels, according to recent model calculations (88). This model allows back
calculation of the preindustrial atmospheric CO2 level that provides the best
fit to both the isotopic trends and the recently measured atmospheric CO2
values. The best estimate of 266 ppm is consistent with ice core values of 258
to 289 ppm (28).
This example illustrates the usefulness of the isotopic tracer approach for
solving mass balance problems, especially where a well-mixed reservoir such
as the global atmosphere can be sampled over time. Alternative methods of
calculating the loss of biospheric carbon through analysis of land-use patterns
and deforestation are uncertain due to the lack of accurate historical land-use
records (49). The existence of two independent methods of computing
changes in biospheric carbon stocks (6'3C vs historical land use records) is
important considering the urgency of gaining accurate understanding of the
controls of atmospheric CO? levels.

Sulfur and Acid Deposition

Human perturbation of the natural sulfur cycle is pronounced, as the majority

of sulfur currently emitted into the atmosphere is probably anthropogenic in
origin (4; R. Howarth, personal communication). Anthropogenic sulfur emis-
sions (SO2, SO3, SO42-) have increased dramatically with industrialization,
and these changes have been recorded by changes in the sulfur content of lake
sediments. The studies of Canadian lakes by Jerome Nriagu and colleagues
illustrate these changes (78, 79, 80, 81).
A very large smelter of copper-nickel sulfide ores located in Sudbury,
Ontario, has accounted for about 20% of total Canadian S emissions. Sudbury
smelters began operation in about 1889, and sharp increases in the percentage
of sulfur in the sediments of nearby lakes have been dated to the 1880-1890
horizons (80). The isotopic composition of the sulfur in lake sediments has
also changed since the 1880s (Figure 10). The weighted average 334S value of
sulfur in ores from seven mines in the Sudbury area is + 2.1 ?/o (range =
0.5 to 7.2?/..; 105), and this is slightly depleted in 34S relative to sulfur
collected from chimney stacks, precipitation, lake water, and older pre-
industrial lake sediments that all average about 4-5?/O O in and near Sudbury
(78, 79, 80). Simple addition of ore-derived sulfur should only slightly
depress 634S values in the Sudbury area towards the + 2. 1 ?/o o average value
of ore sulfides, but in fact much larger 34S depletions of - 30'/. . have been
found in sulfur from recent lake sediments.

This content downloaded from

140.122.143.127 on Sat, 14 Jan 2023 08:49:08 UTC
All use subject to https://about.jstor.org/terms
310 PETERSON & FRY

Analysis of the chemical forms of the sulfur in recent sediments provided a

key to understanding how additions of sulfur at +2 to +50/. could lead to
-10 to -30'/o. values in surface sediments of lakes. Reduced sulfur
species (H2S, So, and pyrite) characteristically formed during dissimilatory
sulfate reduction have been found in large quantities in two of the Sudbury
lakes (81). Bacterial sulfate reduction in these lake sediments appears to be
stimulated by atmospheric sulfate additions; low sulfate levels in lake water
normally limit sulfate reduction. Increased sulfate loading stimulates sulfate
reduction, increases the production of 34S depleted sulfides, and since sulfides
react with iron and organic matter in sediments, results in decreases in the
834S value of total sulfur in lake sediments. The decrease in lake sediment
834S is thus not primarily due to a change in the isotopic composition of
source sulfur but rather to changes at the process level as sulfate reduction is
stimulated.
As in the preceding example of 813C and global CO2 changes, this sulfur
case study shows that isotopic measurements can be used to identify fluxes of
anthropogenic pollutants. However, because sulfate concentrations limit an
important ecosystem process (sulfate reduction), the sulfur isotopic response
is not governed by simple mixing as was the case for anthropogenic CO2
additions. Models that relate sulfur isotopic distributions in lake sediments to
increased sulfate loading are rudimentary at this time (33), but further de-
velopment and testing should improve our understanding of long-term
changes in both sulfate reduction activities and historical rates of an-
thropogenic sulfate loading in lakes.

%S b 34s
0.0 0.2 0.4 0.6 0.8 1.0 -10 -5 0 5 10
0 0

to-10 0

20 - 20 -

a_ 30 , Pristine
ei 30 Lake 30-

40 - 40 -
Sudbury
50 - Lake 50

60 L60'

Figure 10 Sedimentary sulfur profiles differ between a pristine lake in western Ontario and a
lake with high sulfate loading near Sudbury, Canada. Triangles denote lake water sulfate 634S
values. The increases in %S and decreases in 634S in the Sudbury lake contrast strongly with the
uniform vertical profiles from the pristine lake. Increased sulfate loading from smelter activities
at Sudbury has led to production and storage of 34S-depleted sulfides in lake sediments, causing
the recent increases in %S and decreases in 634S of sediments. Sources: 80, 81.

This content downloaded from

140.122.143.127 on Sat, 14 Jan 2023 08:49:08 UTC
All use subject to https://about.jstor.org/terms
 STABLE ISOTOPES IN ECOSYSTEMS 311

75

-14 20 _ff______-------- - 50

-16
-25~
-18
-14
-20 0

-22?-

-24-

-26-

-28
3000 2000 1000 0 1000
Z ~~~~B.C. A. D.
YEAR

Figure 11 Change in diets of North American Indians are recorded in 8'3C contents of the
collagen bone protein. Older collagen averages -21.5?/cc, about 5?/cc enriched in 13C
relative to common C3 plants that average -26.5?/c (1 17). With the introduction of maize,
13C contents rise to - 1 ?/c , indicating a 60% reliance on this C4 species (a 100%
dependence would produce a -4?/cc 6'3C value for collagen). Sources: 116, 117, 107.

Stable Isotopes in Archaeology

The study of human diets in ancient settings is a key to understanding both

cultural evolution and human modification of ancient ecosystems. For ex-
ample, a civilization that relied heavily on corn was probably organized
around farming practices and differed in social structure and land use from a
hunter-gatherer civilization. Archaeologists glean evidence about ancient
diets from charred remnants of foods and from skeletons. Stable isotopic
studies have been used to quantify prehistoric human diets and have proven
valuable supplements of the more qualitative visual estimates.
A case in point is the use of bone collagen 313C to document the transition
to maize cultivation by North American Indians (Figure 11). Collagen ex-
tracted from Indian bones averaged -21 to - 22?/o from 3000 BC to 500AD,
but underwent a rapid increase in 13C to -1 1?/o by 1300 AD. This
increase stems from the advent of widespread maize cultivation by the Upper
Mississippian Indians, and the 13C increase is such that -60% of the dietary
carbon was derived from maize. This finding is interesting because traces of
maize plants were found in earlier sites and might have been judged impor-
tant. However, the isotopic analyses of collagen from those earlier peoples do
not show values expected for a dietary dependence on maize or maize-fed
animals. This supports the idea that for these earlier peoples maize was a

This content downloaded from

140.122.143.127 on Sat, 14 Jan 2023 08:49:08 UTC
All use subject to https://about.jstor.org/terms
312 PETERSON & FRY

minor dietary item whose presence was probably due to an extensive trade
network (118).
Subsequent archaeological investigations have been extended to coastal
areas where marine and terrestrial foods differ in both C and N stable isotopic
compositions, with marine foods being relatively enriched in both 13C and
' 5N. Counterintuitive results have sometimes emerged from those in-
vestigations. For example, prehistoric humans living within 1 km or so of
coastal waters might be expected to depend heavily on marine foods, but this
has not always been the case. Analyses of collagen from skeletons of pre-
historic Danish coastal dwellers showed that mesolithic Danes (before 4000
BC) had relatively heavy C and N isotopic compositions (C-- 12?/oo '13C,
- + 14"/, 3' 5N), reflecting a consumption of predominantly marine foods.
Postmesolithic Danes had substantially lighter C and N values (~-20?/rn
613C, --+9"/ 5 8'5N: 103, 112). The isotopic shifts near 4000 BC are
consistent with shifts from an early subsistence-level marine-based ecology to
that involving increased agriculture and trade. Farming and trade items
apparently came to constitute the bulk of the diet of even the coastal Danes
after about 4000 BC (1 12).
The use of stable isotopes in anthropology has thus led to a more quantita-
tive understanding of food webs and civilizations from the last 6000 years.
Stable isotope studies may be extended to decipher the dietary basis of human
subsistence for other areas and times where archaeological evidence is much
less extensive. Caution must be exercised in selection of samples and in-
terpretation of results, however, as structural analyses show that collagen in
bones slowly degrades and undergoes replacement, sometimes resulting in
large C and N isotopic changes (20, 21, 42). Uses of stable isotopes in
archaeology have been the subject of two excellent reviews (18, 117).

Detrital Organic Matter in Salt Marshes

The need for chemical tracers of organic matter flow and trophic relationships
is greatest in ecosystems dominated by detritus because the origins of detritus
can not be determined visually. Tidal salt-marsh estuaries contain vast areas
of highly productive salt-marsh grasses, often predominantly Spartina alter-
niflora. Large amounts of detrital Spartina become available to the de-
composers and detrital food webs, and it has long been thought that export of
this detritus is largely responsible for the high secondary productivity of
salt-marsh estuaries (82, 113).
One way to test the hypothesis that Spartina detritus is the dominant source
of organic matter in estuarine waters and for consumers is through stable
isotope analysis. Spartina is enriched in 13C (813C -13?/oo) relative to
either plankton (-22?/. .) or upland C-3 plants (-28?/. .). Haines (39) was

This content downloaded from

140.122.143.127 on Sat, 14 Jan 2023 08:49:08 UTC
All use subject to https://about.jstor.org/terms
 STABLE ISOTOPES IN ECOSYSTEMS 313

the first to point out that seston in tidal creeks at Sapelo Island, Georgia was
not isotopically similar to Spartina but was in the range expected for organic
matter derived from phytoplankton. Subsequent studies of consumers in-
dicated that filter feeders such as the oyster (Crassostrea virginica) also had
53I3C values more similar to plankton than to Spartina. Other consumers such
as ribbed mussels had 3'3C values intermediate between plankton and Sparti-
na (40, 72).
The carbon isotope data provided strong evidence that Spartina was prob-
ably not the primary source of detritus or of food for filter feeders in Sapelo
Island salt-marsh creeks. However, since other sources of organic matter,
such as benthic algae (- 17 ? or upland organic matter (-28?/. )
carried by rivers, might be present, the 533C values could not identify the
origin of the detritus unambiguously. This is a common problem of isotopic
tracer studies when more than two sources are present and the sample has an
intermediate isotopic value.
One way of increasing the ability to identify organic matter sources is to
employ simultaneously additional tracers such as sulfur and nitrogen isotopes.
In marshes and estuaries, sulfur is especially appropriate because upland
plants, marine phytoplankton, and Spartina have different 834S values reflect-
ing their different sources of sulfur. Phytoplankton use seawater sulfate
(+ 21 /o ,), whereas Spartina also uses 34S-depleted sulfides which are
produced via sulfate reduction in anoxic sediments (8). Upland plants obtain
sulfates from precipitation or weathering and often have 834S values in-
termediate between Spartina and phytoplankton. A plot of 834S versus 313C
for organic matter producers in and near Sippewissett salt marsh on Cape Cod
illustrates the clear two-dimensional separation (analogous to a two-
dimensional chromatograph) that can be achieved in favorable circumstances
with this dual isotope approach (Figure 12). Analyses of pooled samples of
ribbed mussels from nine locations along an ocean-to-inner marsh transect
demonstrated that the isotopic composition of these filter-feeding mussels was
a function of location where the organisms grew. Mussels collected near
the ocean were isotopically similar to plankton, whereas mussels from the
smallest creeks and the marsh surface were isotopically similar to Spar-
tina. The food resources available to mussels at various sites in the marsh
appeared to be determined in large part by the geometry and physical char-
acteristics of the ecosystem. Transect sampling of stable isotopes in sedi-
ments, plants, and animals can be a powerful way to define gradients and
boundaries of biogeochemical processes and cycles that are not otherwise
apparent.
The multiple isotope approach might be carried further through addition of
more isotopes or other types of tracers. For example, if carbon and sulfur

This content downloaded from

140.122.143.127 on Sat, 14 Jan 2023 08:49:08 UTC
All use subject to https://about.jstor.org/terms
314 PETERSON & FRY

20 Plankton

to 7
6'

Uplandr+,'
'> Plants y '

0 OSport/no
_ u>s ''\\-t2 -30

-30 20 -10

Figure 12 The 8'3C and 834S values for upland plants, plankton and Spartina alterniflora at
Sippewissett salt marsh. The numbers 1-9 indicate the isotopic composition of filter-feeding
ribbed mussels, Geukensia demissa, collected along a transect from the inner marsh (1-4) to the
ocean (9). Source: 91.

isotopes can identify the ultimate sources of the organic matter, the 8'5N
value could give an estimate of trophic level. At our present level of un-
derstanding, there are problems in doing this. For example, there are potential
foods in estuaries such as benthic microalgae that have not been well charac-
terized isotopically, and there are sometimes changes in 8'5N values of
detritus as it ages (124). Nonetheless, the use of multiple tracers has signifi-
cantly increased our ability to trace organic matter flows in the detrital food
webs of marshes and estuaries. Determining the importance of marsh grass
production for secondary production in estuaries is critical because of the
intense development pressures on coastal ecosystems, and stable isotopic
tracer studies can help evaluate these relationships.

CONCLUSIONS

Stable isotope tracers are already present and circulating in natural systems,
and their natural distribution reflects an integrated history of physical and
metabolic processes within ecosystems. As our case studies show, the major
advantage of the stable isotope approach lies in field studies where measure-
ments of existing isotopic distributions show how components of ecosystems

This content downloaded from

140.122.143.127 on Sat, 14 Jan 2023 08:49:08 UTC
All use subject to https://about.jstor.org/terms
 STABLE ISOTOPES IN ECOSYSTEMS 315

are connected. Where clear-cutting, fertilization, burning, or other kinds of

ecosystem-level perturbations alter flows of elements, stable isotopes can be
used to follow these changes and show which processes or components are
most sensitive to perturbation.
It is often impossible to judge on theoretical grounds alone whether or not
stable isotopes will be useful in solving a particular field research problem.
However, by analyzing a few carefully selected samples, one can often
determine whether further analyses will contribute significantly to a solution.
One initial objective is to determine the signal-to-noise ratio. If isotopic
differences between pools are very large and the variation within pools is
small, isotopes may provide a very powerful tool, and a few samples may
be very effective. For example, it is easy to distinguish between C4
(8I3C - 13?/oo) and C3 (8I3C -28'0/c) plants using stable C isotopes. On
the other hand, it would be very difficult to estimate nitrogen fixation using
the 8'5N technique where soil organic matter is about +2?/.. since atmo-
spheric N2 is O'/- O. It would be far better to choose an ecosystem where the
soil organic matter had 8'5N values of +6 to + 1O/oo which contrast
strongly with the atmospheric value.
Measurements of isotopic compositions of standing stocks need to be
balanced by process studies that define fractionations under natural conditions
in ecosystems; the isotopic compositions of pools are determined not only by
inputs but also by any isotopic fractionation occurring during efflux of
materials. Precise understanding of isotopic fractionation allows judgment as
to when the stable isotopes record source information and when they record
process information. It is this combination of source and process information
encoded in the natural distributions of stable isotopes that make these
measurements valuable independent indicators of C, N, and S cycling in
ecosystems.
The specialized chemical knowledge and the sophisticated analytical equip-
ment required for isotope determinations has resulted until now in a partial but
real segregation of isotope studies from other ecological investigations. A
productive development for the future would be an increasing integration of
stable isotope information into a variety of ecosystem studies.

ACKNOWLEDGMENTS

The authors thank Luis Cifuentes, Jon Cole, David DeMarais, Marilyn Fogel,
Anne Giblin, Robert Howarth, Daniel Kohl, and Georgia Shearer for com-
ments on initial drafts of this manuscript. We also thank the many scientists
too numerous to list individually who provided reprints, theses, comments,
and ideas for our consideration. This effort was supported by the Ecosystems
Center, Marine Biological Laboratory, Woods Hole, MA.

This content downloaded from

140.122.143.127 on Sat, 14 Jan 2023 08:49:08 UTC
All use subject to https://about.jstor.org/terms
316 PETERSON & FRY

Literature Cited

1. Abelson, P. H., Hoering, T. C. 1961. 14. Cline, J. D., Kaplan, I. R. 1975. Isotop-
Carbon isotope fractionation in forma- ic fractionation of dissolved nitrate dur-
tion of amino acids by photosynthetic ing denitrification in the eastern tropical
organisms. Proc. Natl. Acad. Sci. USA north Pacific Ocean. Mar. Chem. 3:
47:623-32 271-99
2. Aller, R. C. 1986. Anoxic oxidation of 15. Craig, H. 1957. Isotopic standards for
sulfides in marine sediments. Eos carbon and oxygen and correction fac-
67:996 (Abstr.) tors for mass spectrometric analysis of
3. Altabet, M. A., McCarthy, J. J. 1985. carbon dioxide. Geochim. Cosmochim.
Temporal and spatial variations in the Acta 12:133-49
natural abundance of 15N in PON from16.a Degens, E. T. 1969. Biogeochemistry of
warm-core ring. Deep-Sea Res. 32:755- stable carbon isotopes. In Organic
72 Geochemistry, ed. G. Eglington, M. T.
4. Anderson, J. W. 1978. Sulphur in biolo- J. Murphy, pp. 304-29. Berlin:
gy. Inst. Biol. Stud. Biol. No. 101. 60 Springer-Verlag
pp. 17. Deines, P. 1980. The isotopic composi-
5. Anderson, T. F., Arthur, M. A. 1983. tion of reduced organic carbon. In Hand-
Stable isotopes of oxygen and carbon book of Environmental Isotope Geo-
and their application to sedimentologic chemistry, Vol. I., ed. P. Fritz, J. Ch.
and paleoenvironmental problems. In Fontes, pp. 326-406. Netherlands: Else-
Stable Isotopes in Sedimentary Geology, vier Sci. Publ.
ed. M. A. Arthur, T. F. Anderson, I. R. 18. DeNiro, M. J. 1987. Stable isotopy and
Kaplan, J. Veizer, L. S. Land, 1:1-151. archaeology. Am. Sci. 75:182-91
Dallas: Soc. Econ. Paleontol. Mineral. 19. DeNiro, M. J., Epstein, S. 1978. In-
6. Benner, R., Fogel, M., Hodson, R., fluence of diet on the distribution of car-
Sprague, K. 1986. Depletion of ]3C in bon isotopes in animals. Geochim. Cos-
the lignin fraction of plants: Implications mochim. Acta 42:495-506
for source identification of detritus. Eos 20. DeNiro, M. J., Epstein, S. 1981. In-
67:990 (Abstr.) fluence of diet on the distribution of ni-
7. Binkley, D., Sollins, P., McGill, W. B. trogen isotopes in animals. Geochim.
1985. Natural abundance of nitrogen-15 Cosmochim. Acta 45:341-51
as a tool for tracing alder-fixed nitrogen. 21. DeNiro, M. J., Schoeninger, M. J.,
Soil Soc. Am. J. 49:444-47 Hastorf, C. A. 1985. Effect of heating
8. Carlson, P. R. Jr., Forrest, J. 1982. Up- on the stable carbon and nitrogen isotope
take of dissolved sulfide by Spartina ratios of bone collagen. J. Archaeol.
alterniflora: Evidence from natural sul- Sci. 12:1-7
fur isotope abundance ratios. Science 22. Dzurec, R. S., Boutton, T. W., Cald-
216:633-35 well, M. M., Smith, B. N. 1985.
9. Castleman, A. W. Jr., Munkelwitz, H. Carbon isotope ratios of soil organic
R., Manowitz, B. 1974. Isotopic studies matter and their use in assessing
of the sulfur component of the community composition changes in Cur-
stratospheric aerosol layer. Tellus 26: lew Valley, Utah. Oecologia 66:17-
222-34 24
10. Chambers, L. A., Trudinger, P. A. 23. Epstein, S., Buchsbaum, R., Lowen-
1979. Microbiological fractionation of stam, H. A., Urey, H. C. 1953.
stable sulfur isotopes: A review and cri- Carbonate-water isotopic temperature
tique. Geomicrobiol. J. 1:249-93 scale. Bull. Geol. Soc. Amer. 64:1315-
11. Checkley, D. M. Jr., Entzeroth, L. C. 26
1985. Elemental and isotopic fractiona- 24. Estep, M. L. F., Vigg, S. 1985. Stable
tion of carbon and nitrogen by marine, carbon and nitrogen isotope tracers of
planktonic copepods and implications to trophic dynamics in natural populations
the marine nitrogen cycle. J. Plankton and fisheries of the Lahontan Lake Sys-
Res. 7:553-68 tem, Nevada. Can. J. Fish. Aquat. Sci.
12. Checkley, D. M. Jr., Miller, C. A. 42:1712-19
1986. Nitrogen isotope fractionation by 25. Farquhar, G. D., O'Leary, M. H.,
oceanic zooplankton. Eos 67:988 Berry, J. A. 1982. On the relationship
(Abstr.) between carbon isotope discrimination
13. Chukhrov, F. V., Ermilova, L. P., and the intercellular carbon dioxide con-
Churikov, V. S., Nosik, L. P. 1980. centration in leaves. Aust. J. Plant Phy-
The isotopic composition of plant sulfur. siol. 9:121-37
Org. Geochem. 2:69-75 26. Freyer, H. D. 1978. Preliminary 15N

This content downloaded from

140.122.143.127 on Sat, 14 Jan 2023 08:49:08 UTC
All use subject to https://about.jstor.org/terms
 STABLE ISOTOPES IN ECOSYSTEMS 317

studies on atmospheric nitrogenous trace analyzed using 13C/12C ratios. Ecology

gases. Pageoph 116:393-404 60:48-56
27. Freyer, H. D. 1978. Seasonal trends of 41. Halas, S., Shakur, A., Krouse, H. R.
NH4' and NO-" nitrogen isotope com- 1982. A modified method of SO2 extrac-
position in rain collected at Julich. Ger- tion from sulphates for isotopic analysis
many. Tellus 30:83-92 using NaPO3. Isotopenpraxis 18:433-35
28. Friedli, H., Moor, E., Oeschger, H., 42. Hare, P. E., Estep, M. L. F. 1983. Car-
Siegenthaler, U., Stauffer, B. 1984. bon and nitrogen isotopic composition of
'3C/ '2C ratios in CO, extracted from amino acids in modem and fossil col-
Antarctic ice. Geophvs. Res. Lett. I 1: lagens. Carnegie Inst. Wash. Yrbk.
1145-48 1982-1983 410-14
29. Friedli, H., Lotscher, H., Oeschger. H., 43. Hartmann, V. M., Nielsen, H. 1969.
Siegenthaler, U., Stauffer, B. 1986. Ice V4S-werte in rezenten meeresse-
core record of the -3C/'2C ratio of atmo- dimenten und ihre deutung am beispiel
spheric CO, in the past two centuries. einiger sedimentprofile aus der west-
Nature 324:237-38 lichen ostsee. Geol. Rund. 58:621-55
30. Fritz. P., Drimmie, R. J.. Nowicki, V. 44. Hayes, J. M. 1983. Practice and princi-
K. 1974. Preparation of sulfur dioxide ples of isotopic measurements in organic
for mass spectrometer analyses by com- geochemistry. In Organic Geochemistry
bustion of sulfides with copper oxide. of Contemporaneous and Ancient Sedi-
Anal. Chem. 46:164-66 inents, ed. W. G. Meinschein, 5:1-31.
31. Fry, B. 1981. Tracing shrimp migra- Bloomington, Ind: Great Lakes Sect.,
tions anid diets usinig niatlural variations Soc. Econ. Paleontol. Mineral.
in stable isotopes. PhD thesis. Univ. 45. Hitchcock, D. R., Black, M. S. 1984.
Tex., Austin. 225 pp. 34S/32S evidence of biogenic sulfur ox-
32. Fry, B. 1983. Fish and shrimp migra- ides in a salt marsh atmosphere. Atmos.
tions in the northern Gulf of Mexico Environ. 18:1-17
analyzed using stable C, N and S iso- 46. Hitchon, B., Krouse, H. R. 1972. Hy-
tope ratios. Fish. Bull. U.S. 81:789- drogeochemistry of the surface waters of
801 the Mackenzie River drainage basin,
33. Fry, B. 1986. Stable sulfur isotopic dis- Canada-ll. Stable isotopes of oxygen,
tributions and sulfate reduction in lake carbon and sulphur. Geoc-him. Cos-
sediments of the Adirondack Mountains, mochim. Acta 36:1337-57
New York. BiogeochemistrY 2:329-43 47. Hobbie, J. E., Cole, J., Dungan, J.,
34. Fry, B., Arnold, C. 1982. Rapid 13C/ Houghton, R. A.. Peterson, B. J. 1984.
'2C turnover during growth of brown The controversy on the role of the biota
shrimp (Penaeus aztecus). Oecologia in the global CO, balance. BioScience
54:200-04 34:492-498
35. Fry, B ., Sherr, E. B. 1984. 8' 3C 48. Hoering, T. 1957. The isotopic com-
ineasurements as indicators of carbon position of the ammonia and the nitrate
flow in marine and freshwater ecosys- ion in rain. Geochim. Cosmochim. Acta
tems. Contrib. Mar. Sci. 27:13-47 12:97-102
36. Fry, B., Macko, S. A., Zieman, J. C. 49. Houghton, R. A., Hobbie, J. E., Melil-
1987. Review of stable isotopic in- lo, J. M., Moore, B., Peterson, B. J., et
vestigations of food webs in seagrass al. 1983. Changes in the carbon content
meadows. In Proc. Subtropical-Tropical of terrestrial biota and soils between
Seagrasses of the Southeastern U.S., ed. 1860 and 1980: A net release of CO, to
M. Durako, R. Lewis, R. C. Phillips. the atmosphere. Ecol. Monogr. 53:235-
Fla. Mar. Res. Publ. 42. In press 62
37. Fry, B., Cox, J., Gest, H., Hayes, J. M.50. Jorgensen, B. B. 1979. A theoretical
1986. Discrimination between 334S and model of the stable sulfur isotope dis-
32S during bacterial metabolism of in- tribution in marine sediments. Geochim.
organic sulfur compounds. J. Bacteriol. Cosmochim. Acta 43:363-74
165:328-30 51. Junk, G., Svec, H. 1958. The absolute
38. Goldhaber, M. B., Kaplan, 1. R. 1975. abundance of the nitrogen isotopes in the
Controls and consequences of sulfate atmosphere and compressed gas from
reduction rates in recent marine sedi- various sources. Geochim. Cosmochim.
ments. Soil Sci. 119:42-55 Acta 14:234-43
39. Haines, E. B. 1977. The origins of de- 52. Kaplan, 1. R. 1983. Stable isotopes of
tritus in Georgia salt marsh estuaries. sulfur, nitrogen and deuterium in recent
Qikos 29:254-60 marine environments. In Stable Isotopes
40. Haines, E. B., Montague, C. L. 1979. in Sedimentary Geology, ed. M. A.
Food sources of estuarine invertebrates Arthur, T. F. Anderson, I. R. Kaplan, J.

This content downloaded from

140.122.143.127 on Sat, 14 Jan 2023 08:49:08 UTC
All use subject to https://about.jstor.org/terms
318 PETERSON & FRY

Veizer, L. S. Land, 2:1-108. Dallas: ing Pasture Management, 61-77. Vien-

Soc. Econ. Paleontol. Mineral. na: Int. Atomic Energy Agency
53. Kaplan, I. R., Rittenberg, S. C. 1964. 65. Mariotti, A., Lancelot, C., Billen, G.
Microbiological fractionation of sulphur 1984. Natural isotopic composition of
isotopes. J. Gen. Microbiol. 34:195- nitrogen as a tracer of origin for sus-
212 pended organic matter in the Scheldt es-
54. Kaplan, I. R., Emery, K. O., Ritten- tuary. Geochim. Cosmochim. Acta
berg, S. C. 1963. The distribution and 48:549-55
isotopic abundance of sulphur in recent 66. Mariotti, A., Pierre, D., Vedy, J. C.,
marine sediments off southern Califor- Bruckert, S., Guillemot, J. 1980. The
nia. Geochim. Cosmochim. Acta 27: abundance of natural nitrogen 15 in the
297-331 organic matter of soils along an altitu-
55. Keeling, C. D., Mook, W. G., Tans, P. dinal gradient. Catena 7:293-300
P. 1979. Recent trends in the I3C/ I2C 67. Mariotti, A., Germon, J. C., Hubert, P.,
ratio of atmospheric carbon dioxide. Na- Kaiser, P., Letolle, R., et al. 1981. Ex-
ture 277:121-23 perimental determination of nitrogen
56. Kohl. D. H., Shearer, G. 1980. Isotopic kinetic isotopic fractionation: Some
fractionation associated with symbiotic principles; illustration for the denitrifica-
N2 fixation and uptake of N03 by tion and nitrification processes. Plant
plants. Plant Phvsiol. 66:51-56 Soil 62:413-30
57. Kohl, D. H., Shearer, G., Harper, J. E. 68. Mekhtiyeva, V. L., Pankina, R. G.,
1980. Estimates of N2 fixation based on Gavrilov, Y. Y. 1976. Distributions and
differences in the natural abundance of isotopic compositions of forms of sulfur
'5N in nodulating and nonnodulating iso- in water animals and plants. Geochem.
lines of soybeans. Plant Physiol. 66:61- Int. 13:82-87
65 69. Minagawa, M., Wada, E. 1984. Step-
58. Krouse, H. R. 1980. Sulphur isotopes in wise enrichment of '"N along food
our environment. In Handbook of En- chains: Further evidence and the relation
vironmental Isotope Geochemistrv. between 6'5N and animal age. Geochim.
Volume I, ed. P. Fritz, J. Ch. Fontes, Cosmochim. Acta 48:1135-40
pp. 435-72. Netherlands: Elsevier Sci. 70. Minagawa, M., Winter, D. A., Kaplan,
Publ. I. R. 1984. Comparison of Kjeldahl and
59. LaCroix, M., Mosora, F. 1975. Varia- combustion methods for measurement of
tions du rapport isotopique '3C/'2C dans nitrogen isotope ratios in organic matter.
le metabolisme animal. In (Isotope Anal. Chem. 56:1859-61
Ratios as Pollutant Source and Be- 71. Miyake, Y., Wada, E. 1967. The
haviour Indicators.) Atomic Energy abundance ratio of '5N/'4N in marine
Agency, pp. 343-58. Vienna environments. Rec. Oceanogr. Works,
60. Libes, S. M. 1983. Stable isotope Jpn 9:37-53
geochemistrv of nitrogen in marine72. Montague, C. L., Bunker, S. M.,
par-
ticulates. PhD thesis. Woods Hole Haines, E. B., Pace, M. L., Wetzel, R.
Oceanogr. Inst./MIT. WHOI-83-9. 288 L. 1981. Aquatic macroconsumers. In
PP. The Ecology of a Salt Marsh, ed. L. R.
61. Macko, S. A., Lee, W. Y., Parker, P. Pomeroy, R. G. Wiegert, pp. 69-85.
L. 1982. Nitrogen and carbon isotope New York, Springer-Verlag. 271 pp.
fractionation by two species of marine 73. Montoya, J. P., Horrigan, S. G. 1986.
amphipods: Laboratory and field stud- Rapid changes in the natural abundance
ies. J. Exp. Mar. Biol. Ecol. 63:145- of 15N in plankton and dissolved nitro-
49 gen in the Chesapeake Bay. Eos 67:988
62. Macko, S. A., Estep, M. L. F., Hare, P. (Abstr.)
E., Hoering, T. C. 1983. Stable nitrogen 74. Mook, W. G., Bommerson, J. C.,
and carbon isotopic composition of in- Staverman, W. H. 1974. Carbon isotope
dividual amino acids isolated from cul- fractionation between dissolved bicar-
tured microorganisms. Carnegie Inst. bonate and gaseous carbon dioxide.
Wash. Yrbk 1982-1983, 404-10 Earth Planet. Sci. Lett. 22:169-76
63. Mariotti, A. 1983. Atmospheric nitro- 75. Moore, H. 1974. Isotopic measurement
gen is a reliable standard for natural 5N of atmospheric nitrogen compounds.
abundance measurements. Nature Tellus 26:169-74
303:685-87 76. Moore, H. 1977. The isotopic composi-
64. Mariotti, A. J., Mariotti, F., Amarger, tion of ammonia, nitrogen dioxide and
N. 1983. Use of natural 5N abundance nitrate in the atmosphere. Atmos. En-
in the measurement of symbiotic fixa- viron. 11:1239-43
tion. In Nuclear Techniques in Improv- 77. Nielsen, H. 1978. Sulfur Isotopes in

This content downloaded from

140.122.143.127 on Sat, 14 Jan 2023 08:49:08 UTC
All use subject to https://about.jstor.org/terms
 STABLE ISOTOPES IN ECOSYSTEMS 319

nature. In Handbook of Geochemistrv used to trace the flow of organic matter

11-2, ed. K. H. Wedepohl, 16B:1-40. in estuarine food webs. Science 227:
Springer-Verlag, Berlin 1361-63
78. Nriagu, J. O., Harvey, H. H. 1978. 92. Peterson, B. J., Howarth, R. W., Gar-
Isotopic variation as an index of sulphur ritt, R. H. 1986. Sulfur and carbon iso-
pollution in lakes around Sudbury, topes as tracers of salt-marsh organic
Ontario. Nature 273:223-24 matter flow. Ecology 67:865-74
79. Nriagu, J. O., Coker, R. D. 1978. 93. Rau, G. H. 1978. Carbon-13 depletion
Isotopic composition of sulphur in atmo- in a subalpine lake: Carbon flow im-
spheric precipitation around Sudbury, plications. Science 201:901-02
Ontario. Nature 274:883-85 94. Rau, G. H. 1980. Carbon-13/carbon-12
80. Nriagu, J. O., Coker, R. D. 1983. Sul- variations in subalpine lake aquatic in-
phur in sediments chronicles past sects: Food source implications. Can. J.
changes in lake acidification. Nature Fish. Aquat. Sci. 37:742-46
303:692-94 95. Rees, C. E. 1973. A steady-state model
81. Nriagu, J. O., Soon, Y. K. 1985. Dis- for sulphur isotope fractionation in
tribution and isotopic composition of bacterial reduction processes. Geochim.
sulfur in lake sediments of northern Cosmochim. Acta 37:1141-62
Ontario. Geochim. Cosmochim. Acta 96. Rees, C. E., Jenkins, W. J., Monster, J.
49:823-34 1978. The sulphur isotopic composition
82. Odum, E. P., de la Cruz, A. A. 1967. of ocean water sulphate. Geochim. Cos-
Particulate organic detritus in a Georgia mochim. Acta 42:377-81
salt marsh-estuarine ecosystem. In Es- 97. Roeske, C. A., O'Leary, M. H. 1984.
tuaries, ed. G. H. Lauff, 83:383-88. Carbon isotope effects on the enzyme-
AAAS Publ. New York catalyzed carboxylation of ribulose-
83. O'Leary, M. H. 1981. Carbon isotope bisphosphate. Biochemistrv 23:6275-
fractionation in plants. Phvtochemistrv 84
20:553-67 98. Rotty, R. M., Marland, G. 1986. Fossil
84. O'Leary, M. H. 1984. Measurement of fuel combustion: Recent amounts, pat-
the isotope fractionation associated with terns, and trends of CO,. In The Chang-
diffusion of carbon dioxide in aqueous ing Carbon Cycle: A Global Analysis,
solution. J. Phvs. Chem. 88:823 ed. J. R. Trabalka, D. E. Reichle, pp.
85. Owens, N. J. P. 1985. Variations in the 474-90. New York: Springer-Verlag
natural abundance of '5N in estuarine 99. Rounick, J. S., Winterbourn, M. J.
suspended particulate matter: A specific 1986. Stable carbon isotopes and carbon
indicator of biological processing. Es- flow in ecosystems. BioScience 36:171-
tuarine, Coastal Shelf Sci. 20:505-10 77
86. Pang, P. C., Nriagu, J. 0. 1977. Isotop- 100. Saino, T., Hattori, A. 1980. 15N natural
ic variations of the nitrogen in Lake Su- abundance in oceanic suspended particu-
perior. Geochim. Cosmochim. Acta late matter. Nature 283:752-54
41:811-14 101. Saltzman, E. S., Brass, G. W., Price,
87. Parker, P. L. 1964. The biogeochemis- D. A. 1983. The mechanism of sulfate
try of the stable isotopes of carbon in a aerosol formation: Chemical and sulfur
marine bay. Geochim. Cosmochim. Acta isotopic evidence. Geophys. Res. Lett.
28:1155-64 10:513-16
88. Peng, T.-H., Freyer, H. D. 1986. Re- 102. Schoeller, D. A., Minagawa, M., Sla-
vised estimates of atmospheric CO, vari- ter, R., Kaplan, I. R. 1986. Stable iso-
ations based on the tree-ring '3C record. topes of carbon, nitrogen and hydrogen
In The Changing Carbon Cycle: A Glob- in the contemporary North American hu-
al Analysis, ed. J. R. Trabalka, D. E. man Food web. Ecol. food Nutr.
Reichle, pp. 151-59 New York: 18: 159-70
Springer-Verlag 103. Schoeninger, M. J., DeNiro, M. J.,
89. Peng, T.-H., Broecker, W. S., Freyer, Tauber, H. 1983. Stable nitrogen iso-
H. D., Trumbore, S. 1983. A de- tope ratios of bone collagen reflect
convolution of the tree ring based h13C marine and terrestrial components of
record. J. Geophys. Res. 88:3609-20 prehistoric human diet. Science 220:
90. Peters, K. E., Sweeney, R. E., Kaplan, 1381-83
I. R. 1978. Correlation of carbon and 104. Schwarcz, H. P. 1969. Isotopes in na-
nitrogen stable isotope ratios in ture. I. The stable isotopes of carbon.
sedimentary organic matter. Limnol. In Handbook of Geochemistry II-1,
Oceanogr. 23:598-604 ed. K. H. Wedepohl, 6B:1-16. Berlin:
91. Peterson, B. J., Howarth, R. W., Gar- Springer-Verlag
ritt, R. H. 1985. Multiple stable isotopes 105. Schwarcz, H. P. 1973. Sulfur

This content downloaded from

140.122.143.127 on Sat, 14 Jan 2023 08:49:08 UTC
All use subject to https://about.jstor.org/terms
320 PETERSON & FRY

isotope analyses of some Sudbury, plants. Carniegie Inst. Wash. Yrbk

Ontario, ores. Cani. J. Earth Sci. 73:768-80
10:1444-59 116. Urey, H. C. 1947. The thermodynamic
106. Schwarcz, H. P., Melbye, J., Katzen- properties of isotopic substances. J.
berg, M. A., Knyfe, M. 1985. Stable Chem. Soc. 1947:562-81
isotopes in human skeletons of southern 117. van der Merwe, N. J. 1982. Carbon iso-
Ontario: Reconstructing Palaeodiet. J. topes, photosynthesis, and archaeology.
Archaeol. Sci. 12:187-206 Am. Sci. 70:596-606
107. Shearer, G., Kohl, D. H. 1986. N,- 118. van der Merwe, N. J., Vogel, J. C.
fixation in field settings: Estimations 1978. 13C content of human collagen as
based on natural '5N abundance. Austr. a measure of prehistoric diet in wood-
J. Plant Phvsiol. 13:699-757 land North America. Nature 276:815-16
108. Shearer. G., Kohl, D. H., Virginia, R. 119. Wada, E. 1980. Nitrogen isotope
A., Bryan, B. A., Skeeters, J. L., et al. fractionation and its significance in
1983. Estimates of N,-fixation from biogeochemical processes occurring in
variation in the natural abundance of '5Nmarine environments. In Isotope Marine
in Sonoran Desert ecosystems. Oecolo- Chemistry, ed. E. D. Goldberg, Y.
gia 56:365-73 Horibe, K. Saruhashi, pp. 375-98.
109. Steele, K. W., Daniel, R. M. 1978. Tokyo: Uchida Rotakuho
Fractionation of nitrogen isotopes by an- 120. Wada, E., Hattori, A. 1978. Nitrogen
imals: A further complication to the use isotope effects in the assimilation of in-
of variations in the natural abundance of organic nitrogenous compounds by
' 5N for tracer studies. J. Agric. Sci. marine diatoms. Geomicrobiol. J. 1:85-
90:7-9 101
110. Stuiver, M. 1978. Atmospheric carbon 121. Williams, P. M., Gordon, L. I. 1970.
dioxide and carbon reservoir changes. Carbon- 13: carbon- 12 ratios in dissolved
Science 199:253-58 and particulate organic matter in the sea.
111. Tans, P. 1981. '3C/'2C of industrial Deep-Sea Res. 17:19-27
CO,. In Carboni Cvcle Modelling, 122. Yanagisawa, F., Sakai, H. 1983. Ther-
SCOPE 16, ed. B. Bolin, pp. 127-129. mal decomposition of barium sulfate-
New York: Wiley vanadium pentoxide-silica glass mix-
112. Tauber, H. 1981. 13C evidence for di- tures for preparation of sulfur dioxide in
etary habits of prehistoric man in Den- sulfur isotope ratio measurements. Anal.
mark. Nature 292:332-33 Chem. 55:985-87
113. Teal, J. M. 1962. Energy flow in the salt 123. Yoneyama, T.. Ohta, Y., Ohtani, T.
marsh ecosystem of Georgia. Ecology 1983. Variations of natural 13C and '5N
43:614-24 abundances in the rat tissues and their
114. Tieszen, L. L.. Boutton, T. W., Tes- correlation. Radioisotopes 32:330-32
dahl, K. G., Slade, N. A. 1983. 124. Zieman, J. C., Macko, S. A., Mills, A.
Fractionation and turnover of stable car- L. 1984. Role of seagrasses and man-
bon isotopes in animal tissues: Im- groves in estuarine food webs: Temporal
plications for 8'3 C analysis of diet. and spatial changes in stable isotope
Oecologia 57:32-37 composition and amino acid content dur-
115. Troughton, J. H., Card, K. A., Hendy, ing decomposition. Bull. Mar. Sci.
C. H. 1974. Photosynthetic pathways 35:380-92
and carbon isotope discrimination by

This content downloaded from

140.122.143.127 on Sat, 14 Jan 2023 08:49:08 UTC
All use subject to https://about.jstor.org/terms