Skip to content

Composition of Seawater: Major and Minor Elements, Salinity and Residence Times

CSIR NET Ocean Sciences: salinity scales, major, minor and trace elements, constant proportions, element distribution types and residence-time box models.

  • 21 min read
  • Updated
On this page · 14 sections
  1. Prerequisites
  2. The Hook: A River Is Not Dilute Seawater
  3. Measuring the Total: Salinity
  4. What Is Dissolved: Major, Minor and Trace Constituents
  5. The Principle of Constant Proportions
  6. Conservative and Non-conservative: Three Profile Types
  7. Residence Time
  8. Memory Tricks
  9. Don’t Confuse
  10. Common Mistakes
  11. Practice
  12. Quick Revision Table
  13. Key Takeaways
  14. What’s Next

Prerequisites

The Hook: A River Is Not Dilute Seawater

How CSIR NET asks it. Statement sets on salinity scales and constant proportions; matching elements to their vertical-profile type (conservative, nutrient-type, scavenged); and numericals on residence time, steady state, and how fast a perturbed reservoir relaxes.


Measuring the Total: Salinity

Salinity is, in concept, the mass of dissolved salt in a kilogram of seawater. In practice it has never been measured by drying, because some components (HCl, CO\(_2\), water of hydration) are lost or retained unpredictably. Three operational definitions have been used.

ScaleBasisUnitStatus
Chlorinity (Knudsen, early 20th century; revised 1969)Titration of halides with AgNO\(_3\); \( S = 1.80655\,\text{Cl} \)‰ (g kg\(^{-1}\))Historical; still useful for ratios
Practical salinity \( S_P \) (PSS-78)Conductivity ratio of the sample to a standard KCl solution (32.4356 g KCl in 1 kg of solution) at 15 °C and 1 atm; \( S_P = 35 \) when the ratio is 1DimensionlessStandard for observations; “psu” is informal
Absolute salinity \( S_A \) (TEOS-10, adopted 2010)Reference composition plus a correction for non-conductive solutesg kg\(^{-1}\)Used in the equation of state

The link between the two modern scales:

$$S_A = \frac{35.16504}{35}\,S_P + \delta S_A \approx 1.0047\,S_P + \delta S_A$$

\( \delta S_A \) is small (a few hundredths of a g kg\(^{-1}\) at most) and largest in the deep North Pacific, where regenerated silicate, nitrate and dissolved inorganic carbon add mass without adding much conductivity. Silicic acid is essentially uncharged at seawater pH, so a conductivity probe cannot “see” it.


What Is Dissolved: Major, Minor and Trace Constituents

A practical classification by concentration (textbooks draw the lines slightly differently):

ClassConcentrationMembers
Major> 1 mg kg\(^{-1}\)Cl\(^-\), Na\(^+\), SO\(_4^{2-}\), Mg\(^{2+}\), Ca\(^{2+}\), K\(^+\), HCO\(_3^-\)/CO\(_3^{2-}\), Br\(^-\), Sr\(^{2+}\), B (as B(OH)\(_3\) and B(OH)\(_4^-\)), F\(^-\)
Minor1 µg kg\(^{-1}\) to 1 mg kg\(^{-1}\)Li, Rb, I, Ba, Mo, U, and the nutrients N, P, Si
Trace< 1 µg kg\(^{-1}\) (nmol kg\(^{-1}\) to pmol kg\(^{-1}\))Fe, Mn, Zn, Cu, Ni, Cd, Co, Pb, Al, rare-earth elements, Th

The major ions at \( S_P = 35 \)

Constituentg kg\(^{-1}\)mmol kg\(^{-1}\)Ratio to Cl (by mass)
Cl\(^-\)19.355461
Na\(^+\)10.784690.557
SO\(_4^{2-}\)2.7128.20.140
Mg\(^{2+}\)1.2852.70.066
Ca\(^{2+}\)0.41210.30.021
K\(^+\)0.39910.20.021
Br\(^-\)0.0670.840.0035
Sr\(^{2+}\)0.00790.0900.0004
Total B0.00450.42—
F\(^-\)0.00130.068—

Dissolved inorganic carbon (DIC) is about 2 mmol kg\(^{-1}\), mostly as HCO\(_3^-\). Note the order by moles: Cl > Na > Mg > SO\(_4\) > Ca ≈ K. By mass, sulfate comes before magnesium. Part B options exploit this switch.

Charge balance and ion pairing

The major cations and anions balance to within measurement error, which is a check on any analysis. Not all of each ion is free: Mg\(^{2+}\), Ca\(^{2+}\) and Na\(^+\) form ion pairs with SO\(_4^{2-}\) and CO\(_3^{2-}\). Monovalent ions (Na\(^+\), K\(^+\), Cl\(^-\)) are almost entirely free; a large part of sulfate and most of the carbonate ion are ion-paired. This matters in the carbonate system: equilibrium constants measured in seawater (the “stoichiometric” or conditional constants written \( K' \) or \( K^* \)) already include ion-pairing, which is why oceanographers use them rather than thermodynamic constants with activity coefficients.

Minor and trace constituents worth knowing

ElementTypical concentrationBehaviour
Liabout 26 µmol kg\(^{-1}\)Conservative; enriched in vent fluids
Rbabout 1.4 µmol kg\(^{-1}\)Conservative
Moabout 0.1 µmol kg\(^{-1}\)Conservative as MoO\(_4^{2-}\) in oxic water; removed in sulfidic basins
Uabout 14 nmol kg\(^{-1}\)Conservative as a uranyl carbonate complex; removed in anoxic sediments
Iabout 0.45 µmol kg\(^{-1}\)Iodate (oxic) and iodide (surface, anoxic): redox-sensitive
Batens of nmol kg\(^{-1}\)Nutrient-like, follows silicate and alkalinity
Fesub-nmol in surface HNLC waters to about 1 nmol kg\(^{-1}\) in deep waterHybrid; more than 99% bound to organic ligands
Alnmol kg\(^{-1}\)Scavenged; surface maximum from dust

“Dissolved” is an operational term: whatever passes a 0.2 or 0.45 µm filter. That fraction includes colloids, which is why the “dissolved” iron or thorium in a sample can still behave as a particle.


The Principle of Constant Proportions

Marcet first noticed, and Dittmar’s analysis of the Challenger samples confirmed, that the ratios of the major ions are almost the same everywhere in the open ocean, whatever the total salinity. Salinity varies (about 33 to 37 in the open ocean) because water is added or removed by rain, evaporation, rivers and ice; the salts are hardly touched.

Why it holds:

  1. The major ions have residence times of millions of years or more, far longer than the ocean’s mixing time of about 1,000 years. Any regional input is mixed throughout the ocean long before it can change the ratios.
  2. They are chemically unreactive at seawater concentrations: not needed by organisms in amounts that matter, and not adsorbed strongly onto particles.

Where it fails, and why:

SettingConstituent that departsCause
Estuaries, river plumesCa, HCO\(_3^-\), Si relative to ClRiver water has very different ratios
Anoxic basins and pore watersSO\(_4^{2-}\) lowBacterial sulfate reduction
Hydrothermal vent fluidsMg and SO\(_4^{2-}\) near zero; Ca, Li, K enrichedWater–basalt reaction at high temperature
Sea-ice brinesSO\(_4^{2-}\) and Na relative to ClMirabilite (Na\(_2\)SO\(_4\)·10H\(_2\)O) precipitates at very low temperature
Deep Pacific (small)Ca and Sr slightly higherCaCO\(_3\) dissolution
Restricted evaporating lagoonsCa, SO\(_4^{2-}\)Gypsum and carbonate precipitation

The calcium excess in deep water is real but small (about 1% of the calcium), which is why Ca is still listed among the “conservative” majors in Part B answers. Part C, however, can ask why deep-water Ca/Cl is a little higher than surface Ca/Cl: CaCO\(_3\) shells formed at the surface dissolve at depth.


Conservative and Non-conservative: Three Profile Types

The vertical profile of an element in the open ocean tells you how it is cycled.

graph TD
    A["Element in seawater"] --> B{"Reacts with
organisms or particles
within ~1000 yr?"} B -->|"No"| C["Conservative
uniform ratio to salinity
Na, Cl, Mg, K, SO4,
Li, Rb, Cs, Mo, U"] B -->|"Taken up by plankton,
regenerated at depth"| D["Nutrient-type
depleted surface, enriched deep
N, P, Si, Cd, Zn, Ni, Ba"] B -->|"Adsorbed on sinking
particles, not released"| E["Scavenged
surface or source maximum,
decrease with depth/age
Al, Mn, Pb, Th, Co"] D --> F["Hybrid: nutrient-like and scavenged
Fe, Cu"] E --> F
TypeProfile shapeDeep Pacific vs deep AtlanticResidence time
ConservativeFlat, tracks salinitySame (normalised to salinity)Long (\( 10^{5} \)–\( 10^{8} \) yr)
Nutrient-type (recycled)Surface depletion, deep enrichmentPacific higher (deep water older, has collected more regenerated material)Intermediate (\( 10^{3} \)–\( 10^{5} \) yr)
ScavengedMaximum near the source (surface dust, vents, margins), decreasing awayAtlantic often higher (nearer the dust and the source; Pacific deep water has been scavenged longer)Short (\( 10 \)–\( 10^{3} \) yr)

The inter-ocean contrast is a powerful Part C clue. Deep water forms in the North Atlantic and flows to the Pacific, ageing as it goes. Nutrient-type elements accumulate along the way (sinking particles keep releasing them), so the deep Pacific has more. Scavenged elements are removed along the way, so the deep Pacific has less. Details of the nutrient and trace-metal profiles are on the Nutrient and Trace-Metal Cycling page.


Residence Time

Definition and the steady-state box

For a well-mixed reservoir with inventory \( M \) and input and output fluxes \( F_{in} \), \( F_{out} \):

$$\frac{dM}{dt} = F_{in} - F_{out}$$

At steady state \( F_{in} = F_{out} = F \), and the residence time is

$$\boxed{\tau = \frac{M}{F}}$$

the mean time an atom spends in the reservoir before removal. If removal is first order (\( F_{out} = kM \)), then \( \tau = 1/k \) and a perturbed reservoir relaxes towards its new steady state as

$$M(t) = M_{ss} + \left(M_0 - M_{ss}\right) e^{-t/\tau}$$

Two useful consequences:

  • After a step change in input, the reservoir covers 63% of the way to its new steady state in one \( \tau \), and 95% in three.
  • An element whose residence time is short compared with the time scale of an input change tracks that change; one with a long residence time averages it out.

Typical residence times

ElementOrder of magnitude (yr)Why
Cl, Br\( 10^{8} \)Soluble, unreactive; removed mainly in evaporites and pore water
Na\( 10^{7} \)–\( 10^{8} \)Same, plus some cation exchange
Mg, SO\(_4\), K\( 10^{7} \)Mg and SO\(_4\) removed in hydrothermal circulation; K in clays
Li, U, Mo\( 10^{5} \)–\( 10^{6} \)Slow uptake in altered basalt and reducing sediments
Ca\( 10^{6} \)Biogenic CaCO\(_3\) burial
Si, N, P\( 10^{3} \)–\( 10^{5} \) (Si about \( 10^{4} \); P \( 10^{4} \)–\( 10^{5} \) in recent estimates)Biological uptake and burial; some removal each cycle
Mn, Fe, Al, Pb, Th\( 10 \)–\( 10^{3} \)Particle-reactive: scavenged onto sinking particles
Ocean water (relative to river inflow)about \( 3.6\times 10^{4} \)\( 1.34\times10^{9} \) km\(^3\) / \( 3.7\times10^{4} \) km\(^3\) yr\(^{-1}\)

Values in the literature differ by factors of a few because river fluxes and hydrothermal sinks are uncertain. Part B asks for the order of magnitude and the ranking, not a precise figure.

What sets the residence time

Residence time correlates with how strongly an element prefers solution over particles. Elements whose seawater concentration is high relative to their abundance in crustal rock (Cl, Na, Br, B) have long residence times; elements that are scarce in seawater relative to rock (Al, Fe, Ti, Th) have short ones. Part C may give a set of seawater/crust ratios and ask you to rank residence times. The rule is monotonic: larger ratio, longer \( \tau \).

Kinetic steady state versus equilibrium control

Two classic views of why seawater has the composition it does:

ViewIdeaStatus
Equilibrium model (Sillén)Seawater composition is fixed by chemical equilibrium with silicate minerals (clays), carbonates and the atmosphereUseful as a thought experiment; real removal is slow and partly kinetic
Kinetic steady state (box models, Broecker and others)Composition reflects a balance of input and removal fluxes, each with its own rateStandard framework today

Reverse weathering (proposed by Mackenzie and Garrels) is the bridge between them: in sediments, amorphous silica, dissolved cations and bicarbonate react with degraded clays to form new authigenic clay minerals, releasing CO\(_2\). It removes K, Mg and Si and consumes alkalinity. It was long dismissed as too slow, but rapid authigenic clay formation has been documented in tropical deltaic muds such as those of the Amazon shelf.

Budgets: sources and sinks of the major ions

IonMain sourcesMain sinks
Na\(^+\), Cl\(^-\)Rivers (partly recycled sea salt), pore-water expulsionEvaporites, pore-water burial, ion exchange (Na)
Mg\(^{2+}\)RiversHydrothermal circulation (Mg-silicate/clay formation in hot basalt), dolomite, reverse weathering
SO\(_4^{2-}\)Rivers (pyrite oxidation, gypsum dissolution)Pyrite formation in reducing sediments, evaporites (gypsum), hydrothermal anhydrite
Ca\(^{2+}\)Rivers, hydrothermal fluids (Ca released when Mg is taken up)Biogenic CaCO\(_3\) burial, gypsum
K\(^+\)Rivers, high-temperature ventsLow-temperature basalt alteration, authigenic clays
HCO\(_3^-\)Rivers (carbonate and silicate weathering)CaCO\(_3\) burial, reverse weathering

Worked example: calcium

A two-box ocean

The simplest model that separates biology from mixing (Broecker’s two-box ocean) has a warm surface box and a deep box exchanging water at a rate \( V_{mix} \), with rivers feeding the surface and particles sinking from surface to deep.

graph LR
    R["Rivers
Q × C_river"] --> S["Surface box
C_s"] S -->|"V_mix × C_s
(downwelling)"| D["Deep box
C_d"] D -->|"V_mix × C_d
(upwelling)"| S S -->|"Sinking particles B"| D D -->|"Burial"| SED["Sediment"]

For a nutrient at steady state in the surface box:

$$Q\,C_{river} + V_{mix}\,C_d = V_{mix}\,C_s + B$$

Two dimensionless numbers come out of it:

  • \( g = (C_d - C_s)/C_d \): the fraction of the element brought up to the surface that is stripped out by biology. For phosphate and nitrate \( g \) is close to 0.9 or higher (surface almost depleted); for a conservative element \( g = 0 \).
  • \( f \): the fraction of the sinking flux that is buried rather than regenerated, \( f = Q\,C_{river}/B \). For phosphorus only a percent or two of each pass is buried, so an atom makes tens of trips between surface and deep before burial.

Memory Tricks

  • “Cl Na Mg SO\(_4\) Ca K” by moles: “Clever Natives Make Salt Crystals Kindly”. By mass, swap Mg and SO\(_4\).
  • Profile types, “CNS”: Conservative is flat, Nutrient increases downward, Scavenged decreases away from its source.
  • Pacific versus Atlantic: “Nutrients pile up in the Pacific; scavenged metals are picked off before they get there.”
  • Residence time ladder: “Salt is old, shells are middle-aged, metals die young”: Cl/Na \( 10^{8} \), Ca \( 10^{6} \), Si/P \( 10^{4} \), Fe/Al/Th \( 10^{2} \).
  • Relaxation: one \( \tau \) gets you 63% of the way; three get you 95%.

Don’t Confuse

ThisIs not this
Practical salinity (dimensionless, from conductivity)Absolute salinity (g kg\(^{-1}\), TEOS-10, includes non-ionic solutes)
Constant proportions of major ionsConstant salinity (salinity varies; ratios do not)
Residence time (inventory/flux, one element in one reservoir)Mixing time (about 1,000 yr, the same for all dissolved species)
Nutrient-type profile (deep enrichment, Pacific > Atlantic)Scavenged profile (source maximum, Atlantic often > Pacific)
Conservative in the oceanographic sense (changes only by mixing and water balance)Inert (conservative elements are still added and removed, just slowly)
Stoichiometric constants \( K' \) (include ion pairing, in mol kg\(^{-1}\) seawater)Thermodynamic constants \( K \) (activities, infinite dilution reference)

Common Mistakes

Practice

All questions below are original practice questions written for this page. They are not from any CSIR NET paper.

Level 1: Foundation

  1. What is the unit of practical salinity?
  2. Rank Na\(^+\), Mg\(^{2+}\) and SO\(_4^{2-}\) by molar concentration in seawater.
  3. Give the formula for residence time at steady state.
Answers
  1. None: practical salinity is dimensionless.
  2. Na\(^+\) (469) > Mg\(^{2+}\) (52.7) > SO\(_4^{2-}\) (28.2) mmol kg\(^{-1}\).
  3. \( \tau = M/F \), inventory divided by the input (= output) flux.

Level 2: Part B Style

Practice Q1. A water sample has a chlorinity of 19.00‰. Its salinity on the Knudsen-type relation \( S = 1.80655\,\text{Cl} \) is about:

(a) 33.8 (b) 34.3 (c) 35.0 (d) 36.1

Answer: (b) 34.3

\( S = 1.80655\times19.00 = 34.32 \). Option (c) assumes every sample has \( S = 35 \); (a) and (d) are arithmetic slips.

Practice Q2. Which element is expected to have the highest concentration in deep North Pacific water relative to deep North Atlantic water?

(a) Sodium (b) Silicon (c) Aluminium (d) Lead

Answer: (b) Silicon

Silicate is nutrient-type: it is regenerated from sinking opal along the deep-water path, so it accumulates in the old deep Pacific (well above 100 µmol kg\(^{-1}\)) compared with the young deep North Atlantic. Sodium is conservative (no difference after salinity normalisation). Aluminium and lead are scavenged and are generally higher in the Atlantic, nearer dust and anthropogenic sources.

Practice Q3. The principle of constant proportions holds for the major ions mainly because:

(a) rivers have the same ionic ratios as seawater (b) the residence times of these ions far exceed the ocean mixing time (c) the ions are in equilibrium with sea-floor clays (d) evaporation removes ions in fixed ratios

Answer: (b)

Long residence times relative to about 1,000 yr of mixing homogenise the ratios. River ratios are very different (a); evaporation removes water, not ions (d); equilibrium control (c) is the old Sillén idea and does not explain uniformity.

Practice Q4. Which statement about absolute salinity (TEOS-10) is correct?

(a) It is dimensionless and defined only from conductivity (b) It exceeds practical salinity numerically by about 0.47% for standard seawater, plus a small composition correction (c) It ignores dissolved silicate (d) It is always lower than practical salinity

Answer: (b)

\( S_A \approx 1.0047\,S_P + \delta S_A \), in g kg\(^{-1}\). The correction \( \delta S_A \) exists precisely because silicate and other non-conducting solutes add mass.

Level 3: Part C Style

Practice Q5. A coastal lagoon of volume \( 2.0\times10^{8} \) m\(^3\) receives river water at \( 1.0\times10^{6} \) m\(^3\) day\(^{-1}\) and exchanges with the sea so that its volume stays constant. A dissolved pollutant enters only with the river at 5.0 mmol m\(^{-3}\) and leaves only by outflow; it does not react. Which statements are correct?

  1. The water residence time with respect to river inflow is 200 days.
  2. At steady state the lagoon concentration of the pollutant equals the river concentration if there is no seawater dilution.
  3. If the river input stops suddenly, the pollutant inventory falls to about 37% of its initial value after 200 days.
  4. If the pollutant were instead removed by first-order settling with a rate constant of 0.01 day\(^{-1}\) in addition to outflow, its residence time would be about 67 days.

(a) 1 and 3 only (b) 1, 2 and 3 only (c) 1, 2, 3 and 4 (d) 2 and 4 only

Answer: (c) 1, 2, 3 and 4
  • 1: \( \tau = 2.0\times10^{8}/1.0\times10^{6} = 200 \) days. True.
  • 2: with only river input and outflow at lagoon concentration, steady state requires \( Q C_r = Q C \), so \( C = C_r \). True.
  • 3: first-order flushing with \( k = 1/200 \) day\(^{-1}\): after one \( \tau \), \( e^{-1} = 0.37 \). True.
  • 4: total first-order rate \( k = 0.005 + 0.01 = 0.015 \) day\(^{-1}\); \( \tau = 1/0.015 = 66.7 \) days. True.

Practice Q6. Use the following (illustrative) data: ocean mass \( 1.4\times10^{21} \) kg; dissolved Li 26 µmol kg\(^{-1}\); total Li input (rivers plus vents) \( 2.0\times10^{10} \) mol yr\(^{-1}\). Which statements are correct?

  1. The Li inventory is about \( 3.6\times10^{16} \) mol.
  2. The Li residence time is about \( 1.8\times10^{6} \) yr.
  3. Lithium should show a nutrient-type profile with strong surface depletion.
  4. Lithium/salinity ratios should be nearly uniform in the open ocean.

(a) 1, 2 and 4 only (b) 1 and 3 only (c) 2 and 3 only (d) 1, 2, 3 and 4

Answer: (a) 1, 2 and 4 only
  • 1: \( 26\times10^{-6}\times1.4\times10^{21} = 3.64\times10^{16} \) mol. True.
  • 2: \( 3.64\times10^{16}/2.0\times10^{10} = 1.8\times10^{6} \) yr. True.
  • 3: false. A residence time about 1,800 times the mixing time and negligible biological demand make Li conservative.
  • 4: true, for the same reason.

Practice Q7. Three elements X, Y and Z have the following deep-water concentration ratios (North Pacific / North Atlantic, salinity-normalised): X = 1.00, Y = 4.5, Z = 0.3. Which assignment and inference are correct?

  1. X behaves like sodium; Y like silicate; Z like aluminium.
  2. Y has a shorter residence time than X.
  3. Z has the shortest residence time of the three.
  4. Y is regenerated from sinking particles along the deep-water path.

(a) 1 and 4 only (b) 1, 2, 3 and 4 (c) 2 and 3 only (d) 1, 3 and 4 only

Answer: (b) 1, 2, 3 and 4
  • 1: no inter-basin contrast = conservative (Na); strong Pacific enrichment = nutrient-type (Si); Pacific depletion = scavenged (Al). True.
  • 2: nutrient-type elements (\( 10^{3} \)–\( 10^{5} \) yr) cycle faster than conservative ones (\( 10^{7} \)–\( 10^{8} \) yr). True.
  • 3: scavenged elements have the shortest residence times. True.
  • 4: that is the meaning of nutrient-type accumulation. True.

Quick Revision Table

IdeaFact or formula
Knudsen relation\( S = 1.80655\,\text{Cl} \) (‰)
PSS-78conductivity ratio to KCl standard (32.4356 g kg\(^{-1}\), 15 °C); dimensionless
TEOS-10\( S_A \approx 1.0047\,S_P + \delta S_A \), g kg\(^{-1}\); \( \delta S_A \) largest in deep North Pacific
Major ions (mmol kg\(^{-1}\))Cl 546, Na 469, Mg 52.7, SO\(_4\) 28.2, Ca 10.3, K 10.2
Major ions by massCl > Na > SO\(_4\) > Mg > Ca ≈ K
Constant proportionsMarcet, Dittmar (Challenger); fails in estuaries, vents, anoxic basins, brines
Profile typesconservative (flat), nutrient (deep max, Pacific > Atlantic), scavenged (source max, short \( \tau \))
Residence time\( \tau = M/F \) at steady state; \( \tau = 1/k \) for first-order removal
Relaxation\( e^{-t/\tau} \): 63% after one \( \tau \)
Typical \( \tau \)Cl \( 10^{8} \); Na \( 10^{7} \)–\( 10^{8} \); Ca \( 10^{6} \); Si, P \( 10^{4} \); Al, Fe, Th \( 10 \)–\( 10^{3} \) yr
Mixing timeabout 1,000 yr
Mg sinkhydrothermal circulation (vent fluids Mg ≈ 0)
Reverse weatheringauthigenic clay formation; removes K, Mg, Si, alkalinity; releases CO\(_2\)
Two-box model\( g = (C_d - C_s)/C_d \); \( f \) = buried fraction per pass

Key Takeaways

What’s Next