Odekon M. Encyclopedia of paleoclimatology and ancient environments

Подождите немного. Документ загружается.

Interglacial-glacial extent of permafrost

The extent of permafrost reached an average 44



N latitude at

low altitude during the Late Glacial Maximum (LGM). Its

extent during the Holocene Optimum is restricted to 50



along the east side of continents (Bering Strait) and to 65–



N on the western sides. Its maximal extent, during the cold-

est episodes of the Quaternary, was not more significant than

those of the LGM, due to the landmass configuration. During

the Holocene Optimum, permafrost extent was restricted within

the polar circle in the North. As a result of the lowering of the

solar summer energy input during the second half of the Holo-

cene, the southern re-expansion of permafrost today reaches

about one-third of its LGM extent.

Thermal cracking

Thermal cracking occurs when temperature drops around 15



in a dozen hours, such as from 0 to 15



C and from 15 to

30



C (Mackay, 1990). The efficiency of the thermal contrac-

tion is controlled directly by the ice content of the ground, even

in peat, because the ice retraction coefficient is much larger

than that of most of the rocks or sediments. This is not specific

to the permafrost environment. The crack may be shallow or

many meters deep. It develops either as an isolated event or

in polygonal nets with a 1–40 m mesh, even on rather steep

slopes (up to 22



)(Figure P54). The net is usually pentagonal

or hexagonal in poorly-stratified material. In alluvial plains, it

is often square. When weaknesses such as tension faults pre-

exist, they control the development of the net. In the High Arc-

tic today or in pleniglacial deposits, thermal cracking is located

only in waterlogged sites (oasis in the cold desert).

When thermal cracking occurs in frost-susceptible grounds,

soil wedges form by percolating soil suspension into the open

fissure in spring. In gravel, when the crack opens, gravel and

stones may fall into the furrow.

When permafrost exists, the crack that opens in winter is

filled up by impure melt water (from snow and active layer),

which is preserved in the form of refrozen ice, forming an ice

vein. Year by year, as the thermal contraction follows pre-

existing weaknesses, the accumulation of successive ice veins

leads to the development of ice wedges from about 20 cm wide

for recent forms up to 30 m for very old forms. The ice wedge

develops exclusively in the upper permafrost horizons and the

ice is vertically foliated.

In a very dry climatic context, eolian sand replaces

most of the ice, leading to sand wedges with vertically lami-

nated infill. Usually, lateral stresses deform the permafrost

horizon with upturning of the adjacent strata. In the active

layer, deformations are linked with the width of the ice wedge,

and the rheology of the unfrozen material and its climato-

logical history. Cryoturbation and thermal cracking may be

synchronous.

Figure P54 Examples of periglacial land forms; (a) thermal cracking, NW Svalbard, average width of polygons 30 m; (b) patterned ground (sorted

circles), NW Svalbard; (c) cryoturbation (Ga

sebu, Svalbard); (d) thufur or hummocks, Northern Iceland.

PERIGLACIAL GEOMORPHOLOGY 771

These forms are further deformed by thermokarst and differ-

ential frost heave, according to the lithology and ice content of

the infill when thaw occurs.

The significance of thermal cracking is limited; ice wedges

reveal very short but abrupt cooling events in a continuous per-

mafrost context. Fully vertically stratified sand wedges trace

thermal cracking in dry climate with continuous permafrost

(Romanovskii, 1985).

Patterned ground

These forms are the most spectacular and beautiful of the peri-

glacial landscape (Figure P54).

Ground ice and differential frost heave

Frost may be considered as a peculiar type of thermally

induced desiccation. This means that most of the cryogenic

(micro) fabrics may be linked to shrinkage related to desicca-

tion induced by low vapor tension and crystallization, frost

heave pressure or frost induced swelling, and gravitational

movements during thaw consolidation (sliding or setting). This

heave is due to ice segregating in the soil as lenses and reticu-

lated crystallizations.

Ice can nucleate at the surface of a wet soil during quick

cooling and form pipekrake or needle ice, which lifts stones,

or at depth in desiccation fissures or nets, segregating ice lenses

or reticulate ice. Ice segregations cut the sediment in aggre-

gates and generate the cryogenic fabric.

Frost heave is controlled by: (a) a moisture content that is

optimal at the field capacity and the maximal water retention

in drained conditions, (b) a thermal gradient leading to a cold-

ward water suction, and (c) the particle size distribution

(hydraulic conductivity) of the sediments. Local differences in

temperature, porosity, specific surface of the mineral/organic

matter fraction in connection with the particle size, and vegeta-

tion cover lead to differential frost heaving. Soil horizons with

low water retention freeze first and those richer in fine particles

freeze later because of their higher water retention. They heave

more intensely because of a longer and a more efficient suction

driving migration of capillary water to the growing ice segrega-

tion. Soils that easily develop a high content in segregated ice

are called frost susceptible. Very small differences in water

retention or permeability can lead to differential frost heave

and frost susceptibility gradients (Van Vliet, 1998). Deforma-

tions known as cryoturbations develop by differential heave

where sediments are stratified or translocated. An upward or

downward gradient of frost susceptibility controls the geometry

of the deformations. A regular network of fissures (thermal or

desiccation net) assists the cryoturbation. Deformations develop

both in waterlogged and drained sites. These are noticeable on

flat land surfaces as well as on slopes, irrespective of density

gradients. Deformations evolve in morphology along a topo-

graphic catena; the deformation morphology and its evolution

with ageing will be ruled by the drainage quality and its evolu-

tion with time.

Desiccation cracking, pattern ground and hummocks

Because of the ice segregation process, all the soils are wetter

on the surface from the onset of thaw. Every year, by late fall,

the active layer or the seasonally frozen ground starts freezing

both from the soil surface and from the permafrost table.

The following spring, the soil drains progressively from the

onset of thaw, leading to shrinkage pattern expression. In more

or less frost-susceptible material, desiccation cracks develop

further with drainage, creating small polygonal forms up to

1 m in diameter. Frost fissures or desiccation fissures are the

most common features associated with downward bending

of the stratification. These fissures are at the starting point

for most periglacial deformations.

Polygonal forms evolve from flat microtopography in water-

logged sites to hummocky or raised small mounds in drained

conditions (Tedrow, 1977). Mounds vary from purely mineral

to holorganic. They can be associated with cryoturbation, with

stone jacking, and with solifluction. In this case, the material is

stretched and forms sorted stripes.

Polygonal forms do not necessarily imply permafrost.

Nevertheless, permafrost induces a high water table and

enhances frost heave and frost jacking of stones. Polygons

and hummocks develop during a cooling event such as a

Heinrich event. Further warming, by enhancing the drain-

age, favors raised hummocky forms, leading sometimes to

pseudo-convective forms.

Cryoturbation induces deformation similar to gilgai, pro-

duced by differential swelling by hydration. Cryoturbations

are often confused with loading or explained by “thermal con-

vection” not compatible with the hydrothermal regime of per-

mafrost soils. This is related to the high frequency of

cryoturbations in the same regions as tremors during the onset

of deglaciations.

Stone jacking

Frost jacking is the traction of stone towards the soil surface by

frost heave. This process is very fast and leads to vertically

oriented stones and to loose gravel accumulation at the soil sur-

face. This process is assisted by needle ice, which allows dis-

placement of the lifted stones downslope. When the soil

develops a polygonal pattern of various sizes, stones are accu-

mulated in the furrow at the top of the fissure: the form

becomes sorted. Similarly, if the fissure reaches a level of

frost-shattered rocks at depth, differential frost heave leads to

an upward migration of stones. Both processes generate sorted

polygons. When frost susceptible material rises to the surface

and pushes the superficial gravel laterally, it forms sorted cir-

cles. The morphological shaping induced by drainage is also

valid in this case, with low and raised forms. When the stone

circles are deformed by frost creep on slopes, they evolve into

sorted stripes. These forms do not necessarily imply perma-

frost. Their climatic signature is similar to that of the normal

patterned grounds.

Slope deformation

Periglacial solifluction is mass movement controlled by soil

creep, which results from a combination of thaw settlement

and sliding on slopes under gravity; it is directly related to

the content of ice and to the water supply during thaw. It is at

least ten times more efficient than common soil creep

(thermo-hydric creep) (Figure P55).

Permafrost creep and rock glacier

Soils rich in segregated ice, or regolith rich in ice, flow down at

a very low annual rate in relation to the ice plasticity, just as in

glaciers. The origin of the ice may be segregated, refrozen

meteoric water, or buried snow or ice. Screes and rock falls

present a glacier-like mobility when ice-saturated. The rock

glacier is a peculiar type, intermediate between a mobile scree

and a decaying glacier.

772 PERIGLACIAL GEOMORPHOLOGY

Mobile screes and rock glaciers are representative of dry,

precipitation-starved environments, usually in discontinuous

to continuous permafrost zones. They usually develop during

cooling events, following a deglaciation.

Frost creep, gelifluction, and mud flow

Specific versions of solifluction in periglacial system are

enhanced by soil freezing. Frost creep is a slow laminar flow

of about 1cm per freezing cycle. In soil that is frost heaved

orthogonally to the slope, frost creep develops from the onset

of thaw, starting on very low angle slopes (2–22



). The displa-

cement is the summation of a gravity collapse by thaw conso-

lidation, of a lateral sliding on the roof of the successive

sheets of ice lenses, and of shrinking that results from drainage.

The more abundant that water is during thaw, the more impor-

tant the sliding component is. Below a snow patch or in a down

slope position, this is fulfilled by oblique seepage. This process

is common in deep seasonal frozen soils as well as in the active

layer of permafrost (Harris, 1988). Frost creep is syngenetic

with frost jacking, cryoturbation, and particle translocation.

Gelifluction is an accelerated form, characterized by a turbu-

lent flow on a centimeter scale even if the total mass flows in a

laminar mode. It is linked with more abundant seepage, and is

more frequent on permafrost. The permafrost table restricts the

water flow to the active layer only. In that case, the cryogenic

fabric is stable and the form is drained when thaw proceeds:

oversaturation is never reached.

When the cryogenic fabric is unstable, it collapses at thaw

and the excess water cannot be evacuated: oversaturation is

raised, leading to hydrostatic instability and mud flow. This is

also common in thermokarst conditions.

Common deformations are terminal curvature of beds and

solifluction deforming by buckling and stretching the strata,

sometimes with syngenetic cryoturbation (permafrost). They

are usually shallow (<1 m max) and tend to diminish and van-

ish with depth. Deformations and clast alignments are related to

ice content and moisture supplied during thaw: flat lying clasts

with their long axes oriented down slope indicate excess moist-

ure from snow melt; vertical clasts indicate dominant frost

heave. The morphology is commonly lobbed or elongated par-

allel to the slope. It occurs in pure mineral soils as well as in

organic material. Vegetation, as shrubs or grassy mats, and

frost-jacked gravel accumulations may support the fronts of

some features, leading to terraces or lobate morphology.

Frost creep has no specific climatic association as only frost

is needed. It occurs from subtropical to high latitude or altitude,

and on permafrost as well as in seasonally frozen ground. The

presence of permafrost is indicated by a regionally higher fre-

quency of gelifluction.

Frost shattering

Frost shattering is controlled by the porosity of the rock, espe-

cially the closed porosity and the pre-existing fissility. It can

result from three mechanisms: the change of the pore water

volume by freezing, acting like a wedge; induced hydrofrac-

turation of the rock or water expulsion; and ice growth by seg-

regation in fissures when some accumulated debris or organic

matter (in situ or by translocation) induces water retention –

mechanisms are similar to those in soil. The efficiency of the

shattering is linked with the rapidity of frost, allowing trapping

of water into the porosity and a sudden rise in pressure.

In the field, paleo-weathering or hydrothermal activity, tec-

tonism, and previous thermal and hydric (dilation) shattering

commonly enhance frost shattering. Frost shattering fragments

the rock into blocks, flakes, shards, and particles, down to

sandy or loamy size following the lithology. The sorting is nor-

mally poor except for some isograined lithologies (Lautridou

and Ozouf, 1982). In situ frost shattering and cliff frost shatter-

ing produce screes, stratified slope deposits (“grèzes litées,”

“heads”), and debris flows. Snow patches may locally enhance

the frost shattering and permit the formation of nivation benches

or terraces. Salt can emphasize the process, by lowering the

freezing point. This is common on the coast. Along lakes or sea-

shore, frost shattering, associated with the action of drift ice,

facilitates the formation of abrasion surfaces and littoral

notches. Normally, frost shattering is associated with moisture

and the speed of frost penetration. It is inactivated within perma-

frost as in frozen caves. Climatic significance is thus limited.

Thermokarst

When the ground contains an excess of ice (ground ice,

injected ice, or buried ice of various origins), thawing leads

to subsidence, which results in sink-holes or retrogressive slid-

ing. Simple subsidence of 8–25 m of an important portion of

Figure P55 Examples of slope deformation in a periglacial environment; (a) rock glacier (Chambeyron, French Alps); (b) solifuction lobes

(Northeast Iceland).

PERIGLACIAL GEOMORPHOLOGY 773

the ground allows the formation of sink-holes, often associated

with normal faulting. The hollows are commonly occupied by

a circular lake, which is further extended by the undercutting of

waves. It results in a landscape of oriented lakes. In the pre-

sence of a polygonal network of ice wedges, thermokarst activ-

ity develops firstly polygonal furrows occupied by water and

later, when drained, a specific hilly morphology called “bad-

jarak” by the Russian authors (Solovyev, 1973). As the drai-

nage is modified, melting progresses retrogressively upwards.

On a slope, a modification of the thaw depth and drainage

may induce the thawing of the ice-rich upper horizon of the

permafrost, allowing downward collapse and slumping of

water-saturated sediments. Mud flows can be generated. As

the drainage is modified, melting progresses retrogressively

upwards with slumps (MacRoberts, 1978). This process may

lead to a cyclic landscape rejuvenation, as in the high Arctic.

Thermokarst is generated by various disturbances: (a) by

warming, such as during the late glacial and the early

Holocene; (b) by fire modifying the local energy budget, by

destroying the vegetation cover; (c) by river incision, which

alters the drainage and allows a deeper thaw; and (d) by anthro-

pogenic perturbations or global warming, which like fire can

modify the energy budget (Burn, 1997).

Periglacial mounds

Two main types exist: the pingo and the palsa. The fossil traces

are similar except for size and some specific details (Figure P56).

Pingo

These are the largest forms, often more than 50 m in diameter

(they can be several hundred meters in diameter) and up to

60 m high. They generally occur isolated. They are specifically

linked to the growth of a hydrolaccolith (injected massive ice),

to permafrost aggradation squeezing a water table in a former

lacustrine depression, or to lateral seepage from a slope talik

or unfrozen volume in the permafrost. This form lifts up sedi-

ments until the pressure of the water balances the overburden

pressure. Tension fissures may open at its top. The basal layer

is porous or fractured, allowing water circulation.

During the growth and the decay phases, solifluction devel-

ops on the sides. After the melting, a circular lake remains in

a sink-hole surrounded by a rather high, residual circular

rim. Upturned layers dipping towards the outer part of the

structure exist in the rim and at depth, with a “cone-in-cone”

collapsed structure (thermokarst) (De Gans, 1988). Lacustrine

deposition and peat formation occur in the lake. These mounds

are specific to a continuous permafrost environment and

indicative of a cooling trend during the end of an interstadial

or an interglacial.

Palsa

These forms range from rather small sizes (1–5 m high) up to

rather large forms, reaching up to 15 m high and 50 m wide.

They develop in waterlogged depressions (mire) by segregation

of ice in peat or sediments, usually when the wind drift in win-

ter reduces the thermal insulation by the snow cover (Seppällä,

1988). The surficial layer is desiccated when uplifted, and acts

as a thermal insulation in summer. Some palsas are purely

mineral soils. The mounds occur in clusters, often forming

oriented plateaus, sometimes affected by thermal cracking.

The ice segregation is reticulated in lacustrine or glacio-marine

deposits, or forms ice lenses in sectors affected by impeded

drainage connected with climate cooling.

During the growth and decay phase, solifluction develops

on the sides. The decay is not necessarily linked to a warming:

the drainage of the peat allows its fracturing and warmth pene-

tration, and solifluction develops on the sides, stretching the

peat layer; wind abrasion may also destroy the peat cover and

favor the thermokarst collapse. After the melting, a circular

lake subsists in a sink-hole surrounded by a residual, circular

rim. Upturned (peat) layers dipping towards the outer part of

the structure exist in the rim (Pissart, 1983) with no peculiar

organization at depth (thermokarst). Lacustrine deposition and

peat formation occur in the thaw lake.

These mounds are specific to sporadic to discontinuous per-

mafrost zones. Their development is specifically epigenetic

during a cooling trend that follows a long, humid period such

as an interstadial or an interglacial. Syngenetic forms develop

in a rather stabilized climate.

Brigitte Van Vliet-Lanoë

Bibliography

Burn, C., 1997. Cryostratigraphy, paleogeography and climate change dur-

ing the early Holocene warm interval, western Arctic coast, Canada.

Can. J. Earth Sc., 34, 912–925.

De Gans, W., 1988. Pingo scars and their identification. In Clark, M.J.

(ed.), Advances in Periglacial Geomorphology, Chichester: Wiley, pp.

299–322.

Figure P56 Examples of periglacial mounds; (a) Pingo (Mackensie delta, Canada); (b) Palsa field (Ungava, Canada).

774 PERIGLACIAL GEOMORPHOLOGY

Harris, C., 1988. Mechanisms of mass movement in periglacial environ-

ments. In Anderson, M.G., and Richards, K.S. (eds.), Slope Stability.

Chichester: Wiley, pp. 531–559.

Lautridou, J.P., and Ozouf, J.C., 1982. Experimental frost shattering. 25

years of research at the Centre de Géomorphologie du CNRS. Prog.

Phys. Geogr., 6(2), 215–232.

Mackay, J.R., 1990. Some observations on the growth and deformation of

epigentic, syngenetic and anti-syngenetic ice wedges. Permafrost Peri-

glacial Proc., 1:15–29

MacRoberts, E.C., 1978. Slope Stability in cold regions. In Andersland, O.B.,

and Anderson, D.M. (eds.), Geotechnical Engineering for Cold Regions,

New York: McGraw Hill, pp. 363–404.

Pissart, A., 1983. Remnants of periglacial mounds in the Hautes fagnes

(Belgique); structure and age of the ramparts. Geologie en Mijnbouw,

62(4), 551–555.

Romanovskii, N.N., 1985. Distribution of recently active ice and soil

wedges in the USSR. In Church, M., and Slaymaker, O. (eds),

Field and Theory – Lectures in Geocryology, Vancouver: University

of British Columbia Press, pp. 154–165.

Seppälä, M., 1988. Palsas and related forms. In Clark, M.J. (ed.), Advances

in Periglacial Geomorphology. Chichester: Wiley, 247–278.

Solovyev, P.A., 1973. Thermokarst phenomena and landforms due to frost

heaving in Central Yakoutia. Biuletyn Peryglacjalny, 23, 135–155.

Tedrow, J.C.F., 1977. Soils of the Polar Landscape. New Brunswick, NJ:

Rutgers University Press, 638pp.

Van Vliet-Lanoë, B., 1998. Frost and soils: Implications for paleosols,

paleo-climates and stratigraphy. Catena, 34, 157–183.

Van Vliet-Lanoë, B., and Hallégouët, B., 2001. European permafrost at the

LGM and at its maximal extent. The Geological Approach, 1999. In

Paepe, R., and Melnikov, V. (eds), Permafrost Response on Economic

Development, Environmental Security and Natural Resources, Dor-

drecht: Kluwer, pp. 195–213.

Washburn, A.L., 1979. Geocryology. A Survey of Periglacial Processes and

Environment. London: Edward Arnold publications, 406pp.

Williams, P.J., and Smith, M.W., 1989. The Frozen Earth. Fundamentals

of Geocryology. Studies in Polar Research, Cambridge: Cambridge

University Press, 306pp.

Cross-references

Astronomical Theory of Climate Change

Cenozoic Climate Change

Cryosphere

Glacial Eustasy

Glacial Geomorphology

Glacial Sediments

Heinrich Events

Last Glacial Maximum

Loess Deposits

Millennial Climate Variability

Orbital Variation and Climate

Paleosols–Quaternary

Patterned Ground

Pingos

Pleistocene Climates

PHOSPHORITE

Phosphorite (often referred to as phosphate rock) is a sedimen-

tary rock of which the main mineralogical constituent is fran-

colite (carbonate-fluor-apatite, CFA). The above definition

excludes igneous phosphate-rich rocks as well as continental

deposits (e.g., guano; a few lacustrine deposits) Whereas most

sedimentary rocks contain less than 1% of P

, its content in

phosphorites is above 18%, (>50% in francolite) and may reach

values as high as 37%. Usually phosphorus content is expressed

as P

percentage. An alternative unit is BPL (bone phosphate

of lime or tricalcium phosphate), where BPL = 2.185  %P

Since antiquity have farmers applied lime-treated bones to

soil as a means of improving soil fertility. However it was only

in the mid 19th century that J. von Liebig suggested the use of

phosphorite as a fertilizer, and J. Lawes (1842) invented the

dissolution of phosphorite in sulfuric acid to produce a more

soluble and efficient fertilizer. The study of phosphorites

received much attention beginning with the late thirties of

the 20th century, when the demand for fertilizers increased

dramatically. Research boomed after WW II, and the last dec-

ades of the previous century saw several comprehensive

reviews of the subject in the literature (Cook, 1976; Bentor,

1980; Kolodny, 1981; Sheldon, 1980; Notholt and Jarvis,

1990; Föllmi, 1996, Burnett and Froehlich, 1988). The three

volumes of “Phosphate Deposits of the World” were the fruits

of a concentrated effort of an international research program

(IGCP156) between 1977 and 1984 (Cook and Shergold,

1986; Notholt et al. 1989; Burnett and Riggs, 1990). The three

multi-authored volumes are a comprehensive summary of all

important phosphorite deposits of the world.

Distribution

Phosphorites are found in marine deposits of all ages, from the

Proterozoic to the Recent seafloor (Burnett and Riggs, 1990,

Cook and Shergold., 1986, Notholt et al., 1989, Figures P57,

P57). Phosphorites have been found on all continents except

Antarctica. Cook and McElhinny (1979)(Figure P57)have

shown that the distribution of phosphorites in the record is

uneven: “phosphate giants” dominate geological history with

peaks in the Cambrian, Permian, and Mio-Pliocene. The young

and Recent phosphorites on the shelves around America, the

western coast of Africa, and surrounding Australia

(Figure P58) were the object of intensive study since their dis-

covery by Murray and Renard in 1891, especially since the iden-

tification of localities of present phosphogenesis by Baturin in

1971 (Baturin, 1981). The most common lithological association

is that of phosphorites with cherts and black shales (reflecting the

geochemical nutrient trinity of P-Si-C). This is the association

typical of many phosphorite “giants” such as the Permian Phos-

phoria Formation of the western United States, the Cretaceous-

Eocene phosphorites of the southern Tethys, and the Karatau

(Kazakhstan) and Khubsugul (Mongolia) Cambrian deposits.

In contrast to this “West Coast” type association, an assemblage

of phosphorites with carbonates and siliciclastic rocks has

been described from the Neogoene of the East Coast of the US.

Surface weathering of primary marine phosphates may result in

residual type deposits such as the Holocene brown-rock weather-

ing products of Ordovician phosphatic limestones in Tennessee.

Various authors (e.g., Bentor, 1980) have mentioned evaporites,

dolomite, Mg-rich clays (e.g., palygorskite), and glauconite as

being paragenetically related to phosphorites.

Mineralogy and petrology

The chemical formula of francolite, the major phase in phos-

phorite, can be approximated by

10abc

ðPO

6x

ðCO

xyz

ðCO

FÞ

ðSO

ð1Þ

where x =y+a +2c, x varies from near zero to 1.5, y = 0.4x on

average, and b = 0.4a. c is the number of Ca vacancies

(Nathan, 1984). The key substitution seems to be that of a pla-

nar (CO

)

2–

triangle replacing a tetrahedral (PO

)

3–

ion.

PHOSPHORITE 775

Figure P59 summarizes only the most important substitutions

in the very “open” apatite structure.

Phosphorites may be texturally classified into: (a) bedded

phosphorites, (b) bioclastic phosphorites, (c) nodular phosphor-

ites, and (d) pebble-bed phosphorites. Francolite occurs in

phosphorites as laminae, pellets, ooids, crusts, nodules of var-

ious irregular shapes, skeletal shell and bone fragments, and

cements. Phosphorite types may range from predominantly pel-

letal to purely bioclastic – or “bone beds.” Glenn et al. (1994),

distinguished between pristine phosphate-bearing deposits,

Figure P57 Estimated abundance of phosphorite deposits and P

tonnage in Phanerozoic sediments (from Cook and McElhinny, 1979).

Figure P58 Distribution of major and important sedimentary phosphorite deposits. The symbols mark the age of the deposit. Note the

South-Tethyan Mesozoic belt that extends from Turkey via Morocco to Mauritania and Senegal; it extends across the Atlantic to Venezuela and

Colombia (after Cook, 1976).

776 PHOSPHORITE

which lack any sign of reworking, condensed phosphatic lami-

nae in which particles were concentrated by winnowing, and

allochthonous redeposited particles.

The weathering zone of phosphorites often includes several

Al- and Fe-phosphates such as wavellite and millisite.

Geochemistry

The large amount of substitutions in francolite makes it a most

sensitive recorder of the environments in which it forms.

Major elements

Most unweathered seafloor francolites contain about 6.3% CO

and between 3 and 4% F. Carbonate and fluorine concentrations

are correlated, probably reflecting the coupled substitution of

(CO

3–

for phosphate. Several interpretations have been offered

for this variability: (a) temperature variation of the depositional

basin, causing variable degrees of CO

saturation; (b) late diage-

netic burial effects, metamorphism and weathering; and (c) CO

2–

content of the interstitial water , i.e., carbonate alkalinity.

Trace elements

Phosphorites are enriched in a large number of trace elements.

Many of these enrichment factors can actually be attributed to

the organic matter associated with the phosphorites rather than

to the francolite itself. Still, after accounting for the organic

matter, a strong enrichment in Cd, Ag, Mo, Se, U, Y, Zn,

REE, and V is probably related to apatite. Some of these have

been studied for their importance as valuable by-products

(U, REE), others because of their toxicity, which requires their

removal from the marketable product (Cd and U in production

of phosphogypsum). We shall consider here only those from

which paleo-environmental information has been retrieved.

REE. The rare-earth elements (La–Lu) and Y apparently

substitute for Ca

in the francolite lattice. Both total REE con-

centration and the relative abundance of the different REE in

phosphorites vary. It is customary to consider such abundances

on a shale- (or chondrite-) normalized distribution graph, with

atomic number as the abscissa and log (normalized abundance)

as the ordinate. In such space, “anomalies” can be conveniently

expressed. Thus a Ce anomaly (Ce*) in a rock is defined as



¼ logð3Ce

=ð2La

þ Nd

ÞÞ ð2Þ

where the subscript SN denotes a shale-normalized concentra-

tion. Many phosphorites display a typical “seawater pattern”:

they are strongly depleted in Ce and enriched in the heavy

REE (HREE; Ho–Lu). These include the western US Permian

Phosphoria Formation, as well as many of the Mesozoic Tethys

belt deposits: Morocco, Egypt, Jordan (Jarvis et al., 1994). On

the other hand, numerous deposits (offshore Namibia, Agulhas

Bank, off Peru, Karatau, Kazakhstan) have shale-like (flat) pat-

terns with no Ce anomaly. The Ce anomaly has been attributed

to the oxidation of Ce

in the deep sea to Ce

, which is then

sequestered from seawater by ferromanganese oxides. Indeed

modern coastal and surface seawater display little or no Ce*

and little HREE enrichment. Numerous attempts have been

made to interpret REE patterns in phosphorites as reflecting

depositional environment, specifically depositional oxygena-

tion levels. The very low content of REE in apatitic skeletal

elements of living organisms (e.g., fish bones) compared to

their higher concentration in fossils suggests, however, that at

least a large part of the REE is introduced into francolite during

early diagenesis.

Uranium. Phosphorite typically contains between 50 and

200 ppm U, though numerous deposits (Proterozoic-Cambrian

of Karatau and Khubsugul) contain as little as 2 ppm whereas

others (Eocene of the Central African Republic) may reach

levels of 5,600 ppm. Uranium in phosphorites occurs both as

U(IV) and U(VI), both apparently substituting for Ca in franco-

lite. U(IV) content varies between zero and almost 90%. That

variation in conjunction with petrography and other geochem-

ical parameters has been successfully used for the reconstruc-

tion of marine paleo-redox conditions.

Uranium and uranium series daughter isotopes (specificall y

234

238

230

Th/

234

231

Pa/

235

U) have been successfully used

as dating clocks of young phosphogenesis in the ocean (Baturin,

1981;Jarvisetal.,1994; Kolodny and Kaplan, 1971; see also

summary in Kolodny and Luz, 1992). Dating was instrumental in

demonstrating that: (a) most sites of seafloor phosphorite nodule

occurrence are not sites of present-day phosphogenesis, and (b)

phosphorites have formed recently off the coasts of Peru-Chile,

Namibia, western South Africa, and East Australia. Applying

these techniques, the episodic nature of phosphogenesis –

confined to periods of high eustatic sea level – has been

demonstrated for the region off Peru-Chile and western

South Africa.

Sr and Nd isotopes (Jarvis et al., 1994; Kolodny and Luz,

1992). Since the 1970s, it has been demonstrated that the

ratio of

Sr/

Sr in the world ocean can be described by

the

Sr/

Sr reference curve (see Strontium isotopes). The

high Sr content of francolite makes it potentially possible (within

reasonable assumptions) to date phosphorite formation by

measuring this ratio and comparing it to the reference curve.

Thus, episodes of Neogene phosphogenesis, grain concentra-

tion, and reworking of phosphatic sediments have been estab-

lished for the phosphorites of Florida and the North Carolina

shelf by Glenn et al. (1994), as well as by P. Stille and his

coworkers.

Unlike Sr, the much shorter residence time of Nd (see

Dating, radiometric methods) in the ocean (about 2 kyr) results

in a heterogeneity of the

143

Nd/

144

Nd ratio in seawater: each

major ocean yields a different Nd-isotope evolution curve.

Measurements of eNd in sedimentary apatites showed that the

present day Pacific-Atlantic provincialism is about 50 Ma

old. Prior to that, a single Panthalassa Ocean seems to have

existed for about 400 Ma. Prior to 450 Ma, two separate water

masses can again be distinguished – the South and North Iapetus

Oceans.

Figure P59 Scheme of possible substitutions and of the isotopic tracers

that are commonly analyzed in francolite. Note that whereas the light

stable isotopes (of C, O, S) are major anions in CFA, the heavy isotopes (of

Sr, Nd and U) substitute for Ca in the lattice (after Kolodny and Luz, 1992).

PHOSPHORITE 777

Stable isotopes (Jarvis et al., 1994; Kolodny and

Luz; 1992)

Significant information on the nature of phosphogenesis has

been acquired from the study of the stable isotopes of oxygen,

carbon, and sulfur.

Oxygen. Francolite has the unique property of containing

oxygen in two distinct sites, which are rather simply analyti-

cally separated. These are oxygen in the carbonate – d

and oxygen in phosphate – d

(Figure P59; see Stable iso-

tope analysis for definitions of the d notation used here). It

is assumed that oxygen (and carbon) isotope fractionation

in the carbonate follows the carbonate paleotemperature

equation (Epstein et al., 1953, see Paleotemperatures, proxy

reconstructions)

Tð



CÞ¼16:5  4:3ðd

 d

Þþ0:14ðd

 d

ð3Þ

and d

is governed by (Kolodny and Luz, 1992; Longinelli

and Nuti, 1973)

Tð



CÞ¼111:4  4:3ðd

 d

Þð4Þ

In Equation (3), d

is for all practical purposes equal to d

of the CO

extracted with phosphoric acid and measured vs.

PDB, d

in Equation (4) is the d

O value (vs. SMOW) of

the oxygen of the francolite phosphate, and d

in both equa-

tions is the d

O value of the water with which the francolite

equilibrated. T is the temperature of formation of the francolite.

It was initially hoped that d

might be a sensitive recorder

of environmental temperature and water composition, and a

robust preserver of such records. It turned out, however, that

although d

is more susceptible to alteration than d

both oxygen sites are prone to post depositional exchange.

Indeed d

and d

in phosphorites that span the time

range of the last 10

years are very well correlated. d

and

in phosphorites vary with time in a roughly parallel pat-

tern, both decreasing with increasing age, the greatest changes

occurring within samples of the last 100 Myr. A similar trend

has been observed before for several other sedimentary phases:

carbonates, cherts, and glauconites. The question whether these

trends record changing temperature, change in d

O of sea-

water, or increasing alteration has not yet been resolved. Sev-

eral attempts were made to use d

O in phosphorites for

paleotemperature reconstruction: Karhu and Epstein, 1986

(see also Kolodny and Luz, 1992) deduced high pre-Cambrian

temperatures from the phosphate-chert oxygen isotopic fractio-

nation. Shemesh and Kolodny (1988) concluded from the stra-

tigraphic variation in d

of Tethyan phosphorites that the

phosphogenic event was associated with a cool-water upwel-

ling. In many phosphorites (e.g., Miocene Monterey Formation

of California, Cambrian Georgina basin in Australia), d

values are much too low to represent marine temperatures.

Thus, the interpretation of such results is still problematic.

Carbon. The carbonate ion in francolite (CFA) can also

be analyzed for d

C. Such measurements lead to the conclu-

sion that not all carbonate enters the CFA lattice at the time

of precipitation; usually such carbonate is in equilibrium with

interstitial water rather than with seawater. McArthur et al.

(1986) have demonstrated that organic matter-derived (low

C) carbon was to some degree involved in the diagenesis

of francolites. d

C is negatively correlated with the CO

content of CFA. These studies led to the realization that

most phosphorites formed in dominantly suboxic conditions;

some (e.g., the Blake Plateau) were formed in a clearly oxic

environment, others (Morocco) have been in contact with

anoxic pore waters.

Sulfur. The sulfate ion substitutes for phosphate in CFA.

S varies in phosphorites between values equal to those of

coeval seawater sulfate and much higher values (by as much

as 10%). The former apparently reflects seawater sulfate incor-

poration, whereas the heavier sulfur reflects the residual sulfate

after partial reduction. Such “Rayleigh distillation” in a closed

system results in a negative correlation between SO

2–

and

S, as observed in francolites of the Phosphoria Formation

(Piper and Kolodny, 1987).

Origin and paleoclimatology

Three closely related observations influence strongly all thinking

on the genesis of phosphorites: (a) Though phosphorites were

deposited in a variety of settings, on seamounts, continental mar-

gins, and epeiric seas, all were related to high organic productiv-

ity and nearly all to upwelling. This is evidenced by the very

nature of phosphorus as a biolimiting element and by the ocean

being the major phosphorus reservoir, as well as by the abun-

dance of phosphatic skeletal fragments in phosphorite deposits.

(b) Though most phosphorites on the seafloor were proven to

be relict (see Kolodny and Luz, 1992), all modern phosphori-

tes are found in areas of oceanic upwelling. (c) The strong asso-

ciation of phosphorites with chert and black shales (P-Si-C) in

the geological record further points to such an association. In a

series of studies, J. Parrish and coworkers (see Parrish, 1998

for summary and references) concluded that “the distribution

of phosphorites is mostly simply dependent on the distribution of

upwelling zones over continental shelves for much of the

Phanerozoic.” She then used phosphorite distribution as a tool

for mapping upwelling zones of the past. Cook and McElhinny

(1979) stressed the low paleolatitude distribution of ancient

phosphorites and their association with evaporates.

Thus, the explanation of the formation of phosphorite

deposits amounts to pointing to a suggested break (sink) in

the chain: river input ! dissolved phosphate ! organic matter

formation ! dissolved phosphate. In about two centuries of

research, almost all possible breaks in the chain were offered

in addition to suggestions of additional sources such as volcan-

ism (see Bentor, 1980; Cook et al., 1990; Glenn et al., 1994;

Kolodny, 1981). Of seminal importance was the publication

of Kazakov’s paper in 1937, in which he linked phosphogen-

esis to oceanic upwelling, warming of the rising cold water,

release, and francolite precipitation. Kazakov’s idea,

though ultimately rejected in its original form, became the basis

of most variants of modern theories of phosphogenesis. Most

researchers today agree that francolite is primarily precipitated

interstitially in sediments, either chemically or biochemically.

In many cases, an association between microbial activity with

both mobilization and precipitation of CFA has been documen-

ted. Apatitic microbial structures were identified in phosphorite

deposits (e.g., Krajewski et al., 1994; Soudry and Champetier,

1983). It is not clear whether the role of microbial mats in

phosphogenesis was active, passive, or in-between these

extremes. Apatite was also formed bacterially in the laboratory

by J. Lucas and L. Prevot (see also Krajewski et al., 1994).

Although seawater is oversaturated with respect to CFA, the

slow kinetics of its nucleation and crystallization at the sedi-

ment water interface cause the principal formation site to be a

778 PHOSPHORITE

few centimeters inside the sediment, and associated with sur-

faces of non-deposition or slow sedimentation rates. Whatever

the process of primary francolite formation is, the achievement

of a high-grade phosphorite requires an additional step of con-

centration, most likely by mechanical re-working and winnow-

ing (Baturin, 1981). It has been stressed by Föllmi (1994) that

no exceptionally long times were required to form even the

very large phosphorite deposits. Thus, he estimates that assum-

ing an efficient P supply, only about 12 kyr were required

to form the southern Tethyan Paleocene-Eocene province,

and only 240 kyr of similar conditions for the formation of

the Permian Phosphoria deposit in the Western US. Filippelli

and Delaney (1992) showed that phosphorus accumulation

and burial rates of major phosphorite deposits are comparable

to those of the modern Peru margin.

Following the different models of phosphogenesis, various

authors chose to interpret the occurrence of “phosphate giants”

in the sedimentary record. These were cited as evidence

for times of warm climate (Fisher and Arthur, 1977), as well

as times of glaciation (Sheldon, 1980), times when continents

drifted into low latitudes (Cook and McElhinny, 1979), major

transgressions leading to major anoxia (Arthur and Jenkyns,

1981), or increased continental weathering. The broad spec-

trum of causes cited above probably shows that no single cause

could be regarded as the cause for phosphogenesis. Upwelling

may be an exception, it seemingly being related to most large

phosphorite deposits.

Economics

Phosphorites account for about 85% of the world’s phosphate

consumption (the rest is supplied from igneous and guano ores).

Mineral fertilizers account for approximately 80% of phosphate

use, with the balance divided between detergents (12%), animal

feeds (5%) and specialty applications (3%), e.g. food grade,

metal treatment etc. The production of phosphate rock peaked

in 1988 at a level of 166 million tons product, falling to 125 mil-

lion tons in 2001, and rebouncing to 147 million tons in 2007.

Over 30 countries are currently producing phosphate rock for

use in domestic markets and/or international trade. However,

the world’s top 12 producing countries account for nearly 95%

of the world’s total phosphate production. The main producers

are the USA, China, Morocco Russia and the Middle East.

In 2002 the US Geological Survey estimated that world

phosphate rock reserves amounted to about 12 billion tons,

with a larger reserve base of about 47 billion tons. Of these

reserves the lion’s share is concentrated in Morocco. The last

two decades of the twentieth century have seen a shrp decrease

in the demand for phoosphate, primarily due to an economic

slowdown in developing countries. In 2008 the price of phos-

phate rock reached $200/ton after several years of a stable price

of $50/ton. The expected growth is based on the demand for

more food by a growing and more affluent population, espe-

cially in the developing countries. It also is a reflection of the

fact that there is no substitute for phosphorus in agriculture.

Yehoshua Kolodny

Bibliography

Arthur, M.A., and Jenkyns, H.C., 1981. Phosphorites and paleoceanogra-

phy. Ocenologica Acta, Proceedings of 26th International Geological

Congress, Geology of Oceans Symposium, Paris, 1981, pp. 83–96.

Baturin, G.N. (1981). Phosphorites on the Sea Floor. Developments in Sedi-

mentology, 33. Amsterdam: Elsevier, 343pp.

Bentor, Y.K. (ed.), 1980. Phosphorites – the unsolved problems. Tulsa:

Society Economic Paleontology Mineralogy (SEPM) Special Publica-

tion No. 29, pp. 3–18.

Burnett, W.C., and Riggs, S.R. (eds.), 1990. Phosphate Deposits of

the World, Vol. 3: Neogene to Modern Phosphorites, Cambridge:

Cambridge University Press, 464pp.

Cook, P.J., 1976. Sedimentary phosphate deposits. In Wolf, K.H. (ed.),

Handbook of Strata-Bound and Stratiform Ore Deposits, Vol. 7,

Amsterdam: Elsevier, pp. 505–535.

Cook, P.J., and McElhinny, M.W., 1979. A reevaluation of the spatial and

temporal distribution of sedimentary phosphate deposits in the light of

plate tectonics. Econ. Geol., 74, 315–330.

Cook, P.J., and Shergold, J.H. (eds.), 1986. Phosphate Deposits of the

World, Vol. 1: Proterozoic and Cambrian Phosphorites. Cambridge:

Cambridge University Press. 386pp.

Cook, P.J., Shergold, J.H., Burnett, W.C., and Riggs, S.R., 1990. Phosphor-

ite research: A historical overview. In Notholt, A.J.G., and Jarvis, I.

(eds.), Phosphorite Research and Development. London: Geological

Society, Special Publication 52, pp. 1–22.

Epstein, S., Buchsbaum, R., Lowenstam, H.A., and Urey, H.C., 1953.

Revised carbonate-water isotopic temperature scale. Bull. Geol. Soc.

Am., 64(11), 1315–1326.

Filippelli, G.M., and Delaney, M.L., 1992. Similar phosphorus fluxes in

ancient phosphorite deposits and a modern phosphogenic environment.

Geology, 20(8), 709–712.

Fisher, A.G., and Arthur, M.A. 1977. Secular variations in the pelagic

realm. In: Cook, H.E and Enos, P. (eds.), Deep-Water Carbonate

Environments. Tulsa: Society Economic Paleontology Mineral (SEPM),

Special Publication No. 25, pp. 19–50.

Föllmi, K.B. (ed.), 1994. Concepts and controversies in phosphogenesis.

Eclogae Geol. Helv., 87(3): 639–641.

Föllmi, K.B., 1996. The phosphorus cycle, phosphogenesis and marine

phosphate rich deposits. Earth Sci. Rev., 40,55–124.

Glenn, C.R., Follmi, K.B., Riggs, S.R., Baturin, G.N., Grimm, K.A., Trappe,

J., Abed, A.M., Galliolivier, C., Garrison, R.E., Ilyin, A.V., Jehl, C.,

Rohrlich, V., Sadaqah, R.M.Y., Schidlowski, M., Sheldon, R.E., and

Siegmund, H., 1994. Phosphorus and phosphorites – Sedimentology

and environments of formation. Eclogae Geol. Helv., 87(3), 747

–788.

Jarvis, I., Burnett, W.C., Nathan, Y., Almbaydin, F.S.M., Attia, A.K.M.,

Castro, L.N., Flicoteaux, R., Hilmy, M.E., Husain, V., Qutawnah, A.A.,

Serjani, A., and Zanin, Y.N., 1994. Phosphorite geochemistry –

state-of-the-art and environmental concerns. Eclogae Geol. Helv.,

87(3), 643–700.

Karhu, J., and Epstein, S., 1986. The implication of the oxygen isotope

records in coexisting cherts and phosphates. Geochimica et Cosmochi-

mica Acta, 50, 1745–1756.

Kolodny, Y., 1981. Phosphorites (Chapter 11.3). In Emiliani, C. (ed.), The

Sea, vol. 7. New York: Wiley, pp. 981–1023.

Kolodny, Y., and Kaplan, I.R., 1971. C and O isotopes in apatite CO

and

coexisting calcite from sedimentary phosphorite. J. Sed. Pet., 3,

954–959.

Kolodny, Y., and Luz, B., 1992. Phosphate deposits, formation and diage-

netic history. In Clauer N., and Chaudhuri, S. (eds.), Isotopic Signatures

and Sedimentary Records. New York: Springer, 69–122.

Krajewski, K.P., Vancappellen, P., Trichet, J., Kuhn, O., Lucas, J.,

Martinalgarra, A., Prevot, L., Tewari, V.C., Gaspar, L., Knight, R.I.,

and Lamboy, M., 1994. Biological processes and apatite formation in

sedimentary environments. Eclogae Geol. Helv., 87(3), 701–745.

Longinelli, A., and Nuti, S., 1973. Revised phosphate-water isotopic tem-

perature scale. Earth Planet. Sci. Lett., 20, 337–340.

McArthur, J.M., Benmore, R.A., Coleman, M.L., Soldi, C., Yeh, H.W., and

Obrien, G.W., 1986. Stable isotopic characterization of francolite for-

mation. Earth Planet. Sci. Lett., 77(1), 20–34.

Nathan, Y., 1984. The mineralogy and geochemistry of phosphorites. In

Nriagu, J.O., and More, P.B. (eds.), Phosphate Minerals, Berlin:

Springer, 275–291.

Notholt, A.J.G., Sheldon, R.P., and Davidson, D.F. (eds.), 1989. Phosphate

Deposits of the World, Vol. 2: Phosphate Rock Resources. Cambridge:

Cambridge University Press, 566pp.

Parrish, J.T., 1998. Interpreting Pre-Quaternary Climate from the Geologic

Record. New York: Columbia University Press, 338pp.

Piper, D.Z., and Kolodny, Y., 1987. The stable isotopic composition

of a phosphorite Deposit, d

C, d

S and d

O. Deep Sea Res., 34,

897–911.

PHOSPHORITE 779

Sheldon, R.P., 1980. Ancient marine phosphorites. Annu. Rev. Earth

Planet. Sci., 9, 251–284.

Shemesh, A., and Kolodny, Y., 1988. Oxygen isotope variations in

phosphorites from the southeastern Tethys. Isr. J. Earth Sci., 37,1–15.

Soudry, D., and Champetier, Y., 1983. Microbial processes in the Negev

phosphorites (southern Israel). Sedimentology, 30,411–423.

Cross-references

Carbon Isotopes, Stable

Dating, Radiometric Methods

Geochemical Proxies (Non-Isotopic)

Mineral Indicators of Past Climates

Oxygen Isotopes

Paleotemperatures and Proxy Reconstructions

Phosphorus Cycle

Stable Isotope Analysis

Strontium Isotopes

Sulfur Isotopes

Uranium-Series Dating

PHOSPHORUS CYCLE

Introduction

Phosphorus (P) is a limiting nutrient for terrestrial biological

productivity, and thus it commonly plays a key role in net car-

bon uptake in terrestrial ecosystems. Unlike nitrogen (another

limiting nutrient but one with an abundant atmospheric pool),

the availability of “new” P in ecosystems is restricted by its rate

of release during soil weathering. The release of P places a

limit on ecosystem productivity (Schlesinger, 1997), which in

turn is critical to terrestrial carbon balances. Furthermore, the

weathering of P and transport by rivers is the only appreciable

source of P to the oceans. On longer time scales, this supply of

P limits total primary production in the ocean (Smith, 1984).

Thus, understanding the controls on P cycling is one key ele-

ment of elucidating biogeochemical responses to and forcing

of paleoclimate.

Natural Phosphorus Cycle

The human impact on the global P cycle has been substantial

over the last 150 years. Because this anthropogenic modifica-

tion began well before scientific efforts to quantify the cycle

of P, we can only guess at the “pre-anthropogenic” mass bal-

ance of P. Several aspects of the P cycle are well-constrained

(Figure P60). Phosphorus is initially solubilized, mainly from

apatite minerals, by chemical weathering during soil develop-

ment. Physical weathering also plays a role by producing fine

materials with extremely high surface area/mass ratios, which

enhances chemical weathering in continental environments

(i.e., floodplains, delta systems).

Phosphorus cycling in soils

The cycling of P in soils has received much attention, in terms of

both fertilization and the natural development of ecosystems. Of

the approximately 122,600 Tg P within the soil/biota system on

the continents, nearly 98% is held in soils in a variety of

forms. The exchange of P between biota and soils is relatively

rapid, with an average residence time of 13 years, whereas the

average residence time of P in soils is 600 years (Figure P60).

The most significant weathering source for phosphorus in

soils are apatite minerals. These minerals can be congruently

weathered as a result of reaction with dissolved carbon dioxide:

ðPO

OH þ 4CO

þ3H

O ! 5Ca

2þ

þ3HPO

2

þ4HCO



In soils, P is released from mineral grains by several pro-

cesses. First, the reduced pH produced from respiration-related

in the vicinity of both degrading organic matter and root

hairs dissolves P-bearing minerals (mainly apatites) and re-

leases P to root pore spaces. Second, organic acids released

by plant roots also can dissolve apatite minerals and release P

to soil pore spaces (Jurinak et al., 1986). Phosphorus is very

immobile in most soils, and its slow rate of diffusion from dis-

solved form in pore spaces strongly limits its supply to rootlet

surfaces. Furthermore, much of the available P in soils is in

organic matter, which is not directly accessible for plant nutri-

tion. Plants have developed two specific tactics to increase the

supply of P to roots. Phosphatase, an enzyme that can release

bioavailable inorganic P from organic matter, is often excreted

Figure P60 The natural (pre-human) phosphorus cycle, showing reservoirs (in Tg P, where Tg = 10

g) and fluxes (denoted by arrows,inTgPyr

–1

)

in the P mass balance (after Filippelli, 2002).

780 PHOSPHORUS CYCLE