corresponding to different levels in the food chain –
the net age of phosphorus in any of these reservoirs
can be determined (Table 4). New insights into
phosphorus cycling in oceanic surface waters derived
from recent work using the cosmogenically produced
33
P/
32
P ratio include the following. (1) Turnover rates
of dissolved inorganic phosphorus in coastal and
oligotrophic oceanic surface waters ranges from 1 to
20 days. (2) Variable turnover rates in the DOP pool
range from o1 week to >100 days, suggesting differences in either the demand for DOP, or the lability
of DOP toward enzymatic breakdown. (3) In the Gulf
of Maine, DOP turnover times vary seasonally, increasing from 28 days in July to >100 days in August,
suggesting that the DOP pool may evolve compositionally during the growing seasons. (4) Comparison of the
33
P/
32
P ratio in different particulate
size classes indicates that the age of phosphorus
generally increases at successive levels in the food
chain. (5) Under some circumstances, the
33
P/
32
P
ratio can reveal which dissolved pool is being ingested by a particular size class of organisms. Utilization of this new tool highlights the dynamic nature
of phosphorus cycling in surface waters by revealing
the rapid rates and temporal variability of phosphorus turnover. It further stands to provide new
insights into ecosystem nutrient dynamics by revealing, for example, that low phosphorus concentrations can support high primary productivity
through rapid turnover rates, and that there is preferential utilization of particular dissolved phosphorus
pools by certain classes of organisms.
The Oceanic Residence Time of Phosphorus
As phosphorus is the most likely limiting nutrient on
geological timescales, an accurate determination of
its oceanic residence time is crucial to understanding
how levels of primary productivity may have varied
in the Earth’s past. Residence time provides a means
of evaluating how rapidly the oceanic phosphorus
inventory may have changed in response to variations in either input (e.g., continental weathering,
dust flux) or output (e.g., burial with sediments). In
its role as limiting nutrient, phosphorus will dictate
the amount of surface-ocean net primary productivity, and hence atmospheric CO 2 drawdown that
will occur by photosynthetic biomass production. It
has been suggested that this so-called ‘nutrient–CO 2 ’
connection might link the oceanic phosphorus cycle
to climate change due to reductions or increases in
the atmospheric greenhouse gas inventory. Oceanic
phosphorus residence time and response time (the
inverse of residence time) will dictate the timescales
over which such a phosphorus-induced climate effect
may operate.
Over the past decade there have been several reevaluations of the marine phosphorus cycle in the
literature, reflecting changes in our understanding of
the identities and magnitudes of important phosphorus sources and sinks. One quantitatively important newly identified marine phosphorus sink is
precipitation of disseminated authigenic carbonate
fluoroapatite (CFA) in sediments in nonupwelling
environments. CFA is the dominant phosphorus
mineral in economic phosphorite deposits. Its presence in terrigenous-dominated continental margin
environments is detectable only by indirect methods
(coupled pore water/solid-phase chemical analyses),
because dilution by terrigenous debris renders it
below detection limits of direct methods (e.g. X-ray
diffraction). Disseminated CFA has now been found
in numerous continental margin environments,
bearing out early proposals that this is an important
marine phosphorus sink. A second class of authigenic
phosphorus minerals identified as a quantitatively
significant phosphorus sink in sandy continental
margin sediments are aluminophosphates. Continental margins in general are quantitatively important
sinks for organic and ferric iron-bound phosphorus,
as well. When newly calculated phosphorus burial
fluxes in continental margins, including the newly
identified CFA and aluminophosphate (dominantly
aluminum rare-earth phosphate) sinks are combined
with older estimates of phosphorus burial fluxes in
the deep sea, the overall burial flux results in a much
shorter residence time than the canonical value of
100 000 years found in most text books (Table 5).
This reduced residence time suggests that the oceanic
phosphorus-cycle is subject to perturbations on
shorter timescales than has previously been believed.
The revised, larger burial flux cannot be balanced
by the dissolved riverine input alone. However, when
the fraction of riverine particulate phosphorus that is
believed to be released upon entering the marine
realm is taken into account, the possibility of a balance between inputs and outputs becomes more
feasible. Residence times estimated on the basis of
phosphorus inputs that include this ‘releasable’
riverine particulate phosphorus fall within the range
of residence time estimates derived from phosphorus
burial fluxes (Table 5). Despite the large uncertainties associated with these numbers, as evidenced by
the maximum and minimum values derived from
both input and removal fluxes, these updated residence times are all significantly shorter than the canonical value of 100 000 years. Revised residence
times on the order of 10 000–17 000 y make phosphorus-perturbations of the ocean–atmosphere CO 2
576 PHOSPHORUS CYCLE
the net age of phosphorus in any of these reservoirs
can be determined (Table 4). New insights into
phosphorus cycling in oceanic surface waters derived
from recent work using the cosmogenically produced
33
P/
32
P ratio include the following. (1) Turnover rates
of dissolved inorganic phosphorus in coastal and
oligotrophic oceanic surface waters ranges from 1 to
20 days. (2) Variable turnover rates in the DOP pool
range from o1 week to >100 days, suggesting differences in either the demand for DOP, or the lability
of DOP toward enzymatic breakdown. (3) In the Gulf
of Maine, DOP turnover times vary seasonally, increasing from 28 days in July to >100 days in August,
suggesting that the DOP pool may evolve compositionally during the growing seasons. (4) Comparison of the
33
P/
32
P ratio in different particulate
size classes indicates that the age of phosphorus
generally increases at successive levels in the food
chain. (5) Under some circumstances, the
33
P/
32
P
ratio can reveal which dissolved pool is being ingested by a particular size class of organisms. Utilization of this new tool highlights the dynamic nature
of phosphorus cycling in surface waters by revealing
the rapid rates and temporal variability of phosphorus turnover. It further stands to provide new
insights into ecosystem nutrient dynamics by revealing, for example, that low phosphorus concentrations can support high primary productivity
through rapid turnover rates, and that there is preferential utilization of particular dissolved phosphorus
pools by certain classes of organisms.
The Oceanic Residence Time of Phosphorus
As phosphorus is the most likely limiting nutrient on
geological timescales, an accurate determination of
its oceanic residence time is crucial to understanding
how levels of primary productivity may have varied
in the Earth’s past. Residence time provides a means
of evaluating how rapidly the oceanic phosphorus
inventory may have changed in response to variations in either input (e.g., continental weathering,
dust flux) or output (e.g., burial with sediments). In
its role as limiting nutrient, phosphorus will dictate
the amount of surface-ocean net primary productivity, and hence atmospheric CO 2 drawdown that
will occur by photosynthetic biomass production. It
has been suggested that this so-called ‘nutrient–CO 2 ’
connection might link the oceanic phosphorus cycle
to climate change due to reductions or increases in
the atmospheric greenhouse gas inventory. Oceanic
phosphorus residence time and response time (the
inverse of residence time) will dictate the timescales
over which such a phosphorus-induced climate effect
may operate.
Over the past decade there have been several reevaluations of the marine phosphorus cycle in the
literature, reflecting changes in our understanding of
the identities and magnitudes of important phosphorus sources and sinks. One quantitatively important newly identified marine phosphorus sink is
precipitation of disseminated authigenic carbonate
fluoroapatite (CFA) in sediments in nonupwelling
environments. CFA is the dominant phosphorus
mineral in economic phosphorite deposits. Its presence in terrigenous-dominated continental margin
environments is detectable only by indirect methods
(coupled pore water/solid-phase chemical analyses),
because dilution by terrigenous debris renders it
below detection limits of direct methods (e.g. X-ray
diffraction). Disseminated CFA has now been found
in numerous continental margin environments,
bearing out early proposals that this is an important
marine phosphorus sink. A second class of authigenic
phosphorus minerals identified as a quantitatively
significant phosphorus sink in sandy continental
margin sediments are aluminophosphates. Continental margins in general are quantitatively important
sinks for organic and ferric iron-bound phosphorus,
as well. When newly calculated phosphorus burial
fluxes in continental margins, including the newly
identified CFA and aluminophosphate (dominantly
aluminum rare-earth phosphate) sinks are combined
with older estimates of phosphorus burial fluxes in
the deep sea, the overall burial flux results in a much
shorter residence time than the canonical value of
100 000 years found in most text books (Table 5).
This reduced residence time suggests that the oceanic
phosphorus-cycle is subject to perturbations on
shorter timescales than has previously been believed.
The revised, larger burial flux cannot be balanced
by the dissolved riverine input alone. However, when
the fraction of riverine particulate phosphorus that is
believed to be released upon entering the marine
realm is taken into account, the possibility of a balance between inputs and outputs becomes more
feasible. Residence times estimated on the basis of
phosphorus inputs that include this ‘releasable’
riverine particulate phosphorus fall within the range
of residence time estimates derived from phosphorus
burial fluxes (Table 5). Despite the large uncertainties associated with these numbers, as evidenced by
the maximum and minimum values derived from
both input and removal fluxes, these updated residence times are all significantly shorter than the canonical value of 100 000 years. Revised residence
times on the order of 10 000–17 000 y make phosphorus-perturbations of the ocean–atmosphere CO 2
576 PHOSPHORUS CYCLE