19. Nutrient Manipulations in Terrestrial Ecosystems
the community shifts are a critical factor determining the response of the overall ecosystem net primary productivity (NPP) and biogeochemical processes to nutrient additions (Bowman et al. 1993).
Thus, for questions addressing the overall ecosystem capacity to fix C, the nutrient saturation point
of an ecosystem, or the overall response of a system
to nutrient deposition-the duration of the experiment should be long enough to incorporate these
community shifts, which can be the most significant
mechanism driving overall ecosystem response to
nutrient additions.
Commonly Limiting Nutrients
Which nutrients are limiting can often be predicted
based on the age of the system, climate (especially
how weathered or leached the soil is), soil texture,
parent material, and vegetation type.
Most nutrient limitation experiments focus on nitrogen (N), phosphorus (P), and potassium (K). Nitrogen is the most commonly limiting nutrient to
plant growth in temperate terrestrial systems (Vitousek and Howarth 1991). Phosphorus tends to be
limiting in the lowland wet tropics (Tanner et al.
1998), on very old soils (Vitousek and Farrington
1997), and in legume-rich systems (Jones et al.
1983,1990). Potassium tends to be limiting in areas
of high precipitation, or very late in soil development (Tisdale et al. 1993). While N, P and K may
be the most commonly limiting nutrients, it is not
uncommon to find productivity limited by other nutrients such as calcium (Ca), magnesium (Mg), and
sulfur (S). For example, California grasslands are
often assumed to be N-limited, but many sites respond to P and S fertilization (Jones et al. 1983).
Nitrogen, P, and K fertilizers often contain these
other nutrients, so that some responses to these fertilizers may partially reflect responses to Ca, Mg,
and S. This is especially true for legume-rich sites,
which are often limited by nutrients such as Sand
P (Jones and Martin 1964; Jones et al. 1970).
Single Versus Multiple
Nutrient Limitation
The simplest view of environmental limitation is
that growth is limited by a single resource at any
one time; another resource becomes limiting only
when the supply of the first resource is increased
293
above the point of limitation (Liebig's law of the
minimum). Several processes contribute to the multiple resource limitation observed in most ecosystems: (1) Plants adjust both root/shoot allocation
and nutrient uptake capacity to maximize capture
of (and minimize limitation by) the most strongly
limiting resource (Chapin et al. 1987; Rastetter and
Shaver 1992). (2) Changes in environment (e.g.,
rainstorms, pulses of nutrient supply) change the
relative abundance of resources so that different
factors limit NPP at different times. (3) Different
species in an ecosystem are limited by different resources, so that ecosystem-scale NPP responds to
addition of more than one resource. Each of these
processes contributes to the response of ecosystems
to multiple resources.
Experimental Design
General Approach
The design of nutrient-addition experiments must
be tailored to the question addressed. A single application of nutrients at high addition rates (e.g., 10
to 20 g N or P m - 2) is the simplest test of whether
production of the ecosystem (or its dominant species) is limited by a particular nutrient at the time
of nutrient application. The lack of response to low
nutrient-addition rates could reflect lack of limitation by the nutrient applied or failure of the plants
to gain access to the added nutrient due to chemical
fixation (which frequently happens with phosphorus) or biological immobilization by soil microbes
or vegetation on the ground surface (Chapin et al.
1986).
Experiments intended to examine the sensitivity
of ecosystems to atmospheric deposition should use
application rates that approximate natural deposition, which can range from considerably less that 1
g m -2 yr- 1 to greater than 10 g m -2 yr- 1 in industrialized areas. Low deposition rates typically
occur on the west coasts of continents and other
places remote from pollution sources. These additions should be added repeatedly to determine longterm response. One problem with these experiments
is that control plots inevitably receive ambient deposition, so one can only test the impact of atmospheric deposition that is higher than ambient
levels.
the community shifts are a critical factor determining the response of the overall ecosystem net primary productivity (NPP) and biogeochemical processes to nutrient additions (Bowman et al. 1993).
Thus, for questions addressing the overall ecosystem capacity to fix C, the nutrient saturation point
of an ecosystem, or the overall response of a system
to nutrient deposition-the duration of the experiment should be long enough to incorporate these
community shifts, which can be the most significant
mechanism driving overall ecosystem response to
nutrient additions.
Commonly Limiting Nutrients
Which nutrients are limiting can often be predicted
based on the age of the system, climate (especially
how weathered or leached the soil is), soil texture,
parent material, and vegetation type.
Most nutrient limitation experiments focus on nitrogen (N), phosphorus (P), and potassium (K). Nitrogen is the most commonly limiting nutrient to
plant growth in temperate terrestrial systems (Vitousek and Howarth 1991). Phosphorus tends to be
limiting in the lowland wet tropics (Tanner et al.
1998), on very old soils (Vitousek and Farrington
1997), and in legume-rich systems (Jones et al.
1983,1990). Potassium tends to be limiting in areas
of high precipitation, or very late in soil development (Tisdale et al. 1993). While N, P and K may
be the most commonly limiting nutrients, it is not
uncommon to find productivity limited by other nutrients such as calcium (Ca), magnesium (Mg), and
sulfur (S). For example, California grasslands are
often assumed to be N-limited, but many sites respond to P and S fertilization (Jones et al. 1983).
Nitrogen, P, and K fertilizers often contain these
other nutrients, so that some responses to these fertilizers may partially reflect responses to Ca, Mg,
and S. This is especially true for legume-rich sites,
which are often limited by nutrients such as Sand
P (Jones and Martin 1964; Jones et al. 1970).
Single Versus Multiple
Nutrient Limitation
The simplest view of environmental limitation is
that growth is limited by a single resource at any
one time; another resource becomes limiting only
when the supply of the first resource is increased
293
above the point of limitation (Liebig's law of the
minimum). Several processes contribute to the multiple resource limitation observed in most ecosystems: (1) Plants adjust both root/shoot allocation
and nutrient uptake capacity to maximize capture
of (and minimize limitation by) the most strongly
limiting resource (Chapin et al. 1987; Rastetter and
Shaver 1992). (2) Changes in environment (e.g.,
rainstorms, pulses of nutrient supply) change the
relative abundance of resources so that different
factors limit NPP at different times. (3) Different
species in an ecosystem are limited by different resources, so that ecosystem-scale NPP responds to
addition of more than one resource. Each of these
processes contributes to the response of ecosystems
to multiple resources.
Experimental Design
General Approach
The design of nutrient-addition experiments must
be tailored to the question addressed. A single application of nutrients at high addition rates (e.g., 10
to 20 g N or P m - 2) is the simplest test of whether
production of the ecosystem (or its dominant species) is limited by a particular nutrient at the time
of nutrient application. The lack of response to low
nutrient-addition rates could reflect lack of limitation by the nutrient applied or failure of the plants
to gain access to the added nutrient due to chemical
fixation (which frequently happens with phosphorus) or biological immobilization by soil microbes
or vegetation on the ground surface (Chapin et al.
1986).
Experiments intended to examine the sensitivity
of ecosystems to atmospheric deposition should use
application rates that approximate natural deposition, which can range from considerably less that 1
g m -2 yr- 1 to greater than 10 g m -2 yr- 1 in industrialized areas. Low deposition rates typically
occur on the west coasts of continents and other
places remote from pollution sources. These additions should be added repeatedly to determine longterm response. One problem with these experiments
is that control plots inevitably receive ambient deposition, so one can only test the impact of atmospheric deposition that is higher than ambient
levels.