18. Landscape and Regional Biogeochemistry: Approaches
281
ecologists to assess large-scale movement of nutrients using the natural abundance of stable isotopes
was an assessment of the source of nitrate (fertilizer
vs. soil organic matter) in surface water of a cultivated watershed (Kohl et al. 1971). The work was
met with a great deal of criticism regarding the usefulness of the technique in part because relatively
few samples can be economically analyzed, even
though the data may represent a very large watershed. Nonetheless, a great deal of recent work has
elaborated on the techniques, better establishing
variance estimates for the calculations (Shearer and
Kohl 1993), and providing a better basis for understanding the strengths and limitations of the approach. Evaluation of ol3C within organisms has
provided important information on carbon (C)
sources and fluxes in aquatic ecosystems (Peterson
and Fry 1987; Fry 1991; Michener and Schnell
1994). Multiple isotope (e.g., C, sulfur [S], and nitrogen [N]) methods significantly improve upon our
ability to assess and quantify sources and sinks
(e.g., Fry 1991; Michener and Schell 1994).
Finally, a third approach to studying the twodimensional redistribution of elements within landscapes and regions is using tracers. I distinguish
this from the natural abundance methods above as
being an approach in which the investigator releases an isotope or element into the environment,
and then tracks its distribution at the landscape
scale in space and time. Again, aquatic biogeochemists have taken the lead in this area. Two types
of tracers are utilized, biologically active isotopes,
and nonactive substitutes. Hutchinson and Bowen
(1947) may have been the first to apply such a technique, in a tracer study of phosphorus-32 e 2 p), in
which they studied the sedimentation of 32p from
the epilimnion into the hypolimnion of a lake. For
example, in a current large-scale comparison of N
processing in streams, 15N is released and tracked
downstream in biological and inorganic pools to
assess the biological cycling and spatial redistribution (Holmes, personal communication).
One approach that has proved useful is the analysis of soil movement using cesium-137 (137CS).
Cesium-137 is not technically a tracer, in that the
investigator does not apply it, however humans distributed it relatively uniformly over the landscape
during the bomb tests of the 1950s. Because Cs is
a cation that binds to cation exchange sites on clays
(Tamura 1960) and is relatively inert (Schultz et al.
1960), its current distribution is thought to reflect
the net redistribution of soils on a landscape since
the 1950s (Ritchie and McHenry 1978; Martz and
de Jong 1987; Schlesinger et al. 1989; Coppinger
et al. 1991). A limitation of this method is that actual estimates of rates of soil movement are not
possible, only indices of net movement of soil particles by wind or water.
Otherwise, for terrestrial systems, use of tracers
for the two-dimensional movement of elements
across landscapes does not have a rich literature.
There is likely to be a great deal that can be contributed in methodological development in this
area.
Modeling Movement
How are investigators modeling the movement of
materials and energy across landscapes and
regions? Clearly, the distribution of organic matter,
inorganic nutrients, toxic materials, and sediments
across landscapes and regions is largely the consequences of lateral and vertical movement of those
materials. Models that truly incorporate spatial processes represent a very important advance in our
ability to test our understanding of how systems
function at large scales. However, just as field analysis of movement in ecosystem ecology has been
somewhat limited, so has simulation analysis of
movement (Costanza et al. 1990; Kareiva and Wennergren 1995). Many of the past efforts that address
movement of materials across landscapes and
regions were conducted by hydrologists and soil
scientists studying erosion, and there is a significant
literature on these models (e.g., Favis-Mortlock et
al. 1996; Kirkby et al. 1996). The approach is to
connect a spatial database to a model, as described
above, but the model includes spatial processes in
which materials or biota flow among individual
cells. Rules and probabilities for movement may be
governed by such features as topographic location
(slope angle and aspect, location with respect to
wind direction and force), vegetation type, soil
type, and stochastic processes (e.g., fire) (Wallin et
al. 1996). Transitions in states that influence the
source/sink potential of a particular cell or group of
cells may be implemented using a cellular automata
approach (e.g., Hi and Reynolds 1997), a simple set
of transition rules that can generate relatively complex patterns.
281
ecologists to assess large-scale movement of nutrients using the natural abundance of stable isotopes
was an assessment of the source of nitrate (fertilizer
vs. soil organic matter) in surface water of a cultivated watershed (Kohl et al. 1971). The work was
met with a great deal of criticism regarding the usefulness of the technique in part because relatively
few samples can be economically analyzed, even
though the data may represent a very large watershed. Nonetheless, a great deal of recent work has
elaborated on the techniques, better establishing
variance estimates for the calculations (Shearer and
Kohl 1993), and providing a better basis for understanding the strengths and limitations of the approach. Evaluation of ol3C within organisms has
provided important information on carbon (C)
sources and fluxes in aquatic ecosystems (Peterson
and Fry 1987; Fry 1991; Michener and Schnell
1994). Multiple isotope (e.g., C, sulfur [S], and nitrogen [N]) methods significantly improve upon our
ability to assess and quantify sources and sinks
(e.g., Fry 1991; Michener and Schell 1994).
Finally, a third approach to studying the twodimensional redistribution of elements within landscapes and regions is using tracers. I distinguish
this from the natural abundance methods above as
being an approach in which the investigator releases an isotope or element into the environment,
and then tracks its distribution at the landscape
scale in space and time. Again, aquatic biogeochemists have taken the lead in this area. Two types
of tracers are utilized, biologically active isotopes,
and nonactive substitutes. Hutchinson and Bowen
(1947) may have been the first to apply such a technique, in a tracer study of phosphorus-32 e 2 p), in
which they studied the sedimentation of 32p from
the epilimnion into the hypolimnion of a lake. For
example, in a current large-scale comparison of N
processing in streams, 15N is released and tracked
downstream in biological and inorganic pools to
assess the biological cycling and spatial redistribution (Holmes, personal communication).
One approach that has proved useful is the analysis of soil movement using cesium-137 (137CS).
Cesium-137 is not technically a tracer, in that the
investigator does not apply it, however humans distributed it relatively uniformly over the landscape
during the bomb tests of the 1950s. Because Cs is
a cation that binds to cation exchange sites on clays
(Tamura 1960) and is relatively inert (Schultz et al.
1960), its current distribution is thought to reflect
the net redistribution of soils on a landscape since
the 1950s (Ritchie and McHenry 1978; Martz and
de Jong 1987; Schlesinger et al. 1989; Coppinger
et al. 1991). A limitation of this method is that actual estimates of rates of soil movement are not
possible, only indices of net movement of soil particles by wind or water.
Otherwise, for terrestrial systems, use of tracers
for the two-dimensional movement of elements
across landscapes does not have a rich literature.
There is likely to be a great deal that can be contributed in methodological development in this
area.
Modeling Movement
How are investigators modeling the movement of
materials and energy across landscapes and
regions? Clearly, the distribution of organic matter,
inorganic nutrients, toxic materials, and sediments
across landscapes and regions is largely the consequences of lateral and vertical movement of those
materials. Models that truly incorporate spatial processes represent a very important advance in our
ability to test our understanding of how systems
function at large scales. However, just as field analysis of movement in ecosystem ecology has been
somewhat limited, so has simulation analysis of
movement (Costanza et al. 1990; Kareiva and Wennergren 1995). Many of the past efforts that address
movement of materials across landscapes and
regions were conducted by hydrologists and soil
scientists studying erosion, and there is a significant
literature on these models (e.g., Favis-Mortlock et
al. 1996; Kirkby et al. 1996). The approach is to
connect a spatial database to a model, as described
above, but the model includes spatial processes in
which materials or biota flow among individual
cells. Rules and probabilities for movement may be
governed by such features as topographic location
(slope angle and aspect, location with respect to
wind direction and force), vegetation type, soil
type, and stochastic processes (e.g., fire) (Wallin et
al. 1996). Transitions in states that influence the
source/sink potential of a particular cell or group of
cells may be implemented using a cellular automata
approach (e.g., Hi and Reynolds 1997), a simple set
of transition rules that can generate relatively complex patterns.