4. Atmospheric and Geologic Constraints
61
engelmanii) and subalpine fir (Abies lasiocarpa) on igneous quartz diorite
geology.
As a comparison to these coastal watersheds, we included data from
inland sites along a transect from British Columbia to the western part of
Alberta; Marmot Creek (MC) (51°N, 115°W; Water Survey of Canada,
1979), Tri-Creeks (TC) (53°N, 117°W; Jablonski , 1980), and Streeter
Basin (SB) (50
0N,
114°W; Singh & Kalra , 1976) are located on sandstones
and shales. Annual precipitation is relatively low (ca. 550-950 mm/yr) ,
and vegetation ranges from coniferous forests at MC (Engelmann spruce
and alpine fir) and TC (mainly lodgepole pine, Pinus contorta) to mixed
aspen (Populus tremuloides) and grassland at Streeter Creek. Data from
MC, TC, and SB are based on published growing-season averages.
Information on element losses from South American temperate rainforests are scarce to nonexistent. Our own data derive from 31 watersheds
at Cordillera de Piuchue (CP) (42°22'S, 74°03'W) and 8 watersheds at
Cordillera Pelada (CPE) (40°10'S, 73°35'W) collected during the summers
of 1989, 1992, and 1993 (Hedin & Campos, 1991; Hedin et al., 1995). In
both areas, watersheds were located along west-east transects from the
ocean to approximately 20km inland and from between 0 m to approximately 900m above sea level.
Procedures for sampling, preservation, and analysis are given in Hedin
et al., 1995. The two study areas have similar floristic composition , with
high-elevation conifer forests dominated by Alerce (Fitzroy a cuppressoides), mid-elevation mixed forests dominated by Nothofagus nitida
and Drimys winteri, and low-elevation coastal forests generally dominated
by Olivillo (Aextoxicon punctatum) and Luma apiculata (Armesto &
Fuentes, 1988; Hedin et al., 1995; Villagran, 1985). CP watersheds are
underlaid by strongly weathered metamorphic feldspar schists and have
not been glaciated (Holdgate, 1961 ; Watters & Fleming, 1972). Soils are
moist with infrequent gley horizon at approximately 40 em depth. While
CPE is of similar geological origin as CP, less is known about the
mineralogy, soils, extent of glaciation, and disturbance history at CPE .
Element Contributions from Atmospheric Versus
Geologic Sources
We use three different approaches to evaluate the relative importance
of atmospheric versus geologic sources of elements to watersheds, each
subject to different conceptual advantages and limitations: 1) the socalled Gibbs diagram ; 2) correlation analysis of Na+ versus Cl" ; and 3)
sea-salt correction of stream-water chemistries.
Gibbs (1970) suggested that the relative importance of major sources of
elements to natural waters can be evaluated by graphing the weight ratio
of Na+:(Na+ + Ca
2+)
against total dissolved solids (TDS) . Gibbs used
61
engelmanii) and subalpine fir (Abies lasiocarpa) on igneous quartz diorite
geology.
As a comparison to these coastal watersheds, we included data from
inland sites along a transect from British Columbia to the western part of
Alberta; Marmot Creek (MC) (51°N, 115°W; Water Survey of Canada,
1979), Tri-Creeks (TC) (53°N, 117°W; Jablonski , 1980), and Streeter
Basin (SB) (50
0N,
114°W; Singh & Kalra , 1976) are located on sandstones
and shales. Annual precipitation is relatively low (ca. 550-950 mm/yr) ,
and vegetation ranges from coniferous forests at MC (Engelmann spruce
and alpine fir) and TC (mainly lodgepole pine, Pinus contorta) to mixed
aspen (Populus tremuloides) and grassland at Streeter Creek. Data from
MC, TC, and SB are based on published growing-season averages.
Information on element losses from South American temperate rainforests are scarce to nonexistent. Our own data derive from 31 watersheds
at Cordillera de Piuchue (CP) (42°22'S, 74°03'W) and 8 watersheds at
Cordillera Pelada (CPE) (40°10'S, 73°35'W) collected during the summers
of 1989, 1992, and 1993 (Hedin & Campos, 1991; Hedin et al., 1995). In
both areas, watersheds were located along west-east transects from the
ocean to approximately 20km inland and from between 0 m to approximately 900m above sea level.
Procedures for sampling, preservation, and analysis are given in Hedin
et al., 1995. The two study areas have similar floristic composition , with
high-elevation conifer forests dominated by Alerce (Fitzroy a cuppressoides), mid-elevation mixed forests dominated by Nothofagus nitida
and Drimys winteri, and low-elevation coastal forests generally dominated
by Olivillo (Aextoxicon punctatum) and Luma apiculata (Armesto &
Fuentes, 1988; Hedin et al., 1995; Villagran, 1985). CP watersheds are
underlaid by strongly weathered metamorphic feldspar schists and have
not been glaciated (Holdgate, 1961 ; Watters & Fleming, 1972). Soils are
moist with infrequent gley horizon at approximately 40 em depth. While
CPE is of similar geological origin as CP, less is known about the
mineralogy, soils, extent of glaciation, and disturbance history at CPE .
Element Contributions from Atmospheric Versus
Geologic Sources
We use three different approaches to evaluate the relative importance
of atmospheric versus geologic sources of elements to watersheds, each
subject to different conceptual advantages and limitations: 1) the socalled Gibbs diagram ; 2) correlation analysis of Na+ versus Cl" ; and 3)
sea-salt correction of stream-water chemistries.
Gibbs (1970) suggested that the relative importance of major sources of
elements to natural waters can be evaluated by graphing the weight ratio
of Na+:(Na+ + Ca
2+)
against total dissolved solids (TDS) . Gibbs used