Full text of DOI:10.1007/s10021-001-0073-5

Ecosystems (2002) 5: 300–314
DOI: 10.1007/s10021-001-0073-5
E
COSYSTEMS
© 2002 Springer-Verlag
Phosphorus Limitation in Boreal
Forests: Effects of Aluminum and
Iron Accumulation in the
Humus Layer
Reiner Giesler,* Tomas Petersson, and Peter Ho¨gberg
Soil Science Section, Department of Forest Ecology, Swedish University of Agricultural Sciences, 901 83 Umeå, Sweden
ABSTRACT
Plant growth in boreal forests is generally consid-
humus layer, indicating that organic P is associated
with Al and Fe compounds in the humus. A plant
bioassay using humus substrate from one ground-
water discharge area and the adjacent groundwater
recharge area found that plants grown in ground-
water discharge area humus (with a high P-fixation
capacity) increased their biomass upon P fertiliza-
tion, whereas no growth response was found for N
additions. By contrast, plants grown in humus from
the groundwater recharge area did not respond to
added P unless N was added too. This study suggests
that groundwater discharge can affect the nutrient
availability of N and P both directly, via increased P
fixation due to the redistribution of Al and Fe, and
indirectly, via the inflow of groundwater high in Ca
and alkalinity, maintaining a high pH in the humus
layer that favors in situ N turnover processes.
ered to be predominantly nitrogen (N) limited, but
forested groundwater discharge areas may be ex-
ceptions. In this study, we conducted tests to deter-
mine whether highly productive forested ground-
water discharge areas generally differ from adjacent
groundwater recharge areas in terms of humus
chemistry and the availability of phosphorus (P)
and N to plants. We investigated six forested sites,
divided into groundwater discharge and adjacent
groundwater recharge areas, in northern Sweden.
The humus layers of the forested groundwater dis-
charge areas were clearly distinguished from the
adjacent groundwater recharge areas by having
higher acid-digestible calcium (Ca) and/or alumi-
nium (Al) and iron (Fe) content and higher organic
P and N content. Soil solution inorganic N (NH4ϩ
and NO3Ϫ) and pH were higher in the groundwater
discharge areas than in the groundwater recharge
areas. The organic P content showed a positive lin-
ear relationship to the Al and Fe content in the
Key words: aluminum; groundwater discharge;
iron; phosphorus limitation; phosphate sorption;
nitrogen availability.
fixation is generally restricted to the B horizon due
to the accumulation of aluminum (Al) and iron (Fe)
in this horizon (compare Wood and others 1984;
Yuan and Lavkulich 1994; Li and others 1999).
Wood and others (1984) showed a vertical strat-
ification of the relative distribution of biological
sinks (fine roots, bacteria and fungi) vs geochemical
sinks (surface-reactive Al and Fe sesquioxides),
which contribute to P retention in podzols under
temperate forests. The geochemical sinks domi-
nated in the mineral B horizon, whereas biological
INTRODUCTION
Plant growth in boreal forest ecosystems is gener-
ally considered to be nitrogen (N) limited (Tamm
1991). In contrast, many tropical forests soils are
phosphorus (P) limited due to a high P-fixing ca-
pacity and low available P contents (compare Vi-
tousek and Sanford 1986). In boreal forest soils, P
Received 2 March 2001; Accepted 9 November 2001.
*Corresponding author; e-mail: Reiner.Giesler@sbek.slu.se
300

Phosphorus Limitation in Boreal Forests
301
Ϫ
sinks dominated in the surface humus layer. Organ-
ically bound P dominates the soil P pool in the
humus layer; mineralization of this pool is thus
probably the major pathway for the release of
plant-available P (Wood and others 1984; Cade-
Menun and others 2000).
soil. In Read’s model, NO3 is the predominant N
form in P-limited areas, whereas NH4ϩ or organic N
is predominant in boreal N-limited forests.
Although regional surveys have demonstrated a
linkage between the humus layer N content and
forest productivity (Dahl and others 1967; Lahti
and Va¨isa¨nen 1987), there are no data available on
an eventual P limitation in groundwater discharge
areas other than the single productivity gradient
investigated previously. To gain more insight into
this question, we investigated five additional sites
with features similar to the previous site (Giesler
and others 1998). The objectives of this study were
(a) to compare humus layers from forested ground-
water recharge and groundwater discharge areas
with respect to plant P and N availability, soil solu-
tion composition, and P-fixation capacity; and (b)
to determine if there is a linkage between Al and Fe
in the humus and the above variables.
However, Giesler and others (1998) found large
amounts of Al and Fe in the surface humus layer of
a spodosol in a forested discharge area in northern
Sweden. A re-distribution of Al and Fe into the
humus layer generates surface sites that can adsorb
P (Bloom 1981; Pare and Bernier 1989; Gerke
1993), creating a geochemical sink to the humus
layer. Previous studies on peat soils have shown
that accumulation of Al and Fe in these organic soils
increases the P-adsorption capacity (Cuttle 1983;
Richardson 1985; Nieminen and Jarva 1996). Ac-
cumulation of Al and Fe in the humus layer has
been reported from forested discharge areas (Mul-
der and others 1991; Norrstro¨m 1993, 1995), but it
has been documented in only one case in the con-
text of P adsorption and P availability (Giesler and
others 1998). The presence of a geochemical sink in
the humus layer differs from previous findings on
the vertical stratification in spodosols and suggests
that forested discharge areas might differ in this
context.
MATERIALS AND METHODS
Study Sites
Six forested sites in northern Sweden were selected
for the study (Table 1). Each site was comprised of
a groundwater discharge area and the adjacent
groundwater recharge area. The separation be-
tween the groundwater discharge and groundwater
recharge areas was based on topography and field
layer vegetation. All sites were situated in toe slope
areas except site 3, which was situated in a mid-
slope position. Mean annual temperature and pre-
cipitation are 1°C and 570 mm, respectively. Diur-
nal mean monthly temperatures in June, July, and
August are between 11° and 14°C. At site 5, the
most northern site, mean annual temperature and
precipitation are –1°C and 500 mm, respectively.
On average, the sites are snow-covered from late
October to early May.
The field layer vegetation in the groundwater
recharge areas was dominated by dwarf shrubs
(mainly V. myrtillus L.) and intermingled with
“short-herb” plants such as Gymnocarpium dryopteris
(L.) Newman and Majanthemum bifolium (L.)
Schmidt. The groundwater discharge areas were
dominated by “tall-herb” species (Table 2) except
site 1, where grasses dominated the field layer veg-
etation. The forest stands were dominated by Nor-
way spruce (Picea abies (L.) Karsten) but with indi-
vidual trees of Scots pine (Pinus sylvestris L.) and
aspen (Populus tremula L.). The forest types, accord-
ing to Ha¨gglund and Lundmark (1977), were “tall-
herb spruce forest” for all groundwater discharge
areas except site 1, which was classified as “short-
In a landscape perspective, groundwater dis-
charge areas are found in toe slope areas; hence,
they comprise only a small part of the total area
(about 10%) (Rohde 1987). In Fennoscandian bo-
real forests, these sites often have higher forest
productivity, and have field layer vegetation dom-
inated by tall herbaceous plants rather than dwarf
shrubs and “short” herbs. Regional surveys have
demonstrated that the total N content in the humus
layer is higher in the sites dominated by tall herba-
ceous plants (Dahl and others 1967; Lahti and Va¨i-
sa¨nen 1987). Along a local productivity gradient
encompassing a large part of the natural landscape
variation, Giesler and others (1998) found that in-
organic N concentrations in the soil solution of the
humus layer increased toward the local groundwa-
ter discharge area. They also noted that inorganic N
ϩ
source shifted from NH4 domination in the
Ϫ
groundwater recharge area to NO3 domination in
the groundwater discharge area. In addition, phos-
phate concentrations were close to zero in the
groundwater discharge area, where the P-fixing ca-
pacity of the humus soil was high, and plant bioas-
says indicated a plant P limitation. Read (1991)
postulated a relationship between the predominant
form of plant-available N, mycorrhizal type, and the
most limiting nutrient and suggested that this con-
dition is related to altitude or latitude, climate, and

302
R. Giesler and others
Table 1. Location and Site Index for the Groundwater Discharge (GDI) and the Adjacent Recharge Areas
(GRE) at the Six Sites Studied
Site Index, GDI/GRE
Site No.
Site Name
Latitude/Longitude/Altitude
Area H100a(m)
1
2
3
4
5
6
Rusksele
64°50ЈN, 18°10ЈE, 230 m a.s.l.
64°25ЈN, 19°25ЈE, 225 m a.s.l.
64°25ЈN, 19°30ЈE, 225 m a.s.l.
64°59ЈN, 19°35ЈE, 220 m a.s.l.
66°01ЈN, 19°52ЈE, 230 m a.s.l.
64°39ЈN, 18°30ЈE, 235 m a.s.l.
22/18
23/18
23/18
23/18
21/16
23/18
Flakastugan
Vorrberget
Kryddgrovan
Varjisån
Betsele
aSite index estimated from site properties (vegetation, soil, moisture, latitude and altitude) according to Ha¨gglund and Lundmark (1977). Site index is the height of dominant
trees at a reference age of 100 years and is an index of relative site productivity.
Table 2. Tall-Herb Plant Species Found in the Groundwater Discharge Areas of the Different Sites
Indicating Higher Site Productivity According to Ha¨gglund and Lundmark (1977)
Species
Site No.
1
2
3
4
5
6
Aconitum septentrionale Koelle
Actaea erythrocarpa Fisher
Actaea spicata L.
Cirsium heterophyllum (L.) Hill
Crepis paludosa (L.) Moench
Dryopteris assimilis Walker
Filipendula ulmaria (L.) Maxim
Geranium sylvaticum L.
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
Lactuca alpina (L.) Grey
Matteuccia struthiopteris (L.) Tod
Paris quadrifolia L.
X
X
X
Urtica dioica L.
Nomenclature of species follows Tutin and others (1964–80).
herb spruce forest without dwarf shrubs.” All
groundwater recharge area sites were classified as
“short-herb spruce forests with dwarf shrubs.”
Soil parent materials at sites 3 and 4 were post-
glacial sandy sediments, whereas the other soils
were developed in loamy-sandy till. Soils in the
groundwater recharge areas were classified as Typic
Haplocryods (Soil Survey Staff 1992), whereas soils
in the groundwater discharge areas are Aquic or
Oxyaquic Haplocryods (Soil Survey Staff 1992)
since water saturation within 100 cm from the soil
surface most likely occurs during part of the year.
charge area, and an additional set of five humus
layer cores were taken in the adjacent groundwater
recharge area. The distance between the two areas
was less than 50 m, and the distance between in-
dividual sampling spots was greater than 3 m. Each
sample was kept separately in a polyethylene bag
and kept in a cooler at approximately 5°C during
the transport to the laboratory. Within 24 h of
collection, humus samples were sieved in the labo-
ratory through a sieve (5-mm mesh). A subsample
of approximately 100 g was taken for immediate
extraction of soil solution, and another portion was
used for determination of water content (105°C,
24 h). An additional portion was used for the P-
partitioning experiment and kept at 4°C until the
onset of the experiment within a week after sam-
pling. The rest of the humus, as well as the centri-
fuged humus, was frozen until further treatment.
Soil Sampling
The humus layer (O horizon) was sampled in July
1996 at sites 1–5 and in June 1994 at site 6. Five
humus layer cores were taken randomly with a soil
auger (0.10-m diameter) in the groundwater dis-

Phosphorus Limitation in Boreal Forests
303
In August 1999, centrifuged humus samples were
thawed and dried (40°C, 5 days). The dried samples
were gently milled by hand in a ground mortar.
The milled samples were dried at 70°C 48 h, (in
vacuum) before digestion.
and filtered (Munktell 00H filter paper; Stora AB,
Grycksbo, Sweden). The filtrate was kept at 4°C
until further analyses. The difference between the
acid-digestible P and 1.0 M HCl extractable P is
assumed to be mainly organically bound P and is
hereby denoted Po, whereas the HCl-extractable
fraction mainly constitutes inorganic P associated
with Al, Fe, or Ca.
Plants
Within 0.5 m of each humus sampling spot (sites
1–5), plant shoots of a small herb, Majanthemum
bifolium (L.) Schmidt, and a small fern, Gymno-
carpium dryopteris (L.) Newman, were sampled. The
plant species were selected because they were com-
monly found in both the groundwater discharge
and groundwater recharge areas. The sampled
plants were kept in a cooler at approximately 5°C
during transport to the laboratory, where the
shoots were immediately dried (70°C, 48 h) and
milled in a ball mill to a fine powder before analyses
of N and P (total N between 48 and 49 since some
analyses failed).
Partitioning of Phosphate
The phosphate adsorption capacity of the humus
was determined following the procedure of Nodvin
and others (1986). The ability to remove or release
PO4 can be described by the relationship between
the amount removed or released from the soil so-
lution (RE: in mmoles per kilogram) and the initial
amount of PO4 added (Xi: in millimoles per kilo-
gram); it is defined by the linear isotherm RE ϭ mXi
Ϫ b, where the slope (m) is defined as the partition-
ing coefficient of the initial mass isotherm (Nodvin
and others 1986)—that is, the fraction of total re-
active substance in a soil/water system that is re-
tained by the soil. A value of 1.0 thus indicates total
adsorption, whereas a value of 0 indicates no ad-
sorption.
We equilibrated 25 ml of either 0 or 1 mmol of
KH2PO4 with about 1 g dw of humus on a shaker
(16 h, 20°C) using three replicates from the
groundwater discharge and recharge areas of sites
1–5. After the equilibration period, the humus sus-
pension was filtered (Munktell 00H filter paper;
Stora AB) and the filtrate collected for further anal-
yses.
Soils Solution Extractions
Soil solution was extracted on a portion of the
humus samples within 24 h after sampling using a
centrifugation drainage technique (Giesler and
Lundstro¨m 1993). Humus samples were centri-
fuged for 30 min at 14000 rpm and filtered through
a 0.45-␮m filter (Syrfil-M; Costar Corporation,
Cambridge, MA, USA). Soil solution pH was deter-
mined immediately after centrifugation on a sub-
sample of the soil solution. The remaining soil so-
lution was kept at 4°C until further analyses.
Acid-digestible and Acid-extractable Pools
of Al, Fe, Ca, and P in the Humus
Plant Bioassay
Acid-digestible Al, Fe, calcium (Ca), and P in air-
dried humus were determined using about 0.2 g of
humus. The samples were digested in a mixture of
nitric and perchloric acids (10 and 1 ml, respec-
tively) in 50 ml Teflon tubes at 110°–130°C until
less than 1 ml of the acid remained (but not con-
tinued to dryness). After digestion, de-ionized wa-
ter was added to a total volume of 10 ml and the
solution kept for further analyses. In addition, py-
rophosphate-extractable Al and Fe were deter-
mined on humus samples from the groundwater
discharge areas of sites 1–5 following the procedure
of Buurman and others (1996) and using 2 g dry
weight (dw) of humus and 50 ml sodium pyrophos-
phate. Sodium pyrophosphate is assumed to extract
mainly organically bound Al and Fe.
A plant bioassay to test plant response to N, P, and
NP treatments was conducted using humus layer
soils with different P-fixing capacities. Humus from
the groundwater discharge and the adjacent
groundwater recharge area at site 6 were used for
the plant bioassay (previously described in Giesler
and others 1998; groundwater discharge area hu-
mus equals 90 m position, and groundwater re-
charge area humus equals 80 m position in the
described productivity gradient). Due to the accu-
mulation of Al and Fe, the humus from the ground-
water discharge area has a high P-adsorption capac-
ity in contrast to the humus from the adjacent
groundwater recharge area, which has a low P-
adsorption capacity (Giesler and others 1998).
Acid-washed (0.01 M HCl) silica sand (Silversand
90; Ahlsell Mineral, Sweden) was mixed with hu-
mus so that a loss on ignition (LOI) of approxi-
mately 4% was achieved. The mixed soil was used
The acid-extractable amount of P (PHCl) was de-
termined using 1.0 M HCl. Air-dried humus sam-
ples were shaken (16 h; humus:solution ratio, 1:30)

304
R. Giesler and others
as substrate, and 420 g of the substrate was used for
each 0.3-L pot.
shoot and root biomasses were determined. The
plants were dried (70°C, 24 h) and milled (roots and
shoots together), and P and N concentrations were
determined for the plants.
Six different treatments, control (C), N, P1, P2
(that is, two levels of P addition), NP1, and NP2 and
one plant species (the herb Solidago virgaurea L.,
found in both groundwater recharge and discharge
areas) were used in the plant bioassay. Each treat-
ment was replicated four times (n ϭ 4 replicates ϫ
6 treatments ϫ 2 soils) and only one plant was
growing in each pot. Before the start of the assay, P
(treatments P1, P2, NP1, and NP2) was added as a
single addition to the groundwater discharge area
substrate with a high P-adsorption capacity to
achieve an equilibrium concentration of 300 (P1)
and 600 (P2) ␮M PO4 based on a P-adsorption
isotherm for the humus (Petersson 1998). Equilib-
rium concentrations were selected so that 300 ␮M
was approximately equal to the soil solution con-
centration found in the humus layer of the adjacent
groundwater recharge area (Giesler and others
1998).
Analyses
Plant and humus digests were analyzed for Ca, Al,
and Fe using inductively coupled plasma-mass
spectrometry (ICP-MS; Perkin-Elmer SCIEX, Nor-
walk, CT, USA) and P colorimetrically on a flow
injection analyzer (5020 Analyzer; Tecator, Ho¨ga-
na¨s, Sweden).
Total N in plants sampled in the field was ana-
lyzed on an elemental analyzer coupled to an iso-
tope ratio mass spectrometer (ANCA-NT solid/liq-
uids preparation module coupled to a model 20-20
IRMS; Europa Scientific Limited, Crewe, England).
Total N in plants from the bioassay and humus was
analyzed on a C and N element analyzer (2400 CHN
Element Analyzer; Perkin-Elmer SCIEX).
ϩ
Soil solution PO4, NO3Ϫ, and NH4 concentra-
Each pot was treated with de-ionized water (ev-
ery day) and N and/or P (every 2nd day) during the
experiment. The total amount of added solution
was always the same for all treatments. Phosphate
was added in two concentrations (P1, 0.4 mM; P2,
0.8 mM, respectively) as NaH2PO4 and nitrogen as
NH4NO3 (N, NP1, and NP2; concentration, 1.6
mM). The control treatment (C) received equal
amounts of solution (de-ionized water) as the N
and P treatments. All additions were made with a
dispenser (100-ml dispenser; Socorex, Switzer-
land), and the volume added was noted at each
addition to enable calculation of the total amount of
fertilizer added. The total additions of N and P for
each treatment were 150 and 300 mg P kgϪ1 dw
organic matter (P1 and P2, respectively) and 600
mg N kgϪ1 dw organic matter. The single initial
additions to the groundwater discharge area sub-
strate were 780 and 1470 mg P kgϪ1 dw (P1 and P2,
respectively). The total additions correspond to ap-
proximately 60 kgϪ1 N haϪ1, 15 and 30 kgϪ1 P haϪ1
(groundwater recharge area), and 90 and 180 kgϪ1
P haϪ1 (groundwater discharge area), assuming a
bulk density of 0.1 kg dmϪ3 and a humus layer
thickness of 0.1 m.
tions were determined colorimetrically on a flow
injection analyzer as above. Acid-extractable PO4
(PHCl) was analyzed as above. Soil solution pH was
determined on an ATI Orion pH meter model 370
and sure flow electrode model 9272 (ATI Orion,
Boston, MA, USA).
Statistics
For statistical comparison between groundwater
discharge and groundwater recharge areas, two-
way analysis of variance (ANOVA) was used with
site and area (groundwater discharge or recharge
areas) as fixed factors. Only bulked samples were
available for analyses for site 6 and were thus ex-
cluded from the statistical analyses (except for soil
solution data, where replicates were available).
Treatment effects in the plant bioassay were ana-
lyzed statistically using one-way ANOVA for each
substrate used (substrates considered independent).
Multiple comparisons in variance analyses were
performed with Tukey’s test. Because no difference
was found between the two P-addition levels (P1,
P2, NP1, and NP2), these are treated as one treat-
ment (P and NP) in the statistical analyses. Signifi-
cant differences refer to the P Ͻ 0.05 level unless
stated otherwise.
A climate chamber (Weiss, Germany) was used
for the study, and day length was set to 16 h (full
illumination). Temperature was 20°C at daytime
until day 11 and thereafter 15°C to avoid substrate
drying. Night temperature was 15°C. The relative
air humidity was set to 75% throughout the assay.
Pots were moved systematically during the experi-
ment.
RESULTS
Acid-digestible Pools of Al, Fe, Ca, and P
The groundwater discharge areas had significantly
higher acid-digestible Al, Fe, and Ca concentrations
in the humus layer than the groundwater recharge
The plants were harvested after 2 months and

Phosphorus Limitation in Boreal Forests
305
Figure 1. Comparison of ac-
id-digestible Ca, Al, Fe, and P
concentrations (mean values)
in the humus layer of
groundwater discharge and
the adjacent recharge areas
(standard error for sum of
elements, except for site 6,
where bulked samples were
used). Organic P (Po) is de-
fined as the difference be-
tween the 1.0 M HCl-ex-
tractable PO4 and the acid-
digestible P content. Sites as
in Table 1.
areas. (Figure 1). At sites 1, 3, and 4, Ca concentra-
tions were significantly higher in the groundwater
discharge areas than in the adjacent groundwater
recharge areas, whereas Al and Fe were signifi-
cantly higher in the groundwater discharge areas in
sites 1, 2, 5, and 6. Site 1 was the only site where
both Al and Fe and Ca concentrations were higher
at the groundwater discharge area than in the ad-
jacent groundwater recharge area. The relative dis-
tribution of Fe and Al differed between the sites. At
sites 5 and 6, Fe comprised about 75% of the total
amount of Al ϩ Fe, whereas Al dominated in site 2
(about 77%).
Pyrophosphate-extractable Al ranged from 15%
to 84% of the acid-digestible Al pool (mean values).
The lowest values were found at site 1 and seemed
to differ from the other sites. The amount of pyro-
phosphate-extractable Al was negatively linearly
related to pH (r2 ϭ 0.18, P ϭ 0.04), and the rela-
tionship improved excluding site 1 (r2 ϭ 0.80, P Ͻ
0.001). Pyrophosphate-extractable Fe ranged from
22% to 69% of the acid-digestible Al pool (mean
values) and was negatively linearly related to pH
(r2 ϭ 0.51, P Ͻ 0.001).
Acid-digestible P concentrations were higher in
the humus layer in the groundwater discharge ar-
eas of sites 2, 5, and 6 compared to those in the
adjacent groundwater recharge areas (Figure 1).
However, there was no difference in the concentra-
tion of PHCl between the groundwater discharge
and the groundwater recharge areas (Figure 1). The
increase in total P concentrations in the humus was
due to an increase in Po (Figure 2). Concentrations
of Po were linearly related to the acid-digestible
concentrations of AlϩFe (Figure 2A, r2 ϭ 0.65, P Ͻ
0.001). Similarly, Po was also linearly related to the
pyrophosphate-extractable Al ϩ Fe (r2 ϭ 0.53, P Ͻ
0.001) but not to the difference between the acid-
digestible and pyrophosphate-extractable pool (r2 ϭ
0.14, P ϭ 0.08), assumed to be mainly amorphous
and crystalline Al and Fe. Only a weak linear rela-
tionship was found between PHCl and AlϩFe (Fig-
ure 2B, r2 ϭ 0.16, P ϭ 0.03, only groundwater
discharge area samples included).
Total N concentrations were significantly higher
in the groundwater discharge areas than in the
groundwater recharge areas, whereas C:N and C:P,
and C:Po ratios were significantly lower in the

306
R. Giesler and others
Figure 2. The relationship
between organic P (Po) and
1.0 M HCl–extractable PO4
(PHCl) vs the total content
of Al ϩ Fe in the humus
layer (sites 1–5). Filled cir-
cles, groundwater discharge
area; open circles, ground-
water recharge area.
Table 3. Results from Two-way ANOVA Showing P Values
Variable
n
Type
Site
Type ϫ Site
Acid-digestible Ca
Acid-digestible Al
Acid-digestible Fe
Acid-digestible P
N%
C%
C/N
C/P
C/Po
50
50
50
50
50
50
50
50
50
59
59
59
59
48
48
49
48
48
49
P Ͻ 0.001
P Ͻ 0.001
P Ͻ 0.001
P Ͻ 0.001
P Ͻ 0.001
p ϭ 0.002
P Ͻ 0.001
P Ͻ 0.001
P Ͻ 0.001
P Ͻ 0.001
P Ͻ 0.001
P Ͻ 0.001
P Ͻ 0.001
P Ͻ 0.001
P Ͻ 0.001
P Ͻ 0.001
P Ͻ 0.001
P Ͻ 0.001
P Ͻ 0.001
P Ͻ 0.001
P Ͻ 0.001
P Ͻ 0.001
P Ͻ 0.001
P ϭ 0.001
P ϭ 0.435
P ϭ 0.007
P ϭ 0.001
P ϭ 0.008
P Ͻ 0.001
P ϭ 0.002
P Ͻ 0.001
P Ͻ 0.001
P Ͻ 0.001
P Ͻ 0.001
P Ͻ 0.001
P ϭ 0.004
P Ͻ 0.001
P Ͻ 0.001
P Ͻ 0.001
P ϭ 0.051
P Ͻ 0.001
P ϭ 0.020
P ϭ 0.144
P ϭ 0.071
P ϭ 0.269
P ϭ 0.935
P ϭ 0.617
P Ͻ 0.001
P ϭ 0.015
P Ͻ 0.001
P Ͻ 0.001
P Ͻ 0.001
P Ͻ 0.001
P ϭ 0.002
P Ͻ 0.001
P ϭ 0.002
P Ͻ 0.001
Soil solution PO4
Soil solution NO3
Soil solution NH4
Ϫ
ϩ
Soil solution pH
Plant P/N (M. bifolium)
Plant P/N (G. dryopteris)
Plant N% (M. bifolium)
Plant N% (G. dryopteris)
Plant P% (M. bifolium)
Plant P% (G. dryopteris)
“Type” represents the groundwater discharge and groundwater recharge areas (df ϭ 1); “site” represents the different sites studied (sites 1–5 included in analyses, except for
soil solution data, where site 6 is also included in the analyses).
Type ϫ Site is the interaction term; n ϭ 4–5.
groundwater discharge areas than in the ground-
water recharge areas (Table 3 and 4).
in the groundwater recharge areas of sites 1–5 were
all below 4.5, whereas the average pH value at site
6 was 5.31—that is, higher than some of the
groundwater discharge areas at the other sites (Fig-
ure 3). Soil solution pH was positively related to
acid-digestible Ca and Al ϩ Fe contents (mmol
kgϪ1) in the humus layer (r2 ϭ 0.83, P Ͻ 0.001;
pH ϭ 3.58 ϩ 0.97 ϫ 10Ϫ3 AlϩFe ϩ 4.8 ϫ 10Ϫ3 Ca).
Including Al ϩ Fe contents improved the linear
relationship (r2 ϭ 0.83) compared to just using the
Soil Solution Composition
Average soil solution pH values were always higher
in the humus layer of the groundwater discharge
areas than in the adjacent groundwater recharge
areas (Figure 3). The soil solution pH in the humus
layer encompassed almost four pH units, ranging
from 3.57 to 7.56. Average soil solution pH values

Phosphorus Limitation in Boreal Forests
307
Table 4. Mean Values and Standard Error for C and N Content and C:N, C:P, and C:Po Ratios in Humus
Layer from Groundwater Discharge Areas (GDI) and Adjacent Groundwater Recharge Areas (GRE)
C
C%
N%
C:N
GDI
C:P
C:Po
GDI
Site
No. GDI
GRE
GDI
GRE
GDI
GRE
GRE
GDI
GRE
GRE
(g C/m2 ϫ 103) (%dw)
(%OM)
1
2
3
4
5
6
1.64
2.88
3.70
5.40
2.04
2.32
2.24
2.80
2.96
4.50
2.64
2.15
33 Ϯ 3
36 Ϯ 3
37 Ϯ 1
36 Ϯ 1
34 Ϯ 3
29 Ϯ nd
45 Ϯ 2
35 Ϯ 6
37 Ϯ 3
45 Ϯ 3
44 Ϯ 1
2.5 Ϯ 0.1 1.8 Ϯ 0.1 23 Ϯ 0
1.8 Ϯ 0.1 1.6 Ϯ 0.2 28 Ϯ 1
2.4 Ϯ 0.1 1.8 Ϯ 0.1 21 Ϯ 1
2.3 Ϯ 0.1 1.6 Ϯ 0.1 22 Ϯ 1
2.8 Ϯ 0.1 1.9 Ϯ 0.0 20 Ϯ 1
31 Ϯ 1
33 Ϯ 2
29 Ϯ 2
34 Ϯ 4
30 Ϯ 1
316 Ϯ 25 528 Ϯ 38 397 Ϯ 35 668 Ϯ 44
247 Ϯ 27 445 Ϯ 71 270 Ϯ 33 556 Ϯ 101
342 Ϯ 19 573 Ϯ 30 378 Ϯ 22 678 Ϯ 40
410 Ϯ 58 691 Ϯ 111 452 Ϯ 68 793 Ϯ 122
181 Ϯ 22 423 Ϯ 25 203 Ϯ 28 649 Ϯ 10
33 Ϯ nd 3.1 Ϯ nd 2.3 Ϯ nd 17 Ϯ nd 22 Ϯ nd 148 Ϯ nd 326 Ϯ nd 184 Ϯ nd 597 Ϯ nd
OM, organic matter; nd, not determined; Po, organic P fraction
The C content (expressed as g C/m2) is calculated from the mean humus layer thickness and an assumed bulk density of 100 kg mϪ3
.
Ca content (r2 ϭ 0.63). Soil solution pH values in
sampling spots with a high AlϩFe content were
underestimated in the latter case.
charge areas (mean values, 0.18 and 0.19 for M.
bifolium and G. dryopteris, respectively; P Ͻ 0.001,
two-way ANOVA). Low plant P:N ratios were asso-
ciated with low plant P concentrations rather than
with high plant N concentrations (Figure 5).
Plant N concentrations in the two plant species
were correlated (r ϭ 0.72, P Ͻ 0.001). Similarly, P
concentrations in the plants from the groundwater
discharge areas were strongly correlated (r ϭ 0.80,
P Ͻ 0.001), whereas the relationship was weaker
for plants sampled in the groundwater recharge
area (r ϭ 0.64, P Ͻ 0.001).
The highest inorganic N (NH4ϩ ϩ NO3Ϫ) concen-
trations were generally found in the groundwater
discharge areas (Figure 3). Ammonium was the
dominant inorganic N species, except for site 6 (Fig-
–
ure 3). Soil solution NO3 concentrations above 10
␮M were found only at pH values above 6.
The lowest PO4 concentrations were found in
groundwater discharge areas at sites. 2, 5, and 6
(PO4 concentrations Ͻ10 ␮M), whereas higher PO4
concentrations (Ͼ100 ␮M) were found in the
groundwater recharge areas of sites 1, 5, and 6
(Figure 3). Notably, there is a larger variation in soil
solution PO4 concentrations in the groundwater
recharge areas than in the groundwater discharge
areas (Figure 3).
The three groundwater discharge areas with the
lowest PO4 concentrations (Ͻ10 ␮M, sites 2, 5, and
6) were also those with the highest partitioning
coefficients (Figure 4). The partitioning coefficient,
m, was linearly related to the amount of Al and Fe
in the humus (Figure 4, r2 ϭ 0.82, P Ͻ 0.001; m ϭ
66.2 ϫ 10Ϫ3 ϩ 1.3 ϫ 10Ϫ3 Al (mmol kgϪ1) ϩ 0.4 ϫ
10Ϫ3 Fe (mmol kgϪ1)).
Plant P concentrations could be predicted from
the soil solution composition (PO4, pH, and soil
ϩ
–
solution P:N ratio; N ϭ NH4 ϩ NO3 ) using mul-
tiple linear regression, explaining 58% and 82% of
the variation for M. bifolium and G. dryopteris, re-
spectively (Table 5). The predictability for plants
from the groundwater discharge area was generally
better than for plants from the groundwater re-
charge area. For plant N contents, the degree of
explanation was less (data not shown).
Plant Bioassay
P and NP additions increased plant growth signifi-
cantly in soils from the groundwater discharge area
compared to the untreated control, but no growth
response was found for N additions (Figure 6). In
contrast, no significant plant growth response was
found for P additions in the groundwater recharge
area soil (Figure 6), but the plant biomass tended to
increase with N or NP additions.
P additions (as P or NP) increased the plant P
uptake in plants grown in the groundwater dis-
charge area soil, but only the combination of N and
P (NP) increased P uptake in plants grown in the
groundwater recharge area soil (Figure 6). The
N and P Concentrations in Field Layer
Plants
N concentrations were generally higher and P con-
centrations lower in plants sampled in the ground-
water discharge areas compared to the adjacent
groundwater recharge areas (Figure 5). Plant P:N
ratios were generally lower in plants from the
groundwater discharge areas (mean values, 0.10
and 0.09 for M. bifolium and G. dryopteris, respec-
tively) than in plants from the groundwater re-

308
R. Giesler and others
Figure 3. Comparison of soil
solution pH, soil solution
concentrations of inorganic N
ϩ
(NH4 and NO3N), and soil
solution PO4 concentrations
in the humus layer of
groundwater discharge and
the adjacent recharge areas.
Mean values and standard
error. Sites as in Table 1.
plant N uptake increased with NP additions, but it
tended to decrease with only N additions in plants
grown in the groundwater discharge area soil.
Plants grown in the groundwater recharge area soil
increased their N uptake with N and NP additions,
but P uptake only increased with NP additions. The
average seed content of N and P was 0.010 and
0.001 mg/seed, respectively, and plant uptake of N
and P exceeded seed contents more than 1000
times.
increased only for P additions in the groundwater
recharge area.
Relationship Between Soil Solution
Inorganic N, pH, and C:N Ratios
ϩ
Total inorganic N (NH4 ϩ NO3Ϫ) concentrations
correlated strongly with soil solution pH (r ϭ 0.85,
P Ͻ 0.001, Pearson correlation), indicating that sites
with higher pH also had a higher inorganic N soil
solution concentration (Figure 7). The increase in
soil solution inorganic N also corresponded to a
decrease in humus C:N ratios.
Plant P:N ratios increased from below 0.1 to
about 0.2 for the P and NP additions in the ground-
water discharge area (Figure 6). Plant P:N ratios

Phosphorus Limitation in Boreal Forests
309
Figure 4. (A) Relationship
between partitioning coeffi-
cient (m) and predicted m
(m ϭ 66.2 ϫ 10Ϫ3 ϩ 1.3 ϫ
Al ϩ 0.4 ϫ 10Ϫ3 ϫ Fe, r2
ϭ 0.82, P Ͻ 0.001). (B)
Relationship between the
partitioning coefficient (m)
and the soil solution PO4
concentrations. Filled cir-
cles, groundwater discharge
area; open circles, ground-
water recharge area.
Figure 5. The relationship
between plant N and P
concentrations in (A) M.
bifolium and (B) G.
dryopteris. The slope of the
broken line represents a
plant P:N ratio of 0.1; val-
ues below 0.1 are consid-
ered to indicate a P limita-
tion (Linder 1995).
Table 5. Results from Stepwise Multiple Regression Using Plant Shoot P Concentration as Dependent
Variable and Soil Solution Concentrations (pH, NH4ϩ, NO3Ϫ, PO4 (␮mol/L), and Soil Solution P:N Ratio;
ϩ
N ϭ NH4 ϩ NO3Ϫ) as Independent Variables
Dependent
Variable
Species
n
r2
P value
Model
M. bifolium
G. dryopteris
P
P
47
48
0.56
0.83
Ͻ0.001
Ͻ0.001
P ϭ 0.28 ϩ 0.0004 ϫ PO4–0.02 ϫ pH
P ϭ 0.26 ϩ 0.0007 ϫ PO4–0.02 ϫ pH ϩ 0.004
ϫ soil solution P:N ratio
One sample omitted as outlier
hydrochemistry between groundwater discharge
areas and groundwater recharge areas will affect,
directly or indirectly, soil biological processes and
nutrient availability. In the following sections, we
will discuss how these factors may interrelate.
DISCUSSION
The humus layer in forested groundwater discharge
areas is clearly distinguished from the adjacent
groundwater recharge areas in terms of C:N and C:P
ratios, total N and P content, pH, soil solution inor-
ganic N, and accumulation of total Ca and/or Al and
Fe. A high P-fixation capacity, and thus a low soil
solution PO4 concentration, also occurs in cases
where the Al and Fe content in the humus are
particularly high. We believe that differences in
Impact of Groundwater Discharge
The accumulation of Al, Fe, and Ca in the humus
layer of the groundwater discharge areas is most
likely linked to an upward water flow that redis-

310
R. Giesler and others
Figure 6. Effects of P, N,
and N ϩ P additions in a
plant bioassay (mean and
standard errors) using hu-
mus layer soils from a
groundwater recharge area
and a groundwater dis-
charge area (site 6, Table 1)
with low and high P-fixing
capacity, respectively. Dif-
ferent letters denote signifi-
cant treatment differences.
tributes these elements from the mineral soil into
the humus layer. Other mechanisms for redistribu-
tion, such as litterfall (compare Vogt and others
1987) or bioturbation due to soil animal activity,
seem unlikely explanations for the large differences
we found. For instance, at Betsele (site 6), the shift
in Al and Fe content appears within some meters
between the groundwater recharge and groundwa-
ter discharge areas. Large spruce trees (more than
30 m in height) grow on the margin between the
two areas; litterfall from these trees will be evenly
distributed and unlikely to generate the distinct

Phosphorus Limitation in Boreal Forests
311
Figure 7. Relationship be-
tween (A) soil solution pH
(centrifugate) and soil solu-
ϩ
tion inorganic N (NH4
ϩ
NO3Ϫ) in the humus layer
and (B) between soil solu-
tion pH (centrifugate) and
humus layer C:N ratios.
Data from the 90-m pro-
ductivity gradient at Betsele
are included (from Giesler
and others 1998). Filled
circles, groundwater dis-
charge area; opened circles
groundwater recharge area.
shifts that can be seen in the soil. In the Betsele
case, when the mineral soil was investigated
(Giesler and others 1998), a large increase in Ca
saturation in the groundwater discharge area was
found in the Bs horizon within the same short
distance, supporting the hypothesis that the
changes are linked to hydrochemistry rather than
There is no evidence of systematic differences in
soil mineralogy between groundwater recharge and
groundwater discharge areas. Differences in Ca
contents between soils in groundwater recharge
and groundwater discharge areas are probably due
to the transport of calcium as Ca2ϩ, although or-
ganic complexes might not be excluded. Calcium
accumulation in the groundwater is probably a re-
sult of long-distance transport, generating condi-
tions for a time-dependent accumulation of weath-
ering products and alkalinity. At site 4 (the site with
the highest Ca accumulation), pH values in a small
brook discharging into the groundwater discharge
area were higher than 8.0, suggesting an effect from
the more basic rocks in close proximity to the
groundwater discharge area.
The groundwater discharge of especially Ca in
combination with a high alkalinity would maintain
a high base saturation and thus a high pH in the
groundwater discharge area soils (Giesler and oth-
ers 1998), as indicated by the strong positive rela-
tionship between the contents of Ca in the humus
layer and soil solution pH.
aboveground
redistribution.
Moreover,
the
amounts of Al and Fe in humus documented in
previous studies from forested groundwater re-
charge areas (Giesler and others 2000; Rustad and
Cronan 1995) were much lower than those found
in the groundwater discharge areas.
The particular species of Al, Fe, and Ca redistrib-
uted in the soil solution might differ due to local
conditions. Iron can be transported as Fe2ϩ during
reducing conditions—for example, during water-
logged conditions. At least at two sites (sites 5 and
6), rapid changes can be observed (within hours) in
the ground water table during rain episodes, some-
times resulting in flooding. A high ground water
table that is stagnant for long periods, as is some-
times seen during snowmelt and longer rainfall ep-
isodes, would favor reducing conditions. This could
promote the dissolution of Fe(OH)3 (s) accumu-
lated in the Bs horizon and the formation of soluble
Fe2ϩ and might explain why the humus Fe contents
at sites 5 and 6 are higher compared to the other
sites. Transport as inorganic Fe (except Fe2ϩ) or Al
species is not likely because soil solution pH in the
discharge areas is above 5.0 in the humus layer and
pH increases with depth (Giesler and others 1998).
Alternatively, Al and Fe might be redistributed as
organic complexes or as colloidal hydroxide parti-
cles. High-molecular-weight organic Al and Fe
complexes can be found in the Bs horizon of spo-
dosols (van Hees and others 2000) and may be
redistributed with an upward water flow.
Availability of Phosphate
Both the plant bioassay and plant P:N ratios suggest
that the P availability is lower at least for field layer
plants in groundwater discharge areas with a high
Al and Fe content than in the adjacent groundwater
recharge areas. The results support previous find-
ings from the Betsele site, (Giesler and others
1998), where both plant P:N ratios and 32P root
bioassays indicated a P limitation in the groundwa-
ter discharge area. To what extent low P availability
also affects trees is still unclear, especially since
large trees might extend their roots beyond areas
with low P availability and exploit areas with more
easily available P sources.

312
R. Giesler and others
Low soil solution phosphate concentrations as a
result of a high Al and Fe content in the humus
layer, and thus a high P-fixation capacity, is likely
to be the main cause of the low P availability,
although Ca–phosphate precipitation may play a
role in sites with a high pH. The presence of a
geochemical sink in the humus layer contrasts with
the general view of boreal forests with a geochemi-
cal sink in the mineral soil, namely, the B horizon
(compare Wood and others 1984; Cade-Menun and
others 2000). The P-fixation capacity might actually
be higher in the humus layer than in the B horizon
because the complexation of Al and Fe with organic
material inhibits their crystallization (Schwertman
and others 1986) and thus generates more adsorp-
tion sites per Al and Fe than crystalline forms of Al
and Fe.
The accumulation of excess amounts of organic P
compared to the adjacent groundwater recharge
areas is most likely also related to the presence of Al
and Fe in the groundwater discharge areas and the
adsorption of organic P compounds to Al and Fe
surfaces. Organic P compounds such as inositol
phosphates, glucose phosphates, and phosphonates
have been shown to adsorb strongly to goethite or
Al precipitates (Celi and others 1999; Nowack and
Stone 1999; Ognalaga and others 1994; Shang and
others 1990, 1992; Anderson and others 1974). It is
likely that the adsorption of organic P compounds
also occurs on organically complexed Al and Fe as
well as other Al and Fe surfaces, as has also been
shown for phosphate (Bloom 1981; Gerke 1993).
The higher degree of explanation for the relation-
ship between Po and pyrophosphate-extractable
AlϩFe compared to the linear relationship between
Po and the assumed amorphous and crystalline Al
and Fe forms supports this assumption. Magid and
others (1996) and Stewart and Tiessen (1987) have
argued that adsorbed organic P compounds are less
susceptible to enzymatic biodegradation, since the
organic P compound is bound to the surface via the
phosphate group and thus blocked from attack by
phosphatase enzymes. This could explain why or-
ganic P is accumulating in the forested groundwater
discharge areas. The soil C:P ratios in the ground-
water discharge areas were comparable to those
found in spodic B horizons (Cade-Menun 2000),
whereas those in the groundwater recharge area
were more comparable to ratios in the organic ho-
rizon (Cade-Menun 2000). It has been suggested
that high C:P ratios (above 300–500) (Blair 1988;
Saggar and others 1998) are indicative of net im-
mobilization of P; thus, net immobilization would
be dominating in the humus of the groundwater
recharge areas but not in the groundwater dis-
charge areas. The use of C:P ratios is, however, very
crude because the ratio does not consider actual C:P
ratios for the organic P compounds, but instead uses
the total C content. Also, the lower C:P ratios in the
groundwater discharge areas may reflect a de-
creased biodegradation.
Plant P concentration and P:N ratios could be
predicted from soil solution pH, PO4 concentration,
and soil solution N:P ratios, especially from the
groundwater discharge areas (Table 5). This was
surprising because the soil variables represent only
a single occasion, whereas plant P and N concen-
trations can be assumed to represent integrated val-
ues over a relatively longer time period. The better
predictability for plants growing in the groundwater
discharge areas probably reflects more stable P
availability conditions in the groundwater discharge
areas. Availability of P in the groundwater recharge
areas is most likely primarily determined by the
mineralization of organic P, and spatial and tempo-
ral variability in soil solution PO4 concentrations
may be large. Therefore, plant P concentrations
may not correlate with a single soil solution mea-
surement.
Availability of N
The buildup of a readily available pool of inorganic
N with increasing pH indicates that there is an
excess of N in relation to plant and microbial N
demand. The relationship between inorganic N on
one hand and pH and humus layer C:N ratios on the
other suggests that there is a threshold value
around pH 5, above which inorganic N in the soil
solution starts to increase (Figure 7). Below pH 5,
C:N ratios start to increase from an average of 24 to
values above 40. The range in humus layer C:N
ratios found in these unpolluted sites is similar to
those found in the NITREX project (nitrogen satu-
ration experiments), which encompassed sites from
Switzerland to southwestern Sweden (Emmet and
others 1998). In fact the single 90-m–long transect
at the Betsele site (Giesler and others 1998) shows
the same range in C:N ratios as the whole NITREX
transect.
High nitrate concentrations in the soil solution
(indicative of nitrification) were found at high pH
and C:N ratios below 24. This corresponds to earlier
findings about autotrophic nitrification (Emmet
and others 1998; McNulty and others 1996; Krie-
ϩ
bitzsch 1978). A high availability of NH4 in com-
bination with a decreased availability of P may fa-
vor autotrophic nitrifiers in their competition for
ϩ
NH4 with plants (Riha and others 1986; Zak and
others 1990; Norton and Firestone 1996).
Our results do not provide any evidence to show

Phosphorus Limitation in Boreal Forests
313
whether the higher inorganic N supply with in-
creasing pH is due to in situ turnover processes or
the result of inflow of N. Leaching losses of inor-
ganic N from unpolluted groundwater recharge ar-
eas can, however, be assumed to be small (Johan-
nisson and others 1999; Ho¨gberg 2001); thus, they
would not favor a larger short-term inflow of N to
groundwater discharge areas. Gundersen and oth-
ers (1998) found that gross and net mineralization
rates were highest in the NITREX sites with low C:N
ratios (20–22). Prescott and others (2000) found
appreciable net N mineralization rates only at C:N
ratios lower than 35. These findings suggest that in
situ processes are the main cause for the higher
short-term N supply and that the high rate of N
cycling is faster in N-rich sites.
into account in any evaluation of the effects of
anthropogenic N deposition.
ACKNOWLEDGMENTS
We thank Dr. Ursula Falkengren-Grerup for provid-
ing S. virgaurea seeds and Leif and Birgitta Olsson
for help with many chemical analyses. We are also
indebted to Mona Ho¨gberg for assistance and ad-
vice. The study received financial support from the
Swedish Research Council for Forestry and Agricul-
ture (SJFR) (project 23.0161/98).
REFERENCES
Anderson G, Williams EG, Moir JO. 1974. A comparison of the
sorption of inorganic orthophosphate and inositol hexaphos-
phate by six acid soils. Soil Sci 25:51–62.
Binkley D, Ho¨gberg P. 1997. Does atmospheric deposition of
nitrogen threaten Swedish forests? For Ecol Manage 92:119–
152.
CONCLUSIONS
Blair JM. 1988. Nitrogen, sulphur and phosphorus dynamics in
decomposing deciduous leaf forest floor in the southern Ap-
palachians. Soil Biol Biochem 20:693–701.
Highly productive groundwater discharge areas in
boreal forests can be P limited rather than N limited.
This condition is most likely linked to the influx of
Al and Fe to the humus layer in these sites, decreas-
ing P availability due to P fixation. Sites high in Al
and Fe also seem to accumulate organic P, possibly
due to specific adsorption of organic P compounds
to Al and Fe and a reduced biodegradability due to
the adsorption. In this investigation, we focused on
forested groundwater discharge areas with high
productivity; however, a high P-fixation capacity
due to an accumulation of Al and Fe is not neces-
sarily restricted to richer sites, but may also occur in
other groundwater discharge areas.
Bloom PR. 1981. Phosphorus adsorption by an aluminium–peat
complex. Soil Sci Soci Am J 45:267–72.
Buurman P, van Lagen B, Velthorst EJ, editors. 1996. Manual for
soil and water analyses. Leiden: Backhuys.
Cade-Menun BJ, Berch SM, Preston CM, Lavkulich LM. 2000.
Phosphorus forms and related soil chemistry of podzolic soils
on northern Vancouver Island. I. A comparison of two forest
types. Can J For Res 30:1714–1725.
Celi L, Lamacchia S, Marsan FA, Barberis E. 1999. Interactions of
inositol hexaphosphate on clays: adsorption and charging phe-
nomena. Soil Sci 164:574–585.
Cuttle SP. 1983. Chemical properties of upland peats influencing
the retention of phosphate and potassium ions. J Soil Sci
34:75–82.
Dahl E, Gjems O, Kjelland-Lund J Jr. 1967. On the vegetation of
Norwegian conifer forest in relation to the chemical properties
of the humus layer. Meddelende i Norske Skogsforsoksvesen
85:501–531.
Humus layer solution inorganic N (as well as total
N in the humus layer) increased with pH and was
highest in the groundwater discharge areas. A high
pH may favor in situ N turnover rates; thus, the
effect on N availability is indirectly linked to the
hydrochemistry due to an inflow of Ca and alkalin-
ity maintaining a high pH in the humus layer (see
Ho¨gberg 2001 for a detailed discussion). The com-
bination of high N availability and plant P limitation
may favor nitrification. Further research may reveal
whether nitrifiers have a competitive advantage in
highly productive forest sites with plant P limita-
tion.
Our findings show that inherent site-specific con-
ditions are important to the understanding of N and
P dynamics, especially in the context of the ongoing
discussion on effects of elevated N deposition (com-
pare Binkley and Ho¨gberg 1997; Emmet 1999). Site
conditions prior to increased N deposition rates
probably have a large influence on the response to
elevated N inputs and therefore need to be taken
Emmett BA. 1999. The impact of nitrogen on forest soils and
feedbacks on tree growth. Water Air Soil Pollut 116:65–74.
Emmett BA, Boxman D, Bredemeier M, Gundersen P, Kjonaas
OJ, Moldan F, Schleppi P, Tietema A, Wright RF. 1998. Pre-
dicting the effects of atmospheric nitrogen deposition in coni-
fer stands: evidence from the NITREX ecosystem-scale exper-
iments. Ecosystems 1:352–360.
Gerke J. 1993. Phosphate adsorption by humic/Fe-oxide mix-
tures aged at pH 4 and 7 and by poorly ordered Fe-oxide.
Geoderma 59:279–88.
Giesler R, Ho¨gberg M, Ho¨gberg P. 1998. Soil chemistry and
plants in Fennoscandian boreal forest as exemplified by a local
gradient. Ecology 79:119–37.
Giesler R, Ilvesniemi H, Nyberg L, van Hees P, Starr M, Bishop K,
Kareinen T, Lundstro¨m US. 2000. Distribution and mobiliza-
tion of Al, Fe and Si in three podzolic soil profiles in relation to
the humus layer. Geoderma 94:249–63.
Giesler R, Lundstro¨m U. 1993. Soil solution chemistry: effects of
bulking soil samples. Soil Sci Soci Am J 57:1283–8.
Gundersen P, Emmett BA, Kjonaas OJ, Koopmans CJ, Tietema

314
R. Giesler and others
A. 1998. Impact of nitrogen deposition on nitrogen cycling in
Read DJ. 1991. Myccorhizas in ecosystems. Experientia 47:376–
forests: a synthesis of NITREX data. For Ecol Manage 116:65–
91.
74.
Richardson CJ. 1985. Mechanisms controlling phosphorus re-
Ha¨gglund B, Lundmark JE. 1977. Site index estimations by
means of site properties, Scots Pine and Norway spruce in
Sweden. Studia For Suecica 103:1–38.
tention capacity in freshwater wetlands. Science 22:1424–7.
Riha SJ, Campbell GS, Wolfe J. 1986. A model of competition for
ammonium among heterotrophs, nitrifiers, and roots. Soil Sci
Soc Am J 50:1463–6.
Ho¨gberg P. 2001. Interactions between hillslope hydrochemistry,
nitrogen dynamics, and plants in Fennoscandian boreal forest.
In: Schulze ED, Harrison SP, Heimann M, Holland EA, Loyd J,
Prentice IC, Schimel DS, editors. Global biogeochemical cycles
in the chmate system. San Diego: Academic Press. p 227–33.
Rohde A. 1987. The origin of streamwater traced by oxygen-18
[dissertation]. Uppsala: University of Uppsala. 260 p.
Rustad LE, Cronan CS. 1995. Biogeochemical controls of alumi-
num chemistry in the O horizon of a red spruce (Picea rubens
Sarg.) stand in central Maine, U.S.A. Biogeochemistry 29:107–
129.
Johannisson C, Myrold DD, Hogberg P. 1999. Retention of ni-
trogen by a nitrogen-loaded Scotch pine forest. Soil Sci Soci
Am J 63:383–9.
Saggar S, Parfitt RL, Salt G, Skinner MF. 1998. Carbon and
phosphorus transformations during decomposition of pine for-
est floor with different phosphorus status. Biol Fertil Soils
27:197–204.
Kriebitzsch WU. 1978. Stickstoffnachlieferung in saure Waldbo¨-
den Nordwestdeutschlands. Scripta Geobot 14:1–66.
Lahti T, Va¨isa¨nen RA. 1987. Ecological gradients of boreal forests
in South Finland: an ordination test of Cajander’s forest site
type. Vegetatio 68:145–56.
Schwertman U, Kodama H, Fisher WR. 1986. Mutual interac-
tions between organics and iron oxides. In: Huang PM,
Schnitzer M, editors. Interactions of soil minerals with natural
organics and microbes. Madison (WI): Soil Science Society of
America. p 223–250.
Li YC, Alva AK, Calvert DV. 1999. Transport of phosphorus and
fractionation of residual phosphorus in various horizons of a
spodosol. Water Air Soil Pollut 109:303–12.
Linder S. 1995. Foliar analyses for detecting and correcting nu-
trient imbalance in Norway spruce. Ecol Bull 44:178–90.
Shang C, Huang PM, Stewart JWB. 1990. Kinetics of adsorption
of organic and inorganic phosphates by short-range ordered
precipitate of aluminum. Can J Soil Sci 70:461–470.
McNulty SG, Aber JD, Newman SD. 1996. Nitrogen saturation in
a high elevation New England spruce-fir stand. 84:109–21.
Shang C, Stewart JWB, Huang PM. 1992. pH effects on kinetics
of adsorption of organic and inorganic phosphates by short-
range ordered aluminum and iron precipitates. Geoderma 53:
1–14.
Magid J, Tiessen H, Condron LM. 1996. Dynamics of organic
phosphorus in soils under natural and agricultural systems. In:
Piccolo A, editor Humic substances in terrestrial ecosystems.
Amsterdam: Elsevier. p 429–66.
Soil Survey Staff. 1992. Keys to soil taxonomy. 5th ed. SMSS
technical monograph no. 19. Blacksburg (VA): Pocahontas
Press. 556 p.
Mulder J, Pijpers M, Christophersen N. 1991. Water flow paths
and the spatial distribution of soils and exchangeable cations
in an acid-rain impacted and a pristine catchment in Norway.
Water Resources Res 27:2919–28.
Stewart JWB, Tiessen H. 1987 Dynamics of soil organic phos-
phorus. Biogeochemistry 4:41–60.
Nieminen M, Jarva M. 1996. Phosphorus adsorption by peat
from drained mires in southern Finland. Scand J For Res
11:321–6.
Tamm CO. 1991. Nitrogen in terrestrial ecosystems. Berlin:
Springer-Verlag. 116 p.
Tutin TG, Heywood VH, Burges NA, Valentine DH, Walters SM,
Webb DA. 1964–80. Flora Europea; vols 1–4. Cambridge (En-
gland): Cambridge University Press.
Nodvin SC, Driscoll CT, Likens GE. 1986. Simple partitioning of
anions and dissolved organic carbon in a forest soil. Soil Sci
142:20–8.
van Hees P, Lundstro¨m US, Giesler R. 2000. Low-molecular
weight acids and their Al-complexes in soil solution—compo-
sition, distribution and seasonal variation in three podzolised
soils. Geoderma 94:173–200.
Norrstro¨m AC. 1995. Concentration and chemical species of iron
in soils from groundwater surface water ecotones. Hydrol Sci
J 40:319–29.
Norrstro¨m AC. 1993. Retention and chemistry of aluminium in
groundwater discharge areas. Environ Pollut 81:269–75.
Vitousek PM, Sanford RL Jr. 1986. Nutrient cycling in moist
tropical forest. Annual Rev Ecol Sys 17:137–167.
Norton JM, Firestone M. 1996. N dynamics in the rhizosphere of
Pinus ponderosa seedlings. Soil Biol Biochem 28:351–62.
Vogt KA, Dahlgren R, Ugolini FC, Zabowski D, Moore EE, Za-
soski R. 1987. Aluminium, Fe, Ca, Mg, K, Mn, Zn, and P in
above and below-ground biomass. I. Abies amabilis and Tsuga
mertensiana. Biogeochemistry 4:277–294.
Nowack B, Stone AT. 1999. Adsorption of phosphonates onto
goethite–water interface. J Colloid Interface Sci 214:20–30.
Wood TE, Bormann FH, Voigt GK. 1984. Phosphorus cycling in
a northern hardwood forest: biological and chemical control.
Science 223:341–393.
Ognalaga M, Frossard E, Thomas F. 1994. Glucose-1-phosphate
and myoinositol hexaphophate adsorption mechanisms on
goethite. Soil Sci Soc Am J 58:332–7.
Yanai RD. 1991. Soil solution phosphorus dynamics in a whole-
tree harvested northern hardwood forest. Soil Sci Soci Am J
55:1746–1752.
Pare D, Bernier B. 1989. Origin of the phosphorus deficiency
observed in declining sugar maple stands in the Quebec Ap-
palachians. Can J For Res 19:24–34.
Petersson T. 1998. Plant available phosphorus in humus soils: a
comparison between a recharge and a discharge area [thesis].
Umeå: Swedish University of Agricultural Sciences. 17 p.
Yuan G, Lavkulich M. 1994. Phosphate sorption in relation to
extractable iron and aluminum in spodosols. Soil Sci Soc Am J
58:343–346.
Prescott C, Chappell HN, Vesterdal L. 2000. Nitrogen turnover in
forest floors of costal Douglas fir at sites differing in soil nitro-
gen capital. Ecology 81:1878–86.
Zak DR, Groffman PM, Pregitzer KS, Christensen S, Tiedje JM.
1990. The vernal dam–plant microbe competition for nitrogen
in northern hardwood forests. Ecology 71:651–656.