Patterns and mechanisms of soil acidification in the conversion of grasslands to forests

Abstract

Grassland to forest conversions currently affect some of the world's most productive regions and have the potential to modify many soil properties. We used afforestation of native temperate humid grassland in the Pampas with eucalypts as an experimental system to 1) isolate forest and grassland imprints on soil acidity and base cation cycling and 2) evaluate the mechanisms of soil acidification. We characterized soil changes with afforestation using ten paired stands of native grasslands and Eucalyptus plantations (10–100 years of age). Compared to grasslands, afforested stands had lower soil pH (4.6 vs.5.6, p c.Ha−1.yr−1 across afforested stands, although no aboveground acidic inputs were detected in wet + dry deposition, throughfall and forest floor leachates. Our results suggest that cation cycling and redistribution by trees, rather than cation leaching by organic acids or enhanced carbonic acid production in the soil, is the dominant mechanism of acidification in this system. The magnitude of soil changes that we observed within half a century of tree establishment in the Pampas emphasizes the rapid influence of vegetation on soil formation and suggests that massive afforestation of grasslands for carbon sequestration could have important consequences for soil fertility and base cation cycles.

Full text

Patterns and mechanisms of soil acidification in the
conversion of grasslands to forests
ESTEBAN G. JOBBÁGY
1,2,
*and ROBERT B. JACKSON
1,3
1
Department of Biology, Duke University, Durham, NC 27708, USA;
2
Facultad de Agronomía, Univer-
sidad de Buenos Aires, Avenida San Martín 4453, Buenos Aires, 1417, Argentina;
3
Nicholas School of
the Environment and Earth Sciences, Duke University, Durham, NC 27708, USA; *Author for corre-
spondence (e-mail: jobbagy@agro.uba.ar; phone: +54 11 4524 8000 int 8144; fax: +54 11 4514 8730)
Received 10 June 2002; accepted in revised form 20 December 2002
Key words: Afforestation, Argentina, Eucalyptus, Nutrient cycling, Soil acidity, Vegetation change
Abstract. Grassland to forest conversions currently affect some of the world’s most productive regions
and have the potential to modify many soil properties. We used afforestation of native temperate humid
grassland in the Pampas with eucalypts as an experimental system to 1) isolate forest and grassland
imprints on soil acidity and base cation cycling and 2) evaluate the mechanisms of soil acidification. We
characterized soil changes with afforestation using ten paired stands of native grasslands and Eucalyptus
plantations (10–100 years of age). Compared to grasslands, afforested stands had lower soil pH (4.6
vs.5.6, p < 0.0001) and ⬃40% lower exchangeable Ca (p < 0.001) in the top 20 cm of the soil. At three
afforested stands where we further characterized soil changes to one meter depth, soil became increas-
ingly acidic from 5 to 35 cm depth but more alkaline below ⬃60 cm compared to adjacent grasslands,
with few differences observed between 35 and 60 cm. These changes corresponded with gains of ex-
changeable acidity and Na in intermediate and deeper soil layers. Inferred ecosystem cation balances
(biomass + forest floor + first meter of mineral soil) revealed substantial vertical redistributions of Ca
and Mn and a tripling of Na pools within the mineral soil after afforestation. Soil exchangeable acidity
increased 0.5–1.2 kmol
c
.Ha
−1
.yr
−1
across afforested stands, although no aboveground acidic inputs were
detected in wet + dry deposition, throughfall and forest floor leachates. Our results suggest that cation
cycling and redistribution by trees, rather than cation leaching by organic acids or enhanced carbonic
acid production in the soil, is the dominant mechanism of acidification in this system. The magnitude of
soil changes that we observed within half a century of tree establishment in the Pampas emphasizes the
rapid influence of vegetation on soil formation and suggests that massive afforestation of grasslands for
carbon sequestration could have important consequences for soil fertility and base cation cycles.
Introduction
Plants influence the earth surface through the uptake, transformation and redistri-
bution of materials in the atmosphere, pedosphere and lithosphere (Simonson 1959;
Likens et al. 1977; Schlesinger 1997). Although all plants leave such an imprint
(e.g., input of organic matter and respired CO
2
, enhancement of rock weathering,
etc.), variations in size, growth rate, life span, allocation, tissue chemistry, and many
other attributes affect cycling patterns and the properties of soils differently (Alban
1982; Finzi et al. 1998; Jackson et al. 2000; Jobbágy and Jackson 2000). The widely
applied state factor model of soil formation recognizes such effects and includes
Biogeochemistry 64: 205–229, 2003.
© 2003 Kluwer Academic Publishers. Printed in the Netherlands.

vegetation as a master control of pedogenesis (Jenny 1941, 1980). However, iso-
lating the effects of vegetation on soil development in the field is confounded by
covarying interactions of vegetation and other important influences of soil func-
tioning, including climate, topography, and history (Ugolini et al. 1988).
Understanding the feedback of vegetation type on soils is important because
current rates of vegetation change are high and may lead to further biotic changes
through soil modifications (Roberts 1987; Jackson et al. 2002). Shifts from grass-
lands to forests (afforestation and tree invasion/encroachment) affect some of the
most productive areas still covered by native vegetation globally, especially in the
Southern Hemisphere (Rudel and Ropel 1996; Richardson 1998; Geary 2001). In
the native grasslands of the Pampas, afforestation with pines and eucalypts is be-
coming increasingly common. In the last decade Uruguay and Argentina have in-
creased their afforested areas five- and two-fold, respectively, in this region (MAGP
1998; SAGPyA 2000), with even higher afforestation rates expected for the coming
decades (Wright et al. 2000). In such systems, most base cations are essential plant
nutrients and play a key role balancing ecosystem acidity (Aber and Melillo 1991).
Soil acidity in turn is a master control of soil fertility (Marschner 1995) and affects
many important biogeochemical processes, such as rock weathering and nitrifica-
tion (Richter and Markewitz 2001). In this paper we use the chronosequence of
forested sites compared to native grasslands as a controlled experiment for isolat-
ing the imprint of eucalypt afforestation on soil acidity and base cation cycling.
Spatial comparisons of areas historically dominated by trees and grasses indi-
cate that soil pH and base saturation tend to be lower under forests with similar
parent materials and climates (Jenny 1941; Geis et al. 1970; Ugolini et al. 1988).
Such comparisons do not address the rate and mechanisms of vegetation effects on
soil properties and are usually unable to rule out other controls that influence veg-
etation and soils together (Richter et al. 1994). Nonetheless, such studies suggest
soil acidification as a likely outcome of the conversion of grasslands to forests.
Direct manipulations of vegetation can also yield mechanistic insights into the
ecological consequences of grassland-to-forest conversions. Studies examining the
shift of native grasslands to forests reveal various degrees of surface soil acidifi-
cation after tree establishment (Davis and Lang 1991; Musto 1991; Davis 1995;
Quideau and Bockheim 1996; Parfitt et al. 1997; Alfredsson et al. 1998; Amiotti et
al. 2000). As expected, the largest decline in pH is associated with the loss of ex-
changeable base cations, particularly Ca, and increases in exchangeable Al.
We propose three basic mechanisms of soil acidification following the affores-
tation of grassland ecosystems (Figure 1): 1) organic acid inputs 2) soil respiration,
and 3) sequestration and redistribution of cations. In the first mechanism, organic
acids produced by plants are the acidifying agent. In contrast to grasses, many trees
have acidic litter, canopy leachates, and decomposition products. This acid input,
usually negligible under grassland vegetation, could cause cation leaching and a
decline in soil pH after tree establishment (Ugolini et al. 1988). A second mecha-
nism of soil acidification is carbonic acid inputs derived from soil respiration. Car-
bonic acid is an important agent of weathering in soils (Richter and Markewitz
1995), and higher carbonic acid production has been linked to increased base cat-
206

ion leaching in forests (Andrews and Schlesinger 2001). In order to acidify grass-
land soils by this mechanism, forests should have higher root respiration and/or en-
hanced microbial respiration compared to grasslands. Contrary to this suggestion,
recent reviews suggest that soil respiration rates tend to be lower in forests (Raich
and Tufekcioglu (2000); see also Chen et al. (2000) and Tate et al. (2000), Saviozzi
et al. (2001)). The third potential mechanism is the sequestration and redistribution
of cations within the ecosystem after tree establishment. Trees can store cations in
excess of anions in both aggrading biomass and in the organic floor (litter + or-
ganic soil). This difference in charge should in turn be balanced by a net gain of
protons in the soil (Nilsson et al. 1982). Even when forests approach a steady state
of cation gain and loss, soil cations may be redistributed within the soil as a result
of intense cycling by trees, leading to localized acidification in some layers (Job-
bágy and Jackson 2001).
If each of these mechanisms acts singly in controlling soil acidification, differ-
ent soil vertical patterns and ecosystem balances should develop. If organic acid
inputs drive acidification, they should be observed either in throughfall or in forest
floor leachates, causing maximum acidification in the surface soil (Figure 1a). If
soil respiration drives acidification, maximum acidification should not occur in the
surface soil where CO
2
partial pressures are lower, but at depth; there, the loss of
cations would be observed without signs of acidity in throughfall or forest floor
leachates (Figure 1b). In both of these cases, cations will be lost from the ecosys-
tem. In contrast, in the third mechanism cation redistribution should dominate the
process of acidification. Maximum acidification should then occur where root nu-
trient uptake is relatively high but inputs from litterfall and throughfall are low –
below the first ten centimeters of the mineral soil profile (Figure 1c) (see Jobbágy
and Jackson (2001)). No net ecosystem loss of cations should be observed in this
case, as forest sequestration will balance soil losses. In all likelihood, such mecha-
nisms will operate in concert.
Figure 1. Three potential mechanisms of soil acidification and their imprint on soil profiles after affor-
estation: A) organic acid inputs, B) soil respiration, and C) cation redistribution. Figures show the ver-
tical distribution of acidity accumulation and base cation pools before and after afforestation. For pur-
poses of illustration, base cation pools are assumed to be homogenously distributed under grassland
vegetation. The first two mechanisms cause a net ecosystem loss of base cations, the third yields no net
change in cation pools but redistribution from intermediate depths to the topsoil.
207

In this paper we characterize the effects of afforestation on soil chemistry in the
Pampas of Argentina, measuring soil changes in paired grassland and afforested
sites of 10 to 100 years of age. We evaluate the proposed mechanisms of acidifi-
cation in these systems and explore potential feedbacks of soil changes after affor-
estation on the nutrition of trees and grasses.
Study system
The Pampas region covers ⬃50 million ha of Argentina (Figure 2). The word
⬙Pampa⬙, derived from the Quechua language, indicates a flat, treeless extension of
land (Bravo 1967). The climate lacks a dry season and is temperate and sub-humid
to humid (mean annual T ranges from 17 °C to 14 °C north to south; Hall et al.
(1992)). Average annual rainfall ranges from 600 mm in the southwest to 1000 mm
in the northeast, but during the last three decades increased rainfall in the drier ar-
eas of the west has erased this regional gradient (Podesta et al. 1999) (Table 1).
Dominant soils are Mollisols developed on loess-like sediments deposited through-
out the late Pleistocene and the Holocene (Teruggi 1957; Tricart 1973).
The natural vegetation of the Pampas is a mixture of C3 and C4 grasses (Sori-
ano 1991). One of its most conspicuous original features was a complete lack of
trees over most of the region, an observation noted by early travelers such as Azara
(1796) and Rosas and Senillosa (1825), and Darwin (Barlow (1933); see also Ga-
ravaglia (1999)). Phytoliths indicate that grasses dominated the region throughout
the period of soil genesis (Tecchi 1983). The limitations to plant growth in the
Pampas include poor drainage and Na accumulation in lower landscape positions
(Soriano 1991).
With European settlement, plantations ranging from 0.1 to 100 ha were estab-
lished for shade, windbreak, and aesthetics. Most plantations were unfertilized and
had little wood extraction or soil disturbance. Soon after the mid 1800s, Eucalypts
became common in rural plantations (Senillosa et al. 1878; Zacharin 1996), which
today are primarily composed of E. camaldulensis,E. viminalis, and E. globulus.
The cumulative replacement of natural communities in the region with annual grain
crops and cultivated pastures ranges from 30% in the Flooding Pampas to 90% in
the Rolling Pampas (INDEC 1988). Most areas occupied by natural grasslands are
devoted to cattle ranching, where irrigation, fertilization, and feed supplementation
are almost universally absent (Soriano 1991).
The Pampas region offers an ideal opportunity to test the mechanisms of soil
acidification, both because throughout pedogenesis most of the region lacked
woody species and because it has low anthropogenic inputs of acids and N which
might otherwise confound the effect of vegetation on soil chemistry (Lavado 1983;
Bouwman et al. 1997). We used fence-line comparisons to assess the effect of veg-
etation change on soil attributes, identifying adjacent paired Eucalypt stands and
native grasslands throughout the major subregions of the Pampas using aerial pho-
tographs (1:50000) (INTA 1989) and Landsat imagery (1:250000) (IGM 1997).
208

Materials and methods
Site description
We selected 10 paired sites representing the most typical upland soils of the region
(Figure 2, Table 1): The Castelli, Guerrero, and Pila sites represent one of the most
common soil series of upland soils in the Flooding Pampas, Hapludolls over-laying
an older eroded soil that constitutes a textural B horizon at 30 to 60 cm depth. The
horizon sequence is: A-AC-IIBt-IIC and the soils were derived from loess sedi-
ments that were locally redistributed by wind in the Holocene. The Rojas and Ran-
cagua sites correspond to the most conspicuous soil series of uplands in the Rolling
Pampas. These soils are considered among the best agricultural soils of the conti-
nent and have an A-Bt-C horizon sequence with a 40 cm thick A horizon. The
Pereyra site is located in the transition between the Flooding and Rolling Pampas.
It was located in an Argialboll (planosol) with an A-E-Bt-C sequence with a tex-
tural B at 30 cm of depth. The Chascomús site is also in the transition between the
Figure 2. Location of study sites in the Pampas. Each site includes a grassland and tree plantation stand.
The major subregions covered by this study are indicated. Sites are: 1, Castelli; 2, Guerrero; 3, Pila; 4,
Chascomús; 5, Pereyra; 6, Rojas; 7, Rancagua; 8, América; 9, Tandil; and 10, Pinamar.
209

Table 1. List of study sites. Each site includes a grassland and a Eucalyptus stand. Latitude and longitude correspond to the center of the tree stand. Mean annual
precipitation (MAP) and temperature (MAT) values were obtained from the closest available (< 30 km) meteorological stations. Soil series denomination and soil
classification to subgroups (USDA 1998) were derived from 1:50000 soil cartography and from local observations. Initial tree density was derived from the original
planting grid; in the case of irregular planting patterns it could not be estimated (NA). Current tree density was sampled in the field.
Site Lat (S) Long (W) MAP (mm) MAT (C°) Soil series Soil subgroup Tree species Age (yrs) Density (stems/Ha)
Initial Current
Flooding Pampas
Castelli 36°02.0⬘57°50.3⬘980 15.3 Pila thapto Hapludoll E camaldulensis 50 1666 783
Guerrero 35°58.4⬘57°51.1⬘980 15.3 Pila thapto Hapludoll E camaldulensis 95 NA 545
Pila 36°28.4⬘58°11.8⬘895 15.0 Pila thapto Hapludoll E camaldulensis 36 1111 1022
Flooding/Rolling Pampas
Chascomús 35°25.3⬘58°02.4⬘970 15.4 Abbott aquic Argiudoll E camaldulensis 47 1111 767
Pereyra 34°51.9⬘58°09.0⬘1010 15.7 Numancia vertic Argialboll E viminalis 11 1111 1067
Rolling Pampas
Rojas 34°11.0⬘60°55.5⬘990 16.6 Rojas typic Argiudoll E camaldulensis 44 1111 1002
Rancagua 34°07.6⬘60°34.8⬘990 16.8 Rojas typic Argiudoll E camaldulensis 89 NA 220
Inner Pampas
América 35°30.6⬘62°59.2⬘745 16.5 América entic Haplustoll E camaldulensis 41 833 599
Hills
Tandil 37°20.7⬘59°02.7⬘900 12.3 Tandil lithic Hapludoll E camald./vimin. 42 625 1125
Coastal dunes
Pinamar 37°05.1⬘56°51.3⬘837 14.5 No name typic Udipsamment E camald./vimin. 41 767 690
210

Flooding and Rolling Pampas and occupied an aquic Argiudoll with an A-Bt-C se-
quence. A temporarily perched water table over a textural B located at 35 cm is
commonly observed in this type of soils (INTA 1989). The América site corre-
sponds to the most typical upland pedon of the Inner Pampas, which is a deep sandy
soil with no B horizon (A-AC-C sequence). Like soils sampled in the Flooding
Pampas, these soils developed on sediments transported by wind in the Holocene.
The Tandil site is located on the typical soil of slopes and hilltops on the few iso-
lated hills of the Pampas. The horizon sequence is A-AC-rock. The soils developed
on a thin loess layer over-laying Precambrian granite rock that formed a solid
boundary at 40 cm depth in the study site. The Pinamar site is located on the coastal
strip of sand dunes formed in the Holocene. Calcium carbonate from shell frag-
ments represents 20 to 40% (mass) of the sediments, and there is no horizon dif-
ferentiation in these young Udipsamments.
Adjacent forest and grassland sites shared the same soil unit (deduced from
1:50000 soil maps and on-site inspection of soil profiles) and were always located
in uplands. All forest stands were dominated by E. camaldulensis or E. viminalis,
were > 10 Ha in area and ungrazed, and, as confirmed with the owners or manag-
ers (and consistent with common regional practices), were never fertilized or irri-
gated. Biomass was not harvested from these forest stands except at Pila (see be-
low), where the stand was clearcut and left to regrow for 15 years before sampling,
and at Guerrero and Pinamar, where 20–50% of the trees were cut previously.
Grassland stands were neither fertilized nor irrigated but were generally grazed.
None of the grasslands was plowed except at Castelli, Chascomús, and América,
with no plowing in the last 20 years.
Tree stands had little or no grass understory, with Cynodon dactylon being the
only understory species present at a few sites. For the eucalypt stands > 40 years of
age, Celtis tala,Morus alba, and Ligustrum sp. occasionally formed a sparse woody
understory. Grassland stands were dominated by plants of the genera Paspalum,
Bothriochloa,Bromus,Stipa,Piptochaetium, and Festuca. At the coastal dune site
the native C
3
grass Cortaderia selloana was the only dominant grass species.
Soil and plant analyses
We performed three classes of analyses at the study sites. At all sites we sampled
soil to 20-cm depth and analyzed exchangeable nutrient pools, pH, organic C and
total N. Examining three paired sites (Guerrero, Castelli, and América) in more de-
tail, we extended the above sampling throughout the first meter of soil and also
measured exchangeable acidity, soluble Cl, extractable P, and total element pools.
A full accounting of tree biomass and forest floor elemental pools was also per-
formed at these three sites. Finally, at Castelli we sampled the vegetation more in-
tensively and conducted repeated measurements of wet + dry deposition and
throughfall in the forest and grassland between July 2001 and January 2002 to ex-
amine in more detail the mechanisms of soil acidification.
To avoid potential edge effects, sampling areas were > 50 m away from fences
or borders. At each stand we randomly located five soil pits (1 m wide × 0.5 m
211

deep) along a transect parallel to the grassland-forest edge. The soil pits were > 0.5
m away from the nearest tree and > 10 m away from each other. Mineral soil was
sampled at depth intervals of 0–5, 5–10, 10–20, 20–35, 35–50, 50–75, and 75–100
cm. Above 50 cm, individual samples were composited from material alonga1m
horizontal strip of the pit wall. Below 50 cm, samples were obtained with a 10-cm
diameter auger. We obtained 250-ml volumetric soil samples at each depth interval
for bulk density estimates in two soil pits per stand (Elliot et al. 1999). Litter and
organic soil layers (only present in forest stands) were sampled in each soil pit us-
ing 20 × 50 cm frames. Mineral and organic soil samples were air-dried and ag-
gregates were broken to pass a 2-mm sieve; no stones were observed except at
Tandil. Soil pH was measured with an electrode on the supernatant of a 1:1 soil-
water extract (Thomas 1996). We analyzed samples for exchangeable Ca, Mg, K,
Na, and Mn using extracts of 1 M ammonium acetate (pH 7) with a 1:5 soil-water
ratio, shaking 5 minutes and equilibrating 24 hours (Robertson et al. 1999). Con-
centrations in the extracts were measured using inductively coupled plasma emis-
sion spectrometry (ICP, Soltanpour et al. (1996)). Total organic C and N were mea-
sured using dry combustion and a CHN autoanalyzer system (Gill et al. 2002). For
samples having carbonates in the deeper layer, we ran combustion at 550 °C and
650 °C and estimated organic C as the difference between readings (Robenhorst
1988). Carbonates were detected only below 75 cm of depth at some of the sites.
For the additional soil analyses at Castelli, Guerrero, and América, we deter-
mined exchangeable acidity with an alkaline titration of 1 M KCl extracts (Robert-
son et al. 1999), Cl concentrations on 1:1 water extracts using an ion selective
electrode (Frankenberger et al. 1996), and Olsen extractable P colorimetrically (Kuo
1996). We measured total elemental soil pools using ICP after a HNO
3
/H
2
O
2
/HCl
acid – microwave digestion (EPA-3050A, see Chen and Ma (1998)). We also esti-
mated forest nutrient accumulation from the product of biomass and elemental con-
centration. The forest components analyzed were leaves, bark, wood, roots, and
forest floor, divided into litter and organic soil. We used allometric equations de-
veloped with seven trees to estimate leaf, wood and bark biomass from basal area
and height measurements at each site. We used a ratio of 0.19 for below to above-
ground biomass as determined from published equations that link belowground
biomass to aboveground biomass, stand age, and latitudinal region (Cairns et al.
1997). Basal area and tree height measurements were taken in five 0.1-ha plots at
each site; within each plot we collected five samples of fully expanded sun leaves,
bark, and wood from three randomly selected individuals.
We also sampled shoots of Cynodon dactylon as a bioindicator of soil changes.
Five randomly selected individuals were collected whenever the species was present
in both forest and grassland stands. Plant, litter, and organic soil samples were
oven-dried at 70 °C, ground (0.5 mm sieve), and HNO
3
/H
2
O
2
-digested for ICP el-
emental analysis. Carbon and N concentrations were determined by dry combus-
tion with an autoanalyzer (Gill et al. 2002).
212

Deposition and throughfall measurements
At Castelli we measured wet + dry deposition and throughfall in the forest and
grassland stands. For wet + dry deposition, we installed three 17-cm diameter PVC
funnel collectors with glass wool filters and 5-L glass bottles. Collectors were lo-
cated 3.5 meters above the ground in the grassland 100 m away from the forest
edge. We installed six collectors of the same type under the forest canopy, 60 cm
above the ground. We made 9 collections during the sampling period, with the lag
between rainfall events and collection always < 20 days. Collectors were replaced
after each sampling. Collections covered a 184-day period and accumulated 498
mm of precipitation and 316 mm of throughfall in the forest. The release of acidity
by the forest floor was evaluated in the lab with a 1:3 mixture of litter + organic
soil and deionized water shaken for 15 minutes and filtered. By doing this we at-
tempted to magnify any possible exchange between the forest floor and the perco-
lating rain water that could take place in field conditions. We used rinsed Whatman
42 paper to filter precipitation, throughfall, and forest floor extract solutions and
analyzed them using the above methods for soil extracts.
Element balances
We estimated transfers and gain/loss of base cations and protons following affor-
estation at Castelli, Guerrero, and América. For this purpose we used spatial dif-
ferences between forest and grassland stands as a surrogate for temporal changes
within forest stands established on native grasslands, assuming that forest stands
were established on soils similar to those under our native grasslands stands. We
based our estimates of changes on the observed differences between cation and
proton pools in forest and grassland stands for the top meter of mineral soil, con-
sidering the product of bulk density and concentrations from acid digestion or ti-
tration. Although compaction or expansion of the soil after afforestation could have
affected this calculation (Brimhall et al. 1991), soil volumetric contents of Ti (Jer-
sak et al. 1995) did not differ between forest and grassland stands, suggesting that
such effects were negligible (differences were always < 5% and statistically non-
significant). The ecosystem balance for each base cation was estimated from the
difference between forest accumulation and soil losses. We calculated excess cation
accumulation (ECA) of tree plantations considering biomass and forest floor pools
(ECA = Ca + Mg+K+Na+Mn+Fe−S−P). This methodology was adapted from
Richter (1986).
True replication in our study was derived from the comparisons of effects across
sites, which were evaluated using paired t-tests (Zar 1984). Statistical comparisons
between stands within each site were also performed with a t-test in order to con-
strain error estimates. All pH values were transformed to H
+
concentrations for sta-
tistical calculations.
213

Results
Soils under forest and grassland vegetation displayed consistently different acidity
and base cation composition across the study sites (Table 2). With the exception of
Pinamar, located on calcareous sand dunes, soil pH from 0–20 cm depth was one
pH unit lower on average under forests than grasslands (4.6 and 5.6, respectively;
p < 0.0001 across sites, Table 2). Among base cations, Ca and Na showed the
greatest differences. Exchangeable Ca was 40% lower on average under forest
stands than under grasslands (p < 0.001 across sites, Table 2). Maximum differ-
ences were observed at Pila (75%), the only forest that had been completely har-
vested in the past. Exchangeable Na showed the opposite trend and was higher un-
der forests at all sites except Chascomús, Tandil, and Pinamar (p < 0.05 within sites,
p = 0.0038 across sites, Table 2). Exchangeable Mg and K showed variable trends
with significant differences occurring in both directions between forest and grass-
land stands. All forests, except Pila, had higher soil organic C (SOC) (p < 0.05
within sites, p = 0.0096 across sites, Table 2), with forest stands on average having
⬃30% more SOC in the top 20 cm of mineral soil.
Soils under trees and grasses also showed consistently different vertical patterns
to one-meter depth (Figure 3). Soil pH under grasslands ranged from 5.5 to 6.5 and
decreased slightly with depth (Figure 3). In all cases forest stands had significantly
lower soil pH between 5 and 50 cm depth (p < 0.01 within sites, p = 0.0088 across
sites), with the lowest values observed between 10 and 35 cm at Castelli and
América and closer to the surface in Guerrero. All three sites had a distinct cross-
over point at ⬃60 cm, below which soil pH became more alkaline under forests
than grasslands (Figure 3). This was particularly evident at América, where the
same forest profile displayed pH values<4atintermediate depths and>8atone
meter. The pH of the organic soil horizon was ⬃5.7 at all three sites.
The effective cation exchange capacity (ECEC) of soils also differed under for-
ests and grasslands (Figure 3). Guerrero soils displayed lower ECEC under forests
throughout the top 50 cm of the profile, but at Castelli this occurred only between
20 to 35 cm depth. Differences disappeared below 50 cm with the occurrence of
the B horizon. At America significant differences were observed in the top 5 cm
only (p < 0.01). Soils under all three grasslands had 100% base saturation. Base
saturation decreased to as little as 80% under forest vegetation within 10–35 cm
depth (p < 0.05 within sites, p = 0.0024 across sites). We analyzed the relationship
between soil pH and ECEC in more detail for the A horizon (0–35 cm) at Castelli
and Guerrero, which shared the same parent material. ECEC of individual samples
was significantly and positively associated with organic C content (r
2
= 0.29), and
pH (r
2
= 0.26) (n = 80 and p < 0.0001 in both cases). The residuals of the ECEC-
organic carbon content regression showed a stronger relationship with pH (r
2
=
0.56, p < 0.0001). This regression analysis indicated that a decline of pH from 6 to
5 corresponded with an average ECEC loss of 2.24 cmol
c
/kg, a 20% decline.
The composition of the exchange complex also differed markedly between grass-
land and forest soils. After afforestation, exchangeable acidity and Na displaced Ca
at all sites (Figure 4), with changes in exchangeable Ca smallest in the surface soil.
214

Table 2. Chemical properties of the top mineral soil (0–20 cm) under native grasslands (G) and Eucalyptus forests (F). The sampled depth was completely in the A
horizon except at Pinamar were no horizon differentiation was observed. Soil pH
water
was measured in a 1:1 soil-water ratio. Exchangeable (Exch) base cations were
determined by ICP using extracts of 1 M NH4OAc at pH 7. Organic C and total N were obtained by dry combustion.
Site pH
water
Exch Ca (mg/Kg) Exch Mg (mg/Kg) Exch K (mg/Kg) Exch Na (mg/Kg) Org C (%) Total N (%)
GF GF GF GF GF GF GF
Flooding Pampas
Castelli 5.60 4.60* 1783 1191* 248 271 602 442* 0 42* 2.9 3.9* 0.25 0.30*
Guerrero 5.70 4.30* 1362 679* 224 203* 458 316* 4 84* 2.5 2.9* 0.23 0.23
Pila 5.58 4.42* 1229 305* 383 218* 403 244* 50 89* 2.7 2.2* 0.24 0.18*
Flooding/Rolling Pampas
Chascomús6.15 5.76* 1134 1100 187 345* 656 473* 43 31 2.5 3.5* 0.22 0.25*
Pereyra 5.29 4.59* 1703 1013* 193 191 364 247* 0 17* 2.1 2.4* 0.19 0.20
Rolling Pampas
Rojas 5.17 3.73* 1962 887* 258 183* 809 603* 9 100* 1.7 2.9* 0.21 0.29*
Rancagua 5.29 4.37* 1756 833* 259 176* 760 916* 12 49* 1.9 2.6* 0.19 0.31*
Inner Pampas
América 5.76 4.22* 944 604* 256 353* 355 344 14 78* 1.3 1.5* 0.11 0.11
Hills
Tandil 5.80 5.21* 1038 790* 227 310* 133 109* 9 13 3.8 4.1* 0.30 0.30
Coastal dunes
Pinamar 7.76 7.97 10861 12811 53 95 119 298* 103 159 nd nd nd nd
* Significant differences within sites, p < 0.05
215

At intermediate depths, the contribution of Ca to the exchange complex decreased
to < 50%, mainly as a result of gains in acidity. Below 50 cm, Ca was increasingly
replaced by Na (Figure 4). Exchangeable Na, barely detectable in grassland soils,
accumulated under forests to 16 to 24% of saturation.
Organic matter accumulation in forest stands was ⬃500 to 750 Mg/Ha for the
tree, litter, and organic soil pools (Table 3). Two-thirds of this accumulation was in
wood. Despite contributing only 15% of accumulated biomass, bark accounted for
60 to 90% of base cation sequestration by forests (data not shown). Ca was the
dominant cation in forest biomass, followed by K, Mn, Mg, and Na, with mean
sequestration across sites of 5.6, 0.72, 0.29, 0.28, and 0.19 Mg/Ha, respectively
(Table 4).
Figure 3. Soil properties under native grassland and eucalypt forests for the top meter at three sites
(mean + S.D,n=5foreach pair). A) Soil pH measured in 1:1 water-soil extract. B) Effective cation
exchange capacity (ECEC) obtained from the sum of NH
4
OAc-exchangeable base cations and KCl-
exchangeable acidity. C) Base saturation of the effective cation exchange complex. Values for the or-
ganic soil horizon (present only under forests) are indicated in the top of the panel. Asterisks indicate
significant differences between stands at p < 0.05.
216

With the exception of Na, cation pools in mineral soil displayed little or no
change overall for grasslands and forests in the upper meter (Table 4). The differ-
ences between total elemental pools in the forest and grassland ecosystems sug-
Figure 4. Relative composition of the exchange complex in the top meter of mineral soil under native
grasslands and Eucalyptus forests (n = 5). The values for NH
4
OAc-exchangeable base cations and KCl-
exchangeable acidity sum to 100%, with proportions calculated on a charge basis (mol
c
/mol
c
).
217

gested significant net ecosystem gains after forest establishment for Ca and Na at
Castelli, for Na at Guerrero, and for all base cations at América. The only signifi-
cant negative difference between forest and grassland ecosystem pools was ob-
served for Mn at Guerrero (Table 4). When mineral soil changes were examined by
depth intervals, important localized differences were revealed. All sites lost Ca from
5–50 cm depth and two gained Ca from 0–5 cm. Mn followed this trend with even
bigger changes on a relative basis (Table 4). The large difference of total Ca ob-
served from 5–50 cm depth between stands involved smaller pools of both ex-
changeable and non-exchangeable pools in the forests. Based on these differences,
non-exchangeable Ca losses after afforestation could have reached 20% from 5–10
cm depth (p < 0.001 within sites, p = 0.0097 across sites, data not shown). In con-
trast with the rest of the cations, Na pools were three times higher under forest
stands at all three sites (Table 4; p < 0.001 within sites, p = 0.021 across sites).
Estimates of proton releases from excess cation accumulation by trees were
higher than actual proton accumulation in soils. While the first ranged between 300
± 6 and 406 ± 33 kmol
c
.Ha
−1
across sites, the second amounted 44 ± 14 to 53 ± 21
kmol
c
.Ha
−1
. Taking into account the age of the forest stands (Table 1) and their
excess cation accumulation, mean annual rates of proton release by trees should
have ranged from 3.1 to 9.9 kmol
c
.Ha
−1
.yr
−1
, whereas exchangeable acidity accu-
mulation should have ranged 0.5 to 1.2 kmol
c
.Ha
−1
.yr
−1
. Deposition was more
acidic than throughfall at all sampling dates (p < 0.05), indicating no acid inputs
from the forest canopy. During the sampling period the mean pH of precipitation
collections ranged between 5.6 and 6.9, whereas that of throughfall ranged between
6.7 and 7.2. Acidity deposition was 0.013 kmol
c
.Ha
−1
. year
−1
. Our measurements
of precipitation and throughfall yielded the following estimates for annual elemen-
tal fluxes (precipitation and throughfall respectively, in kg. Ha
−1
.yr
−1
): 5.9 and 10.1
Table 3. Forest biomass estimates at three sites: Castelli, Guerrero, and América (mean + S.D.,n=5
for each site). Leaf, bark, and wood biomass were estimated from measurements of tree basal area and
height at each site and from allometric equations developed at Castelli. Basal area and tree height mea-
surements were taken in five 0.1 Ha plots in each stand. Litter and organic soil biomass were measured
in five 0.2 × 0.5 m plots.
Biomass (Mg/Ha)
Castelli Guerrero América
Trees
leaves 7.2 (3.2) 6.5 (2.5) 6.8 (3.6)
bark 116.9 (28.0) 86.5 (27.7) 71.0 (31.2)
wood 514.8 (123.6) 381.0 (121.9) 312.6 (137.5)
roots 97.8 (24.1) 72.4 (23.8) 59.4 (26.8)
Floor
litter 9.6 (2.2) 11.4 (0.8) 12.2 (1.5)
organic soil 8.0 (4.7) 26.2 (2.1) 28.9 (6.0)
TOTAL 754.3 (129.1) 584.0 (127.3) 490.8 (143.7)
218

Table 4. Differences in elemental pools between eucalypt plantation and native grassland stands in the Pampas. Negative values indicate losses (mean + s.d., n = 5).
Asterisks show significant differences (p < 0.05) for mineral soil changes and net balances. Where mineral soil pools differed significantly, the percentage difference
is shown. See Methods for details on the specific calculations.
Castelli Pool size differences (Mg/Ha)
Ca Mg K Na Mn
mean sd % mean sd % mean sd % mean sd % mean sd %
Forest 5.9 0.1 0.27 0.02 0.66 0.02 0.20 0.02 0.29 0.01
Trees 5.6 0.1 0.24 0.02 0.62 0.02 0.18 0.02 0.26 0.01
Floor 0.5 0.0 0.06 0.00 0.08 0.00 0.02 0.00 0.06 0.00
Mineral soil −2.3 1.8 2.14 4.32 2.69 3.92 10.02 2.20* 192 −0.76 0.46
0–5 cm 1.1 0.5* 49 0.05 0.02* 5 −0.08 0.03 0.08 0.02* 63 0.33 0.08* 126
5–50 cm −4.5 1.9* −25 0.40 0.25 0.46 0.31 1.31 0.25* 82 −0.50 0.15* −24
50–100 cm 1.1 1.7 1.69 4.48 2.31 4.00 8.63 2.04* 246 −0.60 0.47
BALANCE 3.7 1.8* 2.44 4.32 3.39 3.92 10.22 2.20* −0.45 0.46
Guerrero
Forest 4.6 0.1 0.25 0.02 0.57 0.02 0.14 0.01 0.26 0.01
Trees 4.1 0.1 0.19 0.02 0.50 0.02 0.12 0.01 0.19 0.01
Floor 0.6 0.0 0.06 0.00 0.07 0.00 0.03 0.00 0.07 0.00
Mineral soil −3.9 4.5 1.20 5.41 −0.85 4.63 15.13 2.12* 263 −1.14 0.50* −21
0–5 cm −0.3 0.2 −0.08 0.04* −7 −0.19 0.06* −14 0.08 0.01* 49 −0.02 0.05
5–50 cm −7.1 1.3* −34 −4.16 2.28* −27 −5.43 2.28* −30 2.04 0.69* 95 −1.37 0.34* −58
50–100 cm 3.5 4.1 5.44 4.25 4.78 3.42 13.01 1.78* 378 0.25 0.38
219

Table 4. Continued
Castelli Pool size differences (Mg/Ha)
Ca Mg K Na Mn
mean sd % mean sd % mean sd % mean sd % mean sd %
BALANCE 0.8 4.5 1.45 5.41 −0.28 4.63 15.27 2.12* −0.88 0.50*
América
Forest 6.4 0.6 0.32 0.04 0.92 0.11 0.23 0.06 0.29 0.03
Trees 5.6 0.6 0.24 0.04 0.84 0.11 0.21 0.06 0.16 0.03
Floor 0.7 0.0 0.08 0.00 0.08 0.00 0.02 0.00 0.13 0.00
Mineral soil −0.2 1.7 3.77 0.62* 10 2.97 0.53* 9 11.34 1.35* 184 −0.43 0.11* −11
0–5 cm 0.2 0.1* 9 0.06 0.03* 5 −0.04 0.03 0.04 0.01* 17 0.12 0.04* 72
5–50 cm −1.9 1.0* −9 1.30 0.40* 9 0.54 0.28* 4 2.41 0.41* 96 −0.45 0.07* −25
50–100 cm 1.5 0.8 2.40 0.30* 12 2.47 0.42* 15 8.89 1.00* 262 −0.09 0.06
BALANCE 6.2 1.8* 4.08 0.62* 3.89 0.54* 11.56 1.35* −0.14 0.11
220

for Ca, 1.5 and 3.1 for Mg, 4.9 and 27.6 for K, and 4.4 and 8.9 for Na. Forest floor
leachates obtained in the lab had a mean pH of 6.2 (SD = 0.4) indicating a low
contribution of protons by litter and organic soil.
The chemistry of Cynodon dactylon shoots differed significantly in grassland and
forest stands (Table 5). C. dactylon plants from the forest had almost triple the Mn
concentrations of those in grasslands (p < 0.05 within sites, p = 0.033 across sites),
with Ca and Mg showing a similar but smaller change (Table 5; p < 0.1 within
sites, p < 0.05 across sites).
Discussion
Soils under Eucalyptus plantations in the Pampas had substantially higher acidity
than native grassland soils. Ten to 100 years after tree establishment, soils that de-
veloped originally under herbaceous vegetation were 1 pH unit more acidic on av-
erage and had lower exchangeable Ca in the surface soil (Table 2). This pattern
was repeated at nine sites across soils typical of the region (Table 2). The differ-
ences in soil acidity that we observed in the Pampas matched observations in af-
forested grasslands of Africa and New Zealand (Davis and Lang 1991; Musto 1991;
Davis 1995; Alfredsson et al. 1998), although the intensity of forest/grassland dif-
ferences was greater in the Pampas (see also Amiotti et al. (2000)). Mirroring our
results, a recent review of vegetation transitions in the Amazon basin revealed large
pH and exchangeable Ca increases in Ultisols and Oxisols after the shift from na-
tive or secondary forest to pastures (McGrath et al. 2001; Krishnaswamy and Rich-
ter 2002).
Table 5. Chemistry of Cynodon dactylon plants growing in grassland and forest stands. Values corre-
spond to non-reproductive shoots and show the mean for grassland and forest stands (n=5ineach
stand) at six sites: Castelli, Guerrero, Dolores, Pereyra, Chascomús, and América. Percentage change
from grassland to forest plants is indicated where differences are significant.
Dry weight concentration (mg/kg)
Grassland Forest Difference
N 17796 16894
P 1735 1503
Ca 2975 3862* 30%
Mg 1375 1971* 43%
K 10965 11354
Na 199 183
Mn 70 201* 187%
S 2500 3087
Al 151 107
Fe 233 177
Si 102 78
221

Soil acidification following afforestation can potentially be explained by three
complementary mechanisms: 1) organic acids inputs, 2) increased soil respiration,
and 3) cation sequestration and redistribution (Figure 1). The vertical patterns of
soil acidification together with the estimated elemental balances and measurements
of aboveground acidity suggest that the last mechanism dominates in the Pampas
(Table 6). Vertical patterns of pH and exchangeable acidity indicated maximum
acidification at intermediate soil depths (10–35 cm), as well as losses of exchange-
able Ca (Figure 3). These vertical patterns suggest that organic acid inputs are un-
likely to have caused the acidification observed; they would have entered the soil
predominantly from the surface, causing maximum acidification in the top soil
(Figure 1), and the throughfall and organic soil leacheates were not acidic. Affor-
ested stands showed pH values well below 5, suggesting that carbonic acid, if im-
portant, could not be the only agent of acidification since it would not be dissoci-
ated below that pH.
The differences of base cation pools between forest and grassland stands sug-
gested that no net ecosystem losses occurred after afforestation (Table 4), and hence
base cation sequestration and redistribution, rather than increased soil respiration
and leaching, was the dominant mechanism of acidification in afforested plots
(Table 6). An important redistribution took place within mineral soil from interme-
diate depths to the surface soil (Table 4). This vertical redistribution, particularly
large for Ca, most likely reflects the intense cycling (uptake and release) of base
cations by trees (Jobbágy and Jackson 2001). In the past, excess cation absorption
and sequestration by trees has been proposed as a transient cause of acidification in
forests, unless harvesting occurred (Nilsson et al. 1982). In the long term, cation
Table 6. Synthesis of major results for the predicted imprint of soil acidification and the mechanisms
proposed in this paper (see Table 1).
H
+
inputs from
throughfall & forest
floor
Vertical location of
acidification peak
Base cation pools,
grassland/forest
difference
Ca higher in forest
Mg no differences
Castelli No H + inputs 10–35 cm K no differences
Na higher in forest
Ca no differences
Guerrero no data 10–35 cm Mg no differences
K no differences
Na higher in forest
Ca higher in forest
América no data 10–20 cm Mg no differences
K no differences
Na higher in forest
222

recycling was expected to balance uptake at some point, neutralizing any early
acidification (Nilsson et al. 1982). We suggest that even when a long-term equili-
bration is achieved, internal redistribution of cations from intermediate depths to
the surface mineral soil could cause the transfer of charge and sustained localized
acidification, with clear implications for mineral weathering in these layers. It is
important to highlight that the substantial redistribution of Ca in afforested soils
was not observed for Mg, suggesting that a ⬙cation-specific⬙process such as root
uptake, rather than a less specific mechanism like acid leaching, was operating.
In the Southeastern US, acidification accompanied an aggrading pine plantation
on a previously limed and cultivated Ultisol, but acidification peaked in the surface
soil and the system had significant Ca losses from leaching (Richter et al. 1994;
Markewitz et al. 1998). A particularly interesting case is a cultivated pasture in
Australia (formerly a forest soil) that was then reforested with Pinus radiata and
Quercus suber (Noble et al. 1999). Soil pH under the oak stand showed the largest
decrease and peaked at intermediate depths, whereas soils under the pine stand had
smaller pH changes observed only in the surface soil (Noble et al. 1999). We sug-
gest that these differences in acidification could result from shifts in the mechanism
of acidification between forest types.
Additional mechanisms of acidification associated with nitrogen cycling may
operate in afforested ecosystems but do not appear to be dominant in our systems.
One possible source of acidity is enhanced nitrification coupled with net nitrate
losses (Schlesinger 1997). Although feasible in our system, we believe that this
mechanism is unlikely, given that afforested stands retained or gained rather than
lost N, as indicated by the higher N contents of the top 20 cm of the soil in five
afforested stands compared with their grassland pairs (Table 2). Increases in the
ammonium/nitrate ratio of plant uptake after afforestation could be a source of
acidity if the forest acts as a N sink (Richter (1986), see also Arnold (1992)). We
made coarse estimates of net cation sequestration by tree stands assuming that all
N was taken up either as ammonium or as nitrate, and our figures of potential pro-
ton release remained positive and large in all study sites under both scenarios (2 to
6 kmol
c
.Ha
−1
.yr
−1
assuming nitrate uptake and 4–9 kmol
c
.Ha
−1
.yr
−1
assuming am-
monium uptake).
Soils under eucalypt plantations at our Pampas sites acidified with the same in-
tensity as soils from heavily industrialized areas affected by acid rain (Andrews et
al. 1996). The annual gain of exchangeable acidity that we estimated, 0.5 to 1.2
kmol
c
.Ha
−1
.yr
−1
, is attributable solely to the influence of tree plantations; acid rain
is effectively absent in the region, as shown both from our deposition and through-
fall data and in other studies (e.g., Lavado (1983) and Bouwman et al. (1997)).
Estimates of proton releases in plantations because of net cation sequestration ex-
ceeded observed exchangeable proton gains in the soil by 300 kmol
c
.Ha
−1
. These
extra charges could have been consumed by weathering reactions. Assuming that
feldspar, an abundant mineral in these loess soils (Teruggi 1957), was weathered to
kaolinite, < 4% of the soil mass of a 20 cm-thick layer would need to be affected in
a 50-year period to explain this proton consumption.
223

Soil chemical changes associated with afforestation had important feedbacks on
plants. Soil acidification increases the mobility and bioavailability of Mn, poten-
tially even leading to Mn toxicity (Gambrell 1996; Marschner 1995). We observed
substantial mobilization of Mn under afforested stands (Tables 4 and 5), with re-
duced Mn at intermediate depths in the total mineral pool and higher Mn concen-
trations in the mineral surface soil and in the leaves of the bioindicator grass C.
dactylon. This redistribution may have been caused by enhanced mineral release
and mobility of Mn due to acidification and by the accumulation of Mn in the sur-
face soil due to its intense uptake and cycling by Eucalypts (mean concentration in
fully expanded leaves was 1420 mg/kg). Mn toxicity could potentially affect fur-
ther agricultural use of afforested land, as suggested by Mn concentrations in Cy-
nodon leaves, which in our forest stands reached levels that are considered toxic
for other grass species (Edwards and Asher 1982).
The exchange capacity of soils is a critical aspect of soil fertility that may drive
future modifications in afforested soils of the Pampas. Soils under eucalypt planta-
tions showed a decline in effective exchange capacity, particularly at the oldest for-
est stand in Guerrero (Figure 3). A decline in the variable charge component of
organic matter due to acidification is a likely cause of this trend together with min-
eralogical alterations (i.e., weathering) of clays. Effective cation exchange capacity
declines could trigger irreversible nutrient losses in the future.
We suggest that the acidification described here could be expected under most
broadleaf tree species that, like Eucalypts, cycle large amounts of Ca. Acidification
should be even higher under commercial tree plantations experiencing regular bio-
mass (and cation) removals, compared to stands, which in most cases were never
harvested. The largest decreases in pH and Ca in our study occurred at the only site
that had been clear-cut in the past (Table 2 – Pila). In contrast with tree-dominated
systems, grasslands should not suffer significant Ca losses, even under intense for-
age harvesting and removal situations, because of their limited use of soil Ca.
Afforestated stands had substantially higher exchangeable Na and more alkaline
soil in deep layers (Figures 3 and 4) (see Parfitt et al. (1997)). The potential sources
of this Na were atmospheric deposition, rock weathering, and groundwater. Atmo-
spheric inputs are an unlikely source since the annual rates of deposition that we
recorded explained < 10% of the accumulated Na. In addition, Na:Cl ratios of deep
soil were four times higher than those of atmospheric inputs (p < 0.001; data not
shown), suggesting either a different Na source or substantial Cl leaching. Rock
weathering, enhanced by acidification, could potentially increase exchangeable Na
pools, but not total Na pools as we observed (Table 4). We believe that groundwa-
ter was the most likely source of the Na accumulated in our afforested stands. Na
accumulation is frequent in lowland soils and groundwater of the Pampas as a re-
sult of poor regional drainage (Tricart 1973; Lavado 1983; Bui et al. 1998). The
pathway of Na movement from groundwater to the upland soils studied here is still
uncertain. Physical transport resulting from enhanced capillary rise, decreased
leaching under tree plantations, and root uptake from deeper soil layers are plau-
sible mechanisms that deserve further study. Eucalypt leaves in Castelli, Guerrero,
and América showed high Na concetrations (> 1000 mg/kg, data not shown).
224

Whether this increased Na uptake is a cause or a consequence of soil Na accumu-
lation is uncertain, but potentially negative consequences of Na accumulation on
forest production should be explored. Groundwater could also be the source for the
apparent net Ca and Mg gains in afforested stands (Table 4).
Tree planting in the Pampas has been considered an aesthetically and even ⬙ethi-
cally⬙correct practice since the beginning of European settlement, with any poten-
tial negative consequences rarely acknowledged (Sarmiento (1855) and Acosta
(1873); but see Panario (1991)). The growing need for wood products in Argentina
and the prospective global market for C sequestration both suggest that expansion
of plantations will continue (Canadell et al. 2000; Wright et al. 2000). Furthermore,
tree invasions (e.g., Gleditsia triacanthos, Melia azedarach) are increasing in the
region (Ghersa et al. 2001; Mazia et al. 2001). The magnitude of soil alterations
that we observed within half a century of tree establishment emphasizes the role of
vegetation type as a dynamic factor of soil formation and suggests large potential
feedbacks of vegetation change on soil fertility and biogeochemistry.
Acknowledgements
We wish to thank DD Richter, D Binkley, WH Schlesinger, JF Reynolds, R Oren,
and MR Aguiar for their insightful comments and suggestions for this work. Field
and laboratory assistance from JC Villardi, P Gundel, N Trillo, J Rotundo, P Roset,
S Barreiro, A Iorio, and CW Cook is deeply appreciated. Farm owners and man-
agers R Lewin, D Gómez, A Gianini, S Zaldúa, K Carreras, O Santander, and J
García Cuerva provided access to plantations and helpful information. EGJ was
supported by Consejo Nacional de Investigaciones Científicas y Técnicas
(CONICET, Argentina – Beca Externa), the Forest History Society of America, and
grants from NSF (Dissertation Enhancement INT 0089494) and Inter American In-
stitute for Global Change (SGP 004). RBJ was supported by a NSF CAREER grant
(DEB 97-33333), the Inter American Institute for Global Change Research, and the
Andrew W. Mellon Foundation. This research contributes to the Global Change and
Terrestrial Ecosystems (GCTE) core project of the International Geosphere Bio-
sphere Programme (IGBP).
References
Aber J.D. and Melillo J.M. 1991. Terrestrial Ecosystems. Saunders College Publishing, Philadelphia,
US.
Acosta M. 1873. Sericicultura y Silvicultura: Proyecto de Legislación. Anales de la Sociedad Rural Ar-
gentina 7: 239–241.
Alban D.H. 1982. Efffects of nutrient accumulation by aspen, spruce, and pine on soil properties. Soil
Science Society of America Journal 46: 853–861.
225

Alfredsson H., Condron L.M., Clarholm M. and Davis M.R. 1998. Changes in soil acidity and organic
matter following the establishment of conifers on former grassland in New Zealand. Forest Ecology
& Management 112: 245–252.
Amiotti N.M., Zalba P., Sanchez L.F. and Peinemann N. 2000. The impact of single trees on properties
of loess-derived grassland soils in Argentina. Ecology 81: 3283–3290.
Andrews J.A. and Schlesinger W.H. 2001. Soil CO
2
dynamics, acidification and chemical weathering in
a temperate forest with experimental CO
2
enrichment. Global Biogeochemical Cycles 15: 149–162.
Andrews J.E., Brimblecombe P., Jickells T.D. and Liss P.S. 1996. An Introduction to Environmental
Chemistry. Blackwell Science, Oxford, UK.
Arnold G. 1992. Soil acidification as caused by the nitrogen uptake patterns of Scots pine (Pinus sylves-
tris). Plant and Soil 142: 41–51.
Azara F. 1796. Diario de reconocimiento de las gurdias y fortines que guarecen la línea de la frontera de
Buenos Aires. CEPEDA 8
a
. Compilation of original documents. Treasure of the National Library,
Buenos Aires, Argentina.
Barlow N. 1933. Diary of the Voyage of H.M.S Beagle. Cambridge University Press, Cambridge, UK.
Binkley D. and Richter D.D. 1987. Nutrient cycles and H+ budgets of forest ecosystems. Advances in
Ecological Research 16: 1–51.
Bouwman A.F., Lee D.S., Asman W.A.H., Dentener F.J., VanderHoek K.W. and Olivier J.G. 1997. A
global high-resolution emission inventory for ammonia. Global Biogeochemical Cycles 11: 561–587.
Bravo D.A. 1967. Diccionario Quichua Santiagueño-castellano. Instituto Amigos del Libro Argentino,
Buenos Aires, Argentina.
Brimhall G.H., Chadwick O.A., Lewis C.J., Compston W., Williams I.S., Danti K.J. et al. 1991. Def-
formational mass transport and invasive processes in soil evolution. Science 255: 695–702.
Bui E.N., Krogh L., Lavado R.S., Nachtergaele F.O., Tóth T. and Fitzpatrick R.W. 1998. Distribution of
sodic soils: The world scene. In: Sumner M.E. and Naidu R. (eds), Sodic Soils: Distribution, Prop-
erties, Management and Environmental Consequences. Oxford University Press, New York, NY,
USA, pp. 19–33.
Cairns M.A., Brown S., Helmer E.H. and Baumgardner G.A. 1997. Root biomass allocation in the
world’s upland forests. Oecologia 111: 1–11.
Canadell J.G., Mooney H.A., Baldocchi D.D., Berry J.A., Ehleringer J.R., Field C.B. et al. 2000. Car-
bon metabolism of the terrestrial biosphere: a multi-technique approach for improved understanding.
Ecosystems 3: 115–130.
Chen C.R., Condron L.M., Davisand M.R. and Sherlock R.R. 2000. Effects of afforestation on phos-
phorus dynamics and biological properties in a New Zealand grassland soil. Plant and Soil 220:
151–163.
Chen M. and Ma L.Q. 1998. Comparison of four USEPA digestion methods for trace metal analysis
using certified and Florida soils. J of Environmental Quality 27: 1294–1300.
Davis M.R. and Lang M.H. 1991. Increased nutrient availability in topsoils under conifers in the South
Island high country. New Zealand Journal of Forestry Science 2: 165–179.
Davis M.R. 1995. Influence of radiata pine seedlings on chemical properties of some New Zealand
montane grassland soils. Plant & Soil 176: 255–262.
Edwards D.G. and Asher C.J. 1982. Tolerance of crop and pasture species to manganese toxicity. In:
Scaife (ed.), Proceedings of the ninth plant nutrition colloquium, Warwick, England. Commonwealth
Agricultural Bureau, Farnham Royal, Bucks, UK, pp. 140–150.
Elliot E.T., Heil J.W., Kelly E.F. and Curtis H.C. 1999. Soil structural and other physical properties. In:
Robertson G.P., Coleman D.C., Bledsoe C.S. and Sollins P. (eds), Standard Soil Methods for long-
term Ecological Research. LTER, New York, NY, USA.
Finzi A.C., van Breemen N. and Canham C.D. 1998. Canopy tree-soil interactions within temperate
forests: species effects on soil carbon and nitrogen. Ecological Applications 8: 440–446.
Frankenberger W.T., Tabaitabai M.A., Adriano D.C. and Doner H.E. 1996. Bromine, Chlorine and Fluo-
rine. In: Sparks S.L., Page A.L., Helmke P.A., Loeppert R.H., Soltanpour P.N., Tabaitabai M.A. et
al. (eds), Methods of Soil Analysis – Part 3 Chemical Methods. Soil Science Society of America,
Madison, Wisconsin, USA, pp. 833–868.
226

Gambrell R.P. 1996. Manganese. In: Sparks S.L., Page A.L., Helmke P.A., Loeppert R.H., Soltanpour
P.N., Tabaitabai M.A. et al. (eds), Methods of Soil Analysis – Part 3 Chemical Methods. Soil Sci-
ence Society of America, Madison, Wisconsin, USA, pp. 665–682.
Garavaglia J.C. 1999. Pastores y labradores de Buenos Aires. Ediciones de la Flor, Buenos Aires, Ar-
gentina.
Geary T.F. 2001. Afforestation in Uruguay – Study of a changing landscape. J. Forestry 99: 35–39.
Geis J.W., Boggess W.R. and Alexander J.D. 1970. Early effects of forest vegetation and topographic
position on dark-colored, prairie-derived soils. Soil Science Society of America Proceedings 34:
105–111.
Ghersa C.M., de la Fuente E., Suárez S. and León R.J.C. 2001. Woody species invasion in the Rolling
Pampas grasslands, Argentina. Agriculture, Ecosystem and Environment 1792: 1–8.
Gill R.A., Polley H.W., Johnson H.B., Anderson L.J., Maherali H. and Jackson R.B. 2002. Nonlinear
grassland responses to past and future atmospheric CO
2
. Nature 417: 279–282.
Grieve I.C. 1980. Some contrasts in soil development between grassland and deciduous woodland sites.
J. of Soil Science 31: 137–145.
Hall A.J., Rebella C.M., Ghersa C.M. and Culot J.P. 1992. Field-crop systems of the Pampas. In: Pear-
son C.J. (ed.), Field Crop Ecosystems – Ecosystems of the World. Elsevier, Amsterdam, The Neth-
erlands, pp. 413–450.
IGM 1997. Carta de imágen satelitaria de la República Argentina 1:250000. Instituto Geográfico Mili-
tar, Buenos Aires, Argentina.
INDEC 1988. Censo nacional agropecuario 1988. Resultados generales, Instituto Nacional de Estadís-
ticas y Censos, Buenos Aires, Argentina.
INTA 1989. Mapa de suelos de la provincia de Buenos Aires – Escala 1:500.000. Instituto Nacional de
Tecnología Agropecuaria, Buenos Aires, Argentina.
Jackson R.B., Schenk H.J., Jobbágy E.G., Canadell J., Colello G.D., Dickinson R.E. et al. 2000. Be-
lowground consequences of vegetation change and their treatment in models. Ecol. Appl. 10: 470–
483.
Jackson R.B., Banner J.L., Jobbágy E.G., Pockman W.T. and Wall D.H. 2002. Ecosystem carbon loss
with woody plant invasion of grasslands. Nature 418: 623–626.
Jenny H. 1941. Factors of Soil Formation. McGraw-Hill, New York.
Jenny H. 1980. The Soil Resource, Origin and Behavior. Springer-Verlag, New York, US.
Jersak J., Amundson R. and Brimhall G. 1995. A mass balance analysis of podzolization: Examples
from the northeastern United States. Geoderma 66: 15–42.
Jobbágy E.G. and Jackson R.B. 2000. The vertical distribution of soil organic carbon and its relation to
climate and vegetation. Ecological Applications 10: 423–436.
Jobbágy E.G. and Jackson R.B. 2001. The distribution of soil nutrients with depth: Global patterns and
the imprint of plants. Biogeochemistry 53: 51–77.
Judd T.S., Attiwill P.M. and Adams M.A. 1996. Nutrient concentrations in Eucalyptus: A synthesis in
relation to differences between taxa, sites and components. In: Attiwill P.M. and Adams M.A. (eds),
Nutrition of Euclaypts. CSIRO publishing, Collingwood, Victoria.
Krishnaswamy J. and Richter D.D. 2002. Properties of advanced weathering-stage soils in tropical for-
ests and pastures. Soil Science Society of America Journal 66: 244–253.
Kuo S. 1996. Phosphorus. In: Sparks S.L., Page A.L., Helmke P.A., Loeppert R.H., Soltanpour P.N.,
Tabaitabai M.A. et al. (eds), Methods of Soil Analysis – Part 3 Chemical Methods. Soil Science
Society of America, Madison, Wisconsin, USA, pp. 869–920.
Lavado R.S. 1983. Evaluación de la relación entre composición química del agua de lluvia y el grado de
salinidad y sodicidad de distintos suelos. Revista Facultad de Agronomía 4: 135–139.
Likens G.E., Bormann F.H., Pierce R.S., Eaton J.S. and Johnson N.M. 1977. Biogeochemistry of a For-
ested Ecosystem. Springer-Verlag, New York, US.
MAGP 1998. Anuario estadístico agropecuario. Ministerio de Agricultura y Producción, Montevideo,
Uruguay.
227

Markewitz D., Richter D.D., Allenand H.L. and Urrego J.B. 1998. Three decades of observed soil acidi-
fication in the Calhoun Experimental Forest: has acid rain made a difference? Soil Science Society
of America Journal 62: 1428–1439.
Marschner H. 1995. Mineral Nutrition of Higher Plants. Academic Press, San Diego, US.
Mazia C.N., Chaneton E.J., Ghersa C.M. and León R.J.C. 2001. Limits to tree species invasion in pam-
pean grassland and forest plant communities. Oecologia 128: 594–602.
McGrath D.A., Smith C.K., Gholz H.L. and Oliveira F.D. 2001. Effects of land-use change on soil nu-
trient dynamics in Amazonia. Ecosystems 4: 625–645.
Musto J.W. 1991. Impacts of plantation forestry on various soil types. In: Institute for Commercial For-
estry Research Annual Report for 1991. ICFR, 37–39.
Nilsson S.I., Miller H.G. and Miller J.D. 1982. Forest growth as a possible cause of soil and water
acidification: an examination of the concepts. Oikos 39: 40–49.
Noble A.D., Little I.P. and Randall J. 1999. The influence of Pinus radiata, Quercus suber and improved
pasture on soil chemical properties. Australian J. of Soil Research 37: 509–526.
Panario D.H. 1991. Desarrollo forestal y medio ambiente en Uruguay. Serie investigaciones 85. Centro
Interdisciplinario de estudios sobre desarrollo, Montevideo, Uruguay.
Parfitt R.L., Percival H.J., Dahlgren R.A. and Hill L.F. 1997. Soil and solution chemistry under pasture
and radiata pine in New Zealand. Plant and Soil 191: 279–290.
Podesta G.P., Messina C.D., Grondona M.O. and Magrin G.O. 1999. Associations between grain crop
yields in central-eastern Argentina and EI Nino-Southern Oscillation. Journal of Applied Meteorol-
ogy 38: 1488–1498.
Quideau S.A. and Bockheim J.G. 1996. Vegetation and cropping effects on pedogenic processes in a
sandy prairie soil. Soil Science Society of America 60: 536–545.
Raich J.W. and Tufekcioglu A. 2000. Vegetation and soil respiration: Correlations and controls. Bio-
geochemistry 48: 71–90.
Richardson D.M. 1998. Forestry trees as invasive aliens. Conservation Biology 12: 18–26.
Richter D.D. 1986. Sources of acidity in some forested Udults. Soil Science Society of America Journal
50: 1584–1589.
Richter D.D. and Markewitz D. 1995. How deep is soil? BioScience 45: 600–609.
Richter D.D. and Markewitz D. 2001. Understanding Soil Change: Soil Sustainability Over Millennia,
Centuries and Decades. Cambridge University Press, NY, USA.
Richter D.D., Markewitz D., Wells C.G., Allen H.L., April R. and Heine P.R. 1994. Soil chemical change
during three decades in a loblolly pine ecosystem. Ecology 75: 1463–1473.
Robenhorst M.C. 1988. Determination of organic and carbonate carbon in calcareous soils using dry
combustion. Soils Science Society of America J. 52: 965–969.
Roberts D.W. 1987. A dynamical system perspective on vegetation theory. Vegetatio 69: 27–33.
Robertson G.P., Sollins P., Ellis B.G. and Lajtha K. 1999. Exchangeable ions, pH and cation exchange
capacity. In: Robertson G.P., Coleman D.C., Bledsoe C.S. and Sollins P. (eds), Standard Soil Meth-
ods for Long-Term Ecological Research. LTER, New York, NY, USA.
Rosas J.M. and Senillosa F. 1825. Diario de la comisión nombrada para establecer la nueva línea de la
frontera de Buenos Aires. CEPEDA 8
a
. Compilation of original documents. Treasure of the National
Library, Buenos Aires, Argentina.
Rudel T. and Ropel J. 1996. Regional patterns and historical trends in tropical deforestation, 1976–1990:
A qualitative comparative analysis. AMBIO 25: 160–166.
SAGPyA 2000. Primer inventario nacional de plantaciones forestales en macizo. Secretaría de Agricul-
tura, Ganadería y Pesca Argentina. Buenos Aires – Argentina. Revista Forestal 20: 1–9.
Sarmiento D.F. 1855. Plan Combinado de Educación Común, Silvicultura e Industria Pasturil. Docu-
ment in Archivo General de la Nación Argentina..
Saviozzi A., Levi-Minzi R., Cardelli R. and Riffaldi T. 2001. A comparison of soil quality in adjacent
cultivated, forest and native grassland soils. Plant & Soil 233: 251–259.
Schlesinger W.H. 1997. Biogeochemistry, an Analysis of Global Change. Academic Press, San Diego.
Senillosa F., Newton R. and Jurado J.M. 1878. Estancia San Juan. Anales de la Sociedad Rural Argen-
tina 12: 565–570.
228

Simonson R.M. 1959. Outline of a generalized theory of soil genesis. Soil Science Society of America
Proceedings 23: 152–156.
Soltanpour P.N., Jonson G.W., Workman S.M., Jones J.B. and Miller R.O. 1996. Inductively coupled
plasma emission spectrometry and inductively coupled plasma-mass spectroscopy. In: Sparks S.L.,
Page A.L., Helmke P.A., Loeppert R.H., Soltanpour P.N., Tabaitabai M.A. et al. (eds), Methods of
Soil Analysis – Part 3 Chemical Methods. Soil Science Society of America, Madison, Wisconsin,,
pp. 91–140.
Soriano A. 1991. Río de la Plata grasslands. In: Coupland R.T. (ed.), Natural Grasslands – Ecosystems
of the World. Elsevier, Amsterdam, The Netherlands, pp. 367–405.
Tate K.R., Scott N.A., Ross D.J., Parshotam A. and Claydon J.J. 2000. Plant effects on soil carbon stor-
age and turnover in a montane beech (Nothofagus) forest and adjacent tussock grassland in New
Zealand. Australian Journal of Soil Research 38: 685–698.
Tecchi R.A. 1983. Contenido de silicofitolitos en suelos del sector sudoriental de la Pampa Ondulada.
Ciencia del Suelo 1: 75–82.
Teruggi M. 1957. The nature and origin of Argentinean loess. Journal of Sedimentology and Petrology
27: 322–332.
Thomas G.W. 1996. Soil acidity and soil pH. In: Sparks S.L., Page A.L., Helmke P.A., Loeppert R.H.,
Soltanpour P.N., Tabaitabai M.A. et al. (eds), Methods of Soil Analysis – Part 3 Chemical Methods.
Soil Science Society of America, Madison, Wisconsin, USA, pp. 475–490.
Tricart J.L.F. 1973. Geomorfologia de la Pampa Deprimida. Colección científica XII. Instituto Nacional
de Tecnología Agropecuaria, Buenos Aires, Argentina.
Ugolini F.C., Dahlgren R.A., Shoji S. and Ito T. 1988. An example of andolization and podzolization as
revealed by soil solution studies, southern Hakkoda, northeastern Japan. Soil Science 145: 111–125.
USDA 1998. Keys to Soil Taxonomy. Unites States Department of Agriculture, Natural Resource Con-
servation Service, Washington, DC, USA.
Wright J.A., DiNicola A. and Gaitan E. 2000. Latin American forest plantations – Opportunities for
carbon sequestration, economic development and financial returns. J. Forestry 98: 20–23.
Zacharin R.F. 1996. Emigrant Eucalypts, Gum Trees as Exotics. Synergy Print, Cap Schank, Australia.
Zar J.H. 1984. Biostatistical Analysis. Englewoods Cliff, New Jersey, NJ, USA.
229