EFFECTS OF DEFORESTATION AND SUCCESSION ON SOIL AT
CERRO CUERICÍ, COSTA RICA
PAMELA M. COLLINS, BRIGETTE A. JONES, AND MAYDA H. NATHAN
Faculty Editor: David R. Peart
Abstract: The consequences of changes in land use are often apparent in the plant community, but
effects on soil characteristics are frequently overlooked. We attempted to quantify differences in
pH, conductivity, and the depths of the O and A horizons brought about by changes in land use
by analyzing soils of the Cerro Cuericí region of Costa Rica. Three common community types of
this region are pasture, secondary‐growth alder forest, and primary‐growth oak forest. We
predicted that pH would be lowest in the pasture and highest in the alder forest. We expected to
find the thinnest O and A horizons in the pasture, and deepest horizons in the oak forest. By
sampling soil pits in the three sites, we found that oak forest and alder forest pH were similarly
low, while pasture pH was significantly higher. The O horizons of the oak forest and alder forest
were significantly deeper than those of the pasture, and the A horizons of the alder forest and
pasture were significantly deeper than those of the oak forest. Our results suggest that the
creation of pasture through deforestation and burning can significantly alter the chemical and
physical properties of soil.
Key Words: Quercus copeyensis, Q. costaricensis, Alnus alcucuminata, oak, alder, pasture, land use,
pH, soil horizon depth, organic matter
INTRODUCTION content of the organic material in the
soil, and subsequent agriculture or
The characteristics of a given grazing removes some of what
soil are determined by parent remains (NRCS 1996).
material, climate, biota, topography, Until the mid‐20th century, the
and time (USDA 2000). When slopes of Cerro Cuericí, in the San
forested land is cleared, the José province of Costa Rica,
subsequent alteration of organic supported primary‐growth oak
inputs and sunlight and the loss of forest dominated by Quercus
chemical and physical processing of copeyensis and Q. costaricensis
rainfall by the vegetation can have a (Fagaceae). Beginning in the 1940’s,
significant impact on the chemical settlers cut sections of the forest and
and physical properties of the soil burned them to create pasture and
(Schulz 1960, Parker 1985, CONF farmland. Since that time, parts of
2003). That impact is further the deforested areas have
magnified when cleared land is regenerated, producing secondary
burned for agriculture. Burning stands dominated by nitrogen‐fixing
volatilizes much of the nutrient alders, Alnus alcucuminata
(Betulaceae). Other areas remain as predicted that oak soils would have
grazed land (Carlos Solano, personal a lower pH than alder soils.
communication). Nutrient turnover rates
To investigate how the should be highest in the pasture due
historical patterns of deforestation, to the rapid consumption of biomass
burning, grazing, and secondary by cattle, accessibility of nutrients in
growth in this system may have manure, and high decomposition
altered the chemical and physical rates under intense sunlight. Low
properties of the soil, we made three input and rapid uptake result in a
assumptions. First, each of the study low net deposition of organic matter
areas originally consisted of old‐ and nutrients, leading us to expect
growth oak forest with similar soil the O and A horizons to be thinnest
characteristics, including low pH. in the pasture. We predicted the O
Second, the land which is currently horizon to be thickest in the oak
pasture or secondary alder forest forest because of lower litter
was burned after the original oak temperatures under the dense
forest was cleared. Third, the parent canopy and high tannin
material and rainfall regime were concentrations, resulting in slow
constant across all sites. decomposition rates (Coley 1983).
We hypothesized that the Additionally, we predicted that
transition from oak forest to pasture, leaching rates would be low in the
and succession to alder forest, would oak forest due to rapid uptake of
cause significant changes in pH, precipitation by the standing
depth of O horizon (the organic biomass, thereby allowing dissolved
layer), and depth of A horizon (the organic material to accumulate in the
topsoil). To more fully document A horizon. Finally, we predicted that
differences among sites, we also the O and A horizons would be of
measured soil conductivity; intermediate thickness in the alder
however, we had no basis for forest. Decomposition rates should
predicting a particular trend in be faster than in the oak forest
conductivity. because of high concentrations of
Compared to oak and alder nitrogen and lower concentrations of
soils, we expected pasture soils to tannins in the leaf litter, encouraging
have a lower pH due to anticipated microbial activity (Coley 1983).
higher leaching rates in the pasture,
which would cause mineral cations METHODS
to be replaced with hydrogen ions
(Kricher 1997). Because oak Study system: Our study sites
frequently tends to dominate in were located on the property of the
highly acidic soils (Finzi 1998), we Estación Biológica Cuericí (elevation
2600 m), in San José province, Costa Statistical analysis: A
Rica. We selected three sites for our MANOVA was used to determine
study, one for each of three the differences in pH, conductivity,
successional stages: grass pasture, O horizon depth, and A horizon
alder forest, and oak forest. The depth among the three sites. We then
pasture site was immediately to the performed four one‐way ANOVAs
east of the biological station, the comparing each of the four variables
alder site was 60 m to the southwest, among the three sites. Finally, we
and the oak site was 200 m to the conducted Tukey‐Kramer Highly
northeast. To ensure that the soil Significant Differences tests (α =
column was not artificially 0.05) for pH, O horizon depth, and A
homogenized, we chose a pasture horizon depth, and we conducted a
site that had never been tilled. To power analysis for conductivity.
minimize the effects of erosion we
selected sites with similarly shallow RESULTS
Field methods: We sampled on We found significant
28‐29 January, 2007. At each site, we differences in the depth of the O
established a haphazardly‐oriented horizon, the depth of the A horizon,
35 m transect. We randomly selected pH, and conductivity among the
five distances along this transect, three sites (MANOVA: F8,6 = 53.11, P
and for each distance we randomly < 0.0001). There were significant
selected points between zero and differences among sites for pH
three meters from to the left or right (ANOVA: F14 = 74.97, P < 0.0001), O
of the transect. At the first, third, and horizon depth (ANOVA: F8 = 27.62, P
fifth points, we dug a soil pit deep = 0.0009), and A horizon depth
enough to determine depth of the O (ANOVA: F8 = 40.53, P = 0.0003), but
and A horizons. We qualitatively not for conductivity (ANOVA: F14 =
observed the density of the roots in 3.50, P = 0.063). Conductivity was
both the O and A horizons within marginally significantly higher in the
each soil pit. For each of the five pasture than in either the alder or the
points on each transect, we collected oak. A power analysis of
20 mL of A horizon soil from each of conductivity showed that only 3
three holes within a 1.5 m radius and more samples were needed to
combined them into a homogenized demonstrate significant differences.
three‐part composite sample for each pH was significantly higher in the
point. Using this composite soil pasture than in either forest type
sample, we measured pH and (Fig. 1). There was no significant
conductivity using 50 : 50 soil : water difference in pH between alder and
slurries. oak forests (Fig. 1). There was no
measurable O horizon in the pasture. horizon was 35% greater in the oak
Although the mean depth of the O forest than in the alder forest, this
difference was not statistically
significant (Fig. 1). The A horizons in
the pasture and the alder forest were
73% and 77% deeper, respectively,
than that of the oak forest (Fig. 1).
The A horizon was 19% deeper in
the pasture than in the alder forest;
this difference was not statistically
significant (Fig. 1). In the alder and
pasture sites, we noted that the
majority of roots were located in the
A horizon. However, in the oak
forest, fine roots dominated the
Human alteration of forest
and subsequent succession can have
substantial impacts on the physical
and chemical properties of soils. We
found significant differences in pH
and soil horizon depth between
primary oak forest, open pasture,
and secondary alder forest.
The trend observed in
conductivity, with higher
conductivity in the pasture
compared to the oak and alder
forests, may be biologically
significant. Although direct
correlations between conductivity,
Figure 1. Each panel shows comparisons of soil pH, and nutrient availability have
characteristics: (A) pH, (B) O horizon depth, and not been established, it was shown in
(C) A horizon depth across sites. Different letters
(a,b) indicate significantly different values Heinger et al. 2003 that conductivity
among sites. is one of several variables that
influence changes in nutrient
concentrations. This trend in
conductivity is consistent with result is an O horizon of comparable
differences found in the other soil depth in both forest types.
characters measured here, and the The unexpectedly thin A
potential for direct relationships horizon in the oak forest could be
between conductivity and factors related to our observation that fine
such as soil nutrients may be further roots and mycorrhizal hyphae were
investigated in future studies. abundant throughout the O horizon.
Contrary to our predictions, This finding indicates that a large
pH was highest in the pasture and portion of nutrients mineralized
not significantly different between from organic matter is taken up
the two forested sites. A potential directly by vegetation via
explanation for the high pH of the mycorrhizae and is rapidly
pasture lies in the fact that ash is incorporated into standing biomass.
alkaline. Therefore, the slash‐and‐ Therefore, it is likely that organic
burn agriculture practiced here material has no opportunity to
could have increased soil pH accumulate in the A horizon of the
(Kricher 1997). It is possible that the oak forest. In contrast, the alder
alder soils had lower pH than forest lacked the dense mat of fine
expected because the relatively rapid roots and mycorrhizal hyphae
decomposition of alder leaf litter present in the oak forest. We
may produce humic acids (Anon. assumed that the young volcanic
1975). soils of the region are relatively
As predicted, the pasture site phosphorous‐rich; thus, local plant
had the thinnest O horizon. communities are nitrogen‐limited.
However, the alder and oak O Since the alders fix nitrogen, they are
horizon depths were not less constrained by nutrient
significantly different from one availability in the litter layer.
another. The slow decomposition Therefore, alder leaf litter is free to
rate of oak leaves may be offset by move as fine organic material into
relatively low input rates of oak the A horizon.
litter, while the rapid decomposition Despite our expectation that
of alder leaves may be offset by the alder forest A horizon would be
relatively high input rates of alder thicker than that of the pasture, the
litter. Oak litter input rates may be two were not significantly different
lower than alder input rates because from each other and both were
later successional trees, such as oaks, deeper than that of the oak forest.
tend to have leaves with longer Since the alder forest was originally
lifespans than those of secondary pasture, we suggest that its deep A
growth trees, such as alders. The net horizon formed at that time,
probably because decomposition
rates in pasture are high and nutrient
uptake rates are low, contrary to our Coley, P.D., 1983. Herbivory and
original hypothesis. Transition to defensive characteristics of
forest may have simultaneously tree species in a lowland
increased both litter input and tropical forest. Ecological
nutrient uptake, while maintaining Monographs 53: 209‐233.
the alder A horizon depth at levels
characteristic of pasture. Finzi, A.C., C.D. Canham, and N.
Our results show that changes van Breemen. 1998. Canopy
in land use around the Cuericí tree‐soil interactions within
Biological Station have significantly temperate forests: species
altered the chemical and physical effects on pH and cations.
properties of the soil. Specifically, Ecological Applications
the creation of pasture through 8.2:447‐454.
deforestation and burning increased
soil pH, and subsequent biological Heiniger, R.W., R.G. McBride, and
activity deepened topsoil at the D.E. Clay. 2003. Electrical
expense of the organic layer. By conductivity and nutrient
determining the input rates of litter management. Agronomics J.
in alder and oak forests and the 95:508‐519.
nutrient limitations in these
communities, further studies could Anonymous, 1975.“Soils.” The New
substantiate the mechanisms we Encyclopædia Britannica, 15th
propose to explain the observed ed., vol. 16., pp. 1020‐1024.
differences between successional Chicago: Encyclopædia
stages. Although the patterns we Britannica, Inc.
observed may be specific to this oak‐
pasture‐alder system, they Kricher, J. 1997. A Neotropical
demonstrate that the impacts of Companion: An introduction
succession extend beyond the above‐ to the animals, plants, and
ground biotic community to the soil ecosystems of the New World
beneath. tropics. Princeton University
Press. Princeton, NJ.
USDA Natural Resources
Chattahoochee‐Oconee National Conservation Service. 1996.
Forests and Georgia Forestry Soil Quality Information Sheet
Commission. 2003. Forest Soil Quality Indicators: Organic
Ecosystem Study Unit for the Matter.
USDA Natural Resources USDA Natural Resources
Conservation Service. 1998. Conservation Service. 2000.
Soil Quality Information Sheet From the Surface Down: An
Soil Quality Indicators: pH. introduction to soil surveys
for agronomic use.