Oecologia (1998) 113:547±556
D.C. Garcia-Montiel á D. Binkley
Effect of Eucalyptus saligna and Albizia falcataria on soil processes
and nitrogen supply in Hawaii
Received: 28 February 1997 / Accepted: 22 September 1997
Abstract This study investigated the di€erences between
two fast-growing tropical tree species on soil N ¯ux and
availability. The work was conducted in the island of
Hawaii and included three sites located along the
Hamakua coast on the northeastern side of the island.
Within each site pure stands of Eucalyptus saligna
(Sm.) and the N2-®xing Albizia falcataria (L.) Fosberg
[=Paraserianthes falcataria (L.) Nielsen] were arranged
in four randomized complete blocks. For most of the
variables considered in this study, the species e€ects
were usually strong and the site e€ects were signi®cant in
some cases. After 13 years, soils under the Albizia stand
contained larger pools of total soil C and N, and larger
pools of inorganic N. Soil N availability indexed by ion
exchange resin bags revealed a strong pattern of species
and site e€ect on N availability; soils under Albizia
showed a 2.6±9 fold increase in N availability
(P < 0.01). Potential net rates of N transformation (10-
and 30-day aerobic incubations) were more than twice as
high for soils under the Albizia than under the Eucalytus
stands. Nitrogen mineralization during anaerobic incu-
bations were about 10% greater on Albizia soils. Gross
microbial mineralization and immobilization were de-
termined by estimating the gross rates of N transfor-
mation by the 15N-isotope pool dilution techniques.
Across species and sites, a strong linear positive rela-
tionship was obtained for gross immobilization and
gross mineralization indicating faster gross immobiliza-
tion as gross mineralization increases. Soil microbial
biomass on Albizia soils contained larger proportion of
it as bacterial biomass, while larger proportion of fungi
D.C. Garcia-Montiel (&)1 á D. Binkley
Department of Forest Sciences,
Colorado State University,
Fort Collins, CO 80523, USA
Present address:
1
The Ecosystems Center,
Marine Biological Laboratory,
Woods Hole, MA 02543, USA
e-mail: dgarcia@lupine.mbl.edu
Ó Springer-Verlag 1998
biomass comprised the microbial biomass under Euca-
lyptus soils. This study clearly showed that the presence
of Albizia increased total N pools and N supply to the
ecosystem. The overall e€ect on soil fertility will need to
be characterized by the e€ect of the N2-®xer on other
nutrients, especially the e€ect on phosphorus.
Key words Tropical tree plantations á Tree species
e€ect on soil á Silvicultural managed regeneration
N cycling á 15N isotope pool dilution
Introduction
The in¯uence of species on ecological processes is fun-
damental to the understanding of ecosystem function-
ing. Vegetation plays a major role in soil formation
(Jenny 1941) and species di€er in their e€ects on soil
characteristics and processes (Alban 1969; Pastor et al.
1984; Binkley and Sollins 1990; Matson 1990; Gower
and Son 1992; Vinton and Burke 1995; Rhoades and
Binkley 1996) in ways that in¯uence ecosystem func-
tioning and structure (Pastor et al. 1987). Soil organic
matter quality is one of the main soil properties that can
be directly in¯uenced by species through the incorpo-
ration of dead tissue with di€erent chemical composition
(Melillo et al. 1982; Pastor et al. 1984; Gower and Son
1992; Constantinides and Fownes 1993), which may al-
ter nutrient cycling throughout the ecosystem. The use
of fast-growing tropical tree plantations to rehabilitate
and restore many degraded tropical forest ecosystems is
based on this potential of di€erent species to in¯uence
soil fertility and ecological processes (Evans 1982;
Parrota 1992; Brown and Lugo 1994).
This study examined the e€ect of two fast-growing
tropical tree species on soil properties and N ¯ux, with
replication within each site and across three sites. The
species used for this study were Eucalyptus saligna
(Sm.) and the N2-®xing Albizia falcataria (L.) Fosberg
[=Paraserianthes falcataria (L.) Nielsen]. We wanted to
investigate the di€erences between these two fast-grow-
548
ing species in N cycling and availability. Due to higher
N and low polyphenol content of Albizia litter (Binkley
et al. 1992), we expected to see higher N turnover and
availability for soils from the Albizia stands. We also
examined the independent rates of gross microbial
mineralization and immobilization. The balance be-
tween these two rates explains the net availability of N in
the soil.
Materials and methods
Study site
The work was conducted in the island of Hawaii in pure stands of
Eucalyptus and Albizia. These plantations are part of a group of
plantations established in Hawaii in the late 1970s and early 1980s
on abandoned sugar cane ®elds. The main purpose of starting these
plantations was to explore the possible production of energy from
wood biomass. Tree production was strongly responsive to N
fertilization (Whitesell et al. 1992), and N2-®xing legumes were
examined as a possible solution to this N limitation problem.
Several pure and mix cropping experimental trials of Eucalyptus
and Albizia were established to examine species productivity and
changes in Eucalyptus productivity in the presence of a N2-®xer
(DeBell et al. 1989).
The study sites were located along the Hamakua coast on the
northeastern side of the island of Hawaii. The Chin Chuck stand
(site 1) is located at an elevation of 480 m (19°30¢N, 155°15¢W),
with slopes ranging from 0 to 10%. The other two sites, Upper
Kamae (site 2) and Lower Kamae (site 3), are about 1 km away, at
450±550 m of elevation. Site 2 is in a well-drained mid-slope po-
sition, and site 3 is in a wetter, lower-slope position. Average an-
nual precipitation is 4600 mm/year, well distributed through the
year, and annual temperature averages 21°C with little seasonal
variation. The plantation at site 1 was located in a former sugar
cane ®eld cropped for more than 50 years, and the stand was
planted in 1982 shortly after the last harvest of the sugar cane in
1980. Following plowing and herbicide treatment, three month-old
containerized seedlings were planted at a spacing of 2 m by 2 m
(DeBell et al. 1989) in 30 m by 30 m plots (pure Eucalyptus) or 15 m
by 15 m (pure Albizia). All seedlings were fertilized at planting and
4 and 8 months later with 115 g of N-P-K equivalent to a rate of
40, 18, and 33 kg/ha of N, P, K, respectively. Eucalyptus seedlings
were also fertilized at 12, 18, 24, and 36 months after planting
(Binkley et al. 1992). Sites 2 and 3 were abandoned because of poor
cane yields in 1960. Prior to planting in 1981, the vegetation present
after the 20 years of fallow was cleared, the soil plowed, and
resprouting vegetation herbicided. The plots at sites 2 and 3 were
also planted in 1982 from containerized seedlings at a spacing of
1.5 m by 1.5 m (plot size 12 m by 18 m). At these two sites both
Albizia and Eucalyptus seedlings were fertilized only at time of
planting with 110 and 50 kg/ha of N and P, respectively (Whitesell
et al. 1992). Both species grew well at all three sites.
The soils of the three study sites belong to the Akaka series, of
thixotropic isomesic Typic Hydrandepts. They are moderately well
drained soils developed from volcanic ash, low in Ca, Mg, K, and
P, and rich in Fe and Al (US Department of Agriculture 1973).
Bulk density, on an oven-dry basis, is very low, ranging from 0.21
to 0.44 kg/l in the top 0.7 m (Wada and Wada 1976; Rhoades and
Binkley 1996).
Methods
The pure species stands within each site were arranged in four
randomized complete blocks. Site 1 also contained mixed-species
plots that we did not sample. Each site contained four plots of
Albizia and four plots of Eucalyptus stands, giving 12 observations
for each species. The three locations gave us the opportunity to
examine the interactions between species and site e€ects on soil
processes. At the time of sampling all plantations were 12 years old
and the biomass of the stands was 270, 280, and 265 Mg/ha for
Albizia at site 1, 2, and 3, respectively, and 207, 375, and 435 Mg/ha
for Eucalyptus at site 1, 2, and 3, respectively (D. Binkley,
unpublished work).
At each plot, approximately 20 kg soil from 0±20 cm depth was
collected using a shovel. The soil samples were obtained from ®ve
separate points near the center of the plot and then composited into
one sample. Before sampling, the organic material above the
mineral soil was removed. After compositing, fresh soils were
sieved through a 8-mm sieve to remove roots and worms. All of the
sampling was done in one day during April 1994. Sub-samples were
stored in insulated coolers and sent to Colorado State University
within 4 days. Another set of soil sub-samples was sent to Oregon
State University for bacterial and fungal biomass determinations.
The remaining soil was used in bioassay experiments (Binkley
1996).
Soil pH and exchangeable cations
Soil pH was measured in a 1:1 ratio of fresh soil:deionized water,
1:1 ratio fresh soil: 0.01 M CaCl2, and a 2:1 ratio of fresh soil: 1 M
KCl, with a glass electrode. Exchangeable cations were removed
from 10 g fresh soil by shaking for 1 h with 100 ml of 1 M
NH4NO3. After settling for 24 h, soil extracts were ®ltered and
cation concentrations determined with a Jarrell Ash Model 975
Spectrometer (ICP). Values were corrected for moisture content
and reported on mass of oven dry soil. Cation exchange capacity
(CEC) was calculated as the sum of all extractable cations. Base
saturation was calculated as the sum of the basic cations Ca2+,
Mg2+, K+, and Na+ divided by the CEC.
Total carbon and nitrogen
A subsample obtained from each composited soil sample was
sieved through a 2-mm sieve, oven-dried at 105°C for 24 h, and
ground using a three-ball grinder. Total C and N were analyzed
with a LECO-1000 CHN analyzer. Final values are reported on
mass of oven dry soil. Soil C and N divergency values were cal-
culated as the di€erences in C and N content of soils under Albizia
at each site minus the corresponding C and N content of soils under
Eucalyptus and converted to a hectare basis using a bulk density of
0.4 kg/l.
Ion exchange resin bags
In October 1994, ion exchange resin bags were placed in the ®eld
ÿ
for 4 months to adsorb NH‡ 4 and NO3 from the soil solution
(Binkley and Matson 1983; Binkley and Hart 1989). Resin bags
were placed 5 cm below the surface of the mineral soil, every 2 m
along a single transect across each plot. Forest ¯oor over the bags
remained intact. The resin bags were prepared by adding 14 ml
of cation resin (Sybron IONAC c-251, H+ form) into a nylon
stocking, which was sealed by an electric glue gun. Next, 14 ml of
anion resin (Sybron IONAC ASB-IPOH) was added, and the
stocking sealed again. This design provided a 6 cm ´ 6 cm square
of cation resin adjacent to a 6 cm ´ 6 cm square of anion resin,
allowing both resins to sample the same site, while avoiding cross
contamination.
Net N mineralization
Aerobic incubations were performed in the laboratory for 10 and
30 days at 24°C. Two fresh 10-g subsamples of soil were added to
120-ml polyethylene cups and covered with plastic lids to minimize
water loss during incubation. Incubations were performed at ®eld
moisture, with water added weekly to compensate for water loss.