The Evergreen State College Institutional Repository

Hollis_SSMESThesis2008.pdf

Media

Part of Patterns of Mineral Element Retranslocation in Four Species of Trees in Monteverde, Costa Rica

extracted text (extracttext:extracted_text)

PATTERNS OF MINERAL ELEMENT RETRANSLOCATION IN FOUR
SPECIES OF TROPICAL MONTANE FOREST TREES IN MONTEVERDE,
COSTA RICA

by
Steven Scott Hollis

This thesis is submitted in partial fulfillment
of the requirements for the degree of
Master of Environmental Study
The Evergreen State College
December 2008

 2008 by Steven Scott Hollis. All rights reserved

Thesis for the Master of Environmental Study Degree
by
Steven Scott Hollis

has been approved for
The Evergreen State College
by

________________________
Nalini Nadkarni, PhD
Member of the Faculty
________________________
Anne McIntosh, MS
Forest Ecologist and PhD Student
University of Alberta
________________________
Dylan Fischer, PhD
Member of the Faculty

________________________
Date

ABSTRACT
PATTERNS OF MINERAL ELEMENT RETRANSLOCATION IN FOUR
SPECIES OF TROPICAL MONTANE FOREST TREES IN MONTEVERDE,
COSTA RICA
Steven Scott Hollis
Retranslocated nutrients (i.e. those moved out of plants prior to loss through
litterfall) comprise a significant fraction of available nutrients in trees but this
availability can differ by up to 50% or more among species. This high
interspecific variation in nutrient return indicates that tree species differentially
affect stand-level soil quality, as well as overall tree fitness. In this study, I
investigated the amounts and dynamics of nutrient cycling of six elements in the
foliage of four species of trees in the upper montane forest of Monteverde, Costa
Rica, and how much of that nutrient capital was transferred to the soil via
litterfall. For all species combined, as hypothesized, elements useful for tree
growth were retranslocated—with relatively high retranslocation rates of P and
K—while all other elements were transferred to the soil in the litterfall. At the
species level, retranslocation efficiency was highly variable over time. Significant
interspecific differences existed in N (p=0.001), P (p<0.001), K (p<0.001), Na
(p=0.002), but no significant differences existed for Ca and Al. The
retranslocation efficiencies of both N and P were in the lower range when
compared to other cloud forests. Relative to other cloud forests, the high nutrient
contents found in tree foliage in this study suggest nutrients at Monteverde may
not be as limiting as those in other cloud forests. The low rate of retranslocation
suggests these nutrients are being recycled into the soil freely. In the future, the
need for a more complete understanding of cloud forest nutrient cycling and
ecology across multiple scales, one in which broad generalizations can be rooted
in sufficient data, present great challenges for ecologists.

TABLE OF CONTENTS
1. Introduction..................................................................................................................... 1
2. Objectives and Hypotheses ............................................................................................. 6
3. Materials and Methods.................................................................................................... 7
4. Results........................................................................................................................... 12
5. Discussion ..................................................................................................................... 14
6. Conclusions................................................................................................................... 19
7. Bibliography ................................................................................................................. 21
8. Tables and Figures ........................................................................................................ 24

iv

LIST OF TABLES
Table 1: Leaf and litterfall content.................................................................................... 24
Table 2: Cross-study comparison...................................................................................... 25
Table 3: Mean retranslocation proficiency ....................................................................... 26
Table 4: Mean retranslocation efficiency ......................................................................... 27
Table 5: Mean seasonal retranslocation efficiency........................................................... 28

LIST OF FIGURES
Figure 1: Mean annual rainfall.......................................................................................... 29
Figure 2: Map of Costa Rica and Monteverde.................................................................. 30
Figure 3: Picture of primary forest plot ............................................................................ 31
Figure 4: Picture of secondary research plot..................................................................... 32
Figure 5: Mean mineral percent retranslocation efficiency .............................................. 33
Figure 6: Temporal changes in retranslocation efficiency................................................ 34
Figure 7: Seasonal retranslocation differences ................................................................. 35

v

ACKNOWLEDGMENTS
I thank Nalini Nadkarni for her patience, help and understanding, and for
collecting this data and making them available to me; Anne McIntosh and Dylan
Fischer for valuable comments and advice; Greg Stewart, Kathleen Saul and
Natalie Kopytko for statistical help; and the International Canopy Network for
resources. Financial help was provided by the Master of Environmental Studies
program at The Evergreen State College.

vi

1. INTRODUCTION
The pathways of mineral and nutrient fluxes through a forest are complex,
and their loss or recycling depends to a large extent on their nutritional value to
trees. Nutrients are captured and stored in the biomass of trees and used for
biological processes, then released from the tree to be remineralized in the
surrounding environment, and eventually used again as part of the
biogeochemical nutrient cycle (Jordan 1985). Resorbed nutrients (relative to
nutrients lost through litterfall) comprise a significant fraction of available
nutrients in forests, but this availability can differ by up to 50% or more among
species. This high intraspecific variation in nutrient return indicates that tree
species differentially affect stand-level soil quality, as well as overall tree fitness.
By documenting temporal patterns of nutrient return in litterfall, researchers can
begin to understand interactions between tree species and soils in an evolutionary
context (Binkley and Giardina 1998).
In this tree-soil nutrient feedback loop, litterfall is a major pathway for the
return of organic matter and nutrients from aerial portions of the forest to the soil
surface (Vitousek and Sanford 1986, Tanner et al. 1998a, Villela and Proctor
1999). Most organisms are directly or indirectly dependent on the nutrients
available in plant tissues deposited as litter, which account for 70% of all
aboveground litter (Killingbeck 1996). However, nutrients lost in litterfall do not
immediately benefit the plant that shed them because the litter must be
decomposed, and the nutrients contained in that litter must be remineralized to
become available for plant uptake. As these processes are not instantaneous, the

1

nutrients contained in litterfall must be, at least temporarily, considered as losses
to the plant population (Chapin 1980, Jordan 1985, Berendse and Aerts 1987,
Aerts and Berendse 1989, Aerts 1996, Killingbeck 1996).
Trees can compensate for these delays by a process called nutrient
retranslocation in which nutrients are withdrawn from leaves prior to abscission
and redeployed in developing tissues, such as leaves or reproductive structures
such as seeds, or stored for later use, thus extending the mean residence time of
nutrients in the plant (Wright and Westoby 2003). Retranslocation has been
characterized as one of the most important strategies used by trees to conserve
nutrients, which consequently influences competition, nutrient uptake, and
productivity (Killingbeck 1996). Stored nutrients that become retranslocated are
immediately available for plant use, allowing trees to react quickly to changes in
their environment and remain productive during nutrient-limiting periods
(Helmisaari 1992).
Retranslocation can be quantified in two distinct ways (Killingbeck 1996).
Retranslocation “proficiency” is the absolute level to which nutrient
concentrations are reduced in senesced leaves or litter. Retranslocation
“efficiency” is a different (but complementary) index of nutrient conservation that
measures the relative proportion of nutrients resorbed from senesced leaves.
Proficiency values are a more definitive and objective measure of the degree to
which evolution has acted to minimize nutrient loss, but efficiency values are
more useful to measure both nutrient demand (green-leaf nutrient content) and
nutrient withdrawal. Their complementary nature suggests that plants use

2

proficiency and efficiency in combination with one another for optimum nutrient
conservation (Killingbeck 1996).
Retranslocation efficiency at a particular growth phase is subject to many
variables, even within a species. In general, about 50% of leaf N and P can be
recycled via retranslocation (Aerts 1996, Killingbeck 1996), but the
retranslocation efficiency varies widely among species (Wright and Westoby
2003). For example, Aerts and Chapin (2000) documented that plants from all
habitats can retranslocate <5% to 80% of foliar N, and 0-95% of foliar P.
Evergreens have lower litter P concentrations than deciduous species, whereas
N2-fixers have higher litter N concentrations than non-N2-fixers (Killingbeck
1996). Yet, patterns of retranslocation and their governing factors may be similar
among species in the absence of interspecies competition for growth and crown
structure, which occurs in mixed species stands (Fife et al. 2008).
The extent to which an element can be retranslocated depends on its
physical properties, as well as its importance to the plant’s nutrient requirements.
Nitrogen, P, and K are relatively mobile, and are important for metabolism and
growth, so they retranslocate the most efficiently (Fife et al. 2008). Nitrogen and
P are considered limiting in most tropical environments—P is particularly limiting
in lowland tropical rain forests (Townsend et al. 2007) and N most limiting in
montane tropical forests (Tanner et al. 1998a), whereas K tends to be the most
mobile nutrient and easily available relative to concentrations in soils and rainfall.
Calcium is an important macronutrient for cell growth and functioning, but it is
not easily retranslocated because it becomes immobilized in the cell walls and

3

other structural components of leaves (Salisbury and Ross 1992, Fife et al. 2008).
Other elements such as Na are not biologically active, so they are usually
retranslocated in insignificant quantities. In contrast to the other minerals, some
elements such as Al are toxic to most plants, which either resist their uptake using
physiological mechanisms in their roots and leaves, or reduce toxicity by
sequestering them in vacuoles (Delhaize and Ryan 1995).
In addition to retranslocation, plants also use other mechanisms to increase
their overall nutrient-use efficiency (Wright and Westoby 2003). Nutrient-use
efficiency (NUE) is defined as the ratio of litterfall production per unit nutrient to
the litterfall nutrient content (Knops et al. 1997). Evergreen plants increase their
NUE by synthesizing leaves with a long leaf life-span, high leaf mass per area,
low leaf nutrient concentrations and low photosynthetic capacity (Aerts 1996,
Harrington et al. 2001, Wright and Westoby 2003, Fife et al. 2008).
Cloud forests are those that are frequently in direct contact with clouds (or
fog), and receive by condensation a significant amount of water in addition to
rainfall (Mai 1986). This frequent cloud cover influences the water balance as
well as radiation and other climatological, ecological, and soil parameters (Mai
1986). Soils in cloud forests have relatively slow rates of decomposition and
nutrient release (Jordan 1985). The slow rate of litter decomposition in cloud
forests, relative to rates in tropical lowland forests, also has been attributed to low
air and soil temperatures, lack of drying-rewetting cycles, a high degree of
sclerophylly, and waterlogged soils (Nadkarni and Matelson 1992). In cloud
forests, litter dynamics may be especially important because litterfall is the major

4

path of flux for macronutrients, and broadleaf evergreens are the dominant growth
form (Vitousek and Sanford 1986). Plants are expected to minimize nutrient
concentrations in leaves and maximize nutrient retranslocation from leaves before
senescence in response to the slow remineralization of nutrients (Veneklaas
1991).
Compared to lowland forests, however, cloud forest trees may support
foliage with higher nutrient concentrations due to the larger amounts of nutrients
that enter the system via interception of precipitation, especially in the form of
wind-blown mist (Nadkarni 1986, Vitousek and Sanford 1986, Clark et al. 1997),
and from canopy organic matter from the epiphyte community, which reaches
greatest abundance and diversity in cloud forests (Nadkarni 1986, Nadkarni et al.
2004).
To date, there has been extensive ecosystem-level research into forest
nutrient cycling in a few cloud forests (e.g., Luquillo National Forest, Puerto Rico
(Zou et al. 1995, Thompson et al. 2002)), but fewer ecosystem- or landscape-level
studies exist for upper montane cloud forests. The montane forests of
Monteverde, Costa Rica, have relatively high regional plant biodiversity within a
narrow elevational zonation of habitats along upper mountain slopes, which make
them especially interesting for nutrient cycling research. The area is characterized
by an abundance of mosses, epiphytes and tree trunk climbers (Haber 2000). Gaps
exist in our understanding of how cloud forest tree species react to the complex
balance of biotic and abiotic factors in this environment. For example, are trees in
this ecosystem nutrient-conservative as a result of nutrient deficiency? Or do they

5

show nutrient “leakage” that would indicate higher productivity on more fertile
soils? In a previous study at Monteverde, the amounts of N, P and K transferred in
litterfall to the forest floor were high compared to those reported from other
tropical montane forests (Nadkarni and Matelson 1992), and fell more closely
within the range reported for forests growing on alfisols and other moderately
fertile tropical soils (Vitousek and Sanford 1986). This suggests trees in
Monteverde are less conservative and have access to a larger nutrient pool than
other trees in cloud forests. If this is true, then the retranslocation efficiency of
macronutrients might be relatively lower than that in other cloud forests. This is
the first study to observe the interspecies retranslocation variation of dominant
trees in Monteverde that can be compared to other cloud forests. This comparison
may aid future researchers in understanding the relative importance of roles of
cloud forest tree species in nutrient cycling, soil development and tree-soil
interactions.

2. OBJECTIVES AND HYPOTHESES
In this study, I investigated the amounts and dynamics of nutrient cycling
in the foliage of four species of trees in the upper montane forest of Monteverde,
Costa Rica, and how much of that nutrient capital is transferred to the soil via
litterfall. The primary objectives of this study were to: i) investigate the
retranslocation dynamics of six foliar nutrients in three primary forest tree
species, and one secondary forest species; and ii) compare these data to
retranslocation studies in other ecosystems. I hypothesize that:

6

•

Foliage and litterfall nutrient content will be relatively higher than
those in other cloud forests;

•

the nutrients most critical for plant productivity (N, P, K) will be
retranslocated most efficiently;

•

elements not useful for plant productivity will not be
retranlsocated;

•

the elemental nutrient retranslocation efficiency will occur in the
order of K>P>N>Na>Al>Ca;

•

retranslocation efficiency will be lower than for other cloud
forests;

•

each tree species observed in this study will exhibit different
retranslocation efficiencies;

•

seasonality should have an effect on retranslocation efficiency, but
to variable degrees in each species.

3. MATERIALS AND METHODS
STUDY SITE—The montane cloud forests of Monteverde are located in the
Intertropical Convergence Zone (a zone of low pressure associated with intense
solar radiation and heating that follows the seasonal migration of the sun), at an
altitude (1460 m) where orographic precipitation and fog play major roles in
precipitation, nutrient deposition, and plant productivity. The tropical montane
forests at Monteverde are in a relatively narrow altitudinal zone with frequent
cloud cover during much of the year. Solar radiation and evapotranspiration are

7

limiting factors for growth, and precipitation is enhanced by canopy interception
of cloud water. Compared to trees in lower altitude tropical moist forests, trees in
the tropical montane cloud forests at Monteverde tend to be suppressed by wind
and frequent storms, with dense and relatively short, gnarled trunks, compact
crowns and small, thick leaves. Epiphytes are abundant and diverse, and soils are
frequently wet and highly organic (Nadkarni et al. 2000).
The climate of Monteverde is transitional between lowland and montane
sites in terms of ambient air temperature, and between the Caribbean and Pacific
sides of Costa Rica in incident solar radiation and amounts and seasonality of
precipitation (Clark et al. 2000) (Fig. 1). Mean annual temperature at Monteverde
(1460 m) is approximately 18.5o C, with a minimum of 9o C and a maximum of
27o C. From 1956 to 1995, mean annual precipitation depth at 1460 m was 2519
mm, but actual wet deposition is probably much higher because of the prevalence
of wind-driven mist and fog that occurs throughout the year (Nadkarni and
Matelson 1992).
The climate of Monteverde can be roughly divided into three seasons (Fig.
1). The misty-windy season (November-January) is characterized by advective
cloud cover and precipitation dominated by mist borne by the northeast
tradewinds. During the dry season (February-April), advective cloud water and
mist deposition occur, but measurable precipitation is low; bouts of strong
tradewinds abate at the end of this season. The wet season (May-October) is
characterized by low windspeeds and convective precipitation, much of which
originate in the Pacific-side lowlands.

8

Field research was conducted from June 19, 1990, to June 25, 1992, in the
Puntarenas Province of the Monteverde Cloud Forest Preserve (MCFP) of Costa
Rica (10 18′ N, 84 48′ W). This life zone occurs on a restricted area of the upper
Pacific slope, extending from the lower part of the MCFP above Monteverde to
Las Nubes, including the upper Río Negro and Río Chiquito drainages (Fig. 2).
The study area was in tropical lower montane moist forest (1480 m), described as
a leeward cloud forest (Lawton and Dryer 1980). The continually moist soils are
derived from volcanic rhyolites and are classified as Typic Dystrandept. These
volcanically derived soils are considered to be fairly fertile, recently deposited,
and share characteristics with other tropical cloud forests at similar elevations
(Vance and Nadkarni 1990).
In April 1987, a 4-ha study area was established within the primary forest
of the research area of the MCFP (Fig. 3). This forest is composed of trees that are
15-30 m in height, with a well-developed subcanopy. Tree density was 555 ha-1,
with a reverse-J diameter distribution. Tree species composition, density, basal
area and structural characteristics are reported in Nadkarni et al. (1995). The three
most common families of trees in this forest are Moraceae, Lauraceae, Sabiaceae,
respectively (Lawton and Dryer 1980). Species from these families include Ficus
tuerckheimii, Ocotea tonduzii and Meliosma vernicosa.
In 1989, a 1-ha research plot was established in the adjacent secondary
forest, which is also within the research area of the MCFP. In the early 1960s, the
area was cleared for cattle pasture, but was left to regrow because the area was too
cold and wet to be productive for agriculture. In this study, all trees in the plot

9

were measured, identified to species, and tagged. This forest is strongly
dominated (91%) by a single tree species (Conostegia oerstediana,
Melastomataceae), with a density of 1,124 trees ha-1 and a size class distribution
typical of early successional montane forests (Fig. 4). The forest supports a welldeveloped understory, with saplings of some of the primary forest trees from the
adjacent primary forest present. Additional details about study plots, precipitation,
structural characteristics and floristic composition are found in Nadkarni et al.
(1995) and in Nadkarni and Wheelwright (2000).
TREE SELECTION—Nine trees in the largest size class (>80 cm dbh) in the
primary study plot were randomly chosen for sampling live foliage of primary
forest trees. These included three trees for each of the species Ficus tuerckheimii,
Ocotea tonduzii and Meliosma vernicosa. Three trees in the secondary forest plot
(all Conostegia oerstediana) were also sampled. The sample trees were rigged
and climbed with single-rope mountain-climbing methods (Perry 1978, Nadkarni
1988).
FOLIAGE—Foliage

was collected for foliar analysis from the same 12

sample trees at intervals of 20 to 68 days, generally each month for year 1, and
every 2-3 months after that, for a total of 23 collections dates. For each of the four
species, three individual trees per species (N=3) were sampled at each harvest.
Live foliage was collected from at least three locations within accessible areas of
the crowns of the sample trees. Leaves that appeared to have emerged recently
were classified as immature, indicated by light green color and proximity to bud,

10

were not taken. Leaves were bagged separately by tree, dried, processed and
analyzed for nutrients as described below.
At the same time, a sample of litterfall leaves (15-35 leaves) was collected
from under each of these trees, from mesh surfaces (0.5 m x 0.5 m) installed on
the forest floor to differentiate new from old fallen leaves. Any visible frass or
detritus was removed from leaf surfaces. Foliage and litterfall were dried at 60o C
to constant weight (24 to 48 h), ground, and analyzed. The mean nutrient
concentration was calculated by averaging the nutrient concentrations of tissue
from the individual sample trees by species for each time interval.
NUTRIENT ANALYSIS—Foliage was analyzed for five macronutrients (K, P,
N, Na, Ca) and one micronutrient (Al). Nutrient analysis of plant and soil material
was done at Micro-Macro International, Inc. analytical laboratory (Athens, Ga.).
Plant tissue was prepared by weighing a 0.5 g sample into a porcelain crucible
and ashing at 500o C for four hours. The ash was dissolved in 30% aqua regia, and
then the digest assayed by ICP with Cd as an internal standard. A LECO Nitrogen
Determinator was used for N in plant tissue. A 0.25 g sample was placed into an
induction furnace, and the N was reduced to N2, which was measured by thermal
conductivity. Nutrient content is expressed in mean percent of total leaf dry
weight except for the micronutrient Al, which is expressed in parts per million
(ppm) because of its low foliar concentrations.
DATA ANALYSIS—Missed collections were treated as an “NA” and made
up >4% of the total calculations. Following Veneklaas (1991), retranslocation
efficiency was calculated by dividing the difference between elemental

11

concentrations of live foliage and litterfall by the concentration of that element in
live foliage:
Retranslocation Efficiency = [(live foliage-litterfall)/live foliage]* 100
Analysis focused on N, P and K due to their importance to foliage
production. Data were confirmed to be normally distributed with a Shapiro-Wilk
test. Statistical analyses were performed using R (Free Software Foundation,
Boston, Mass., version 2.6.2). Comparisons of retranslocation efficiency among
tree species, nutrients, and seasonality used a one-way ANOVA test for assessing
differences of means; pair-wise comparisons of species and nutrients used the
student t-test (with Bonferroni adjustment) for assessing differences of means;
correlation analysis used the Spearman Rank Test.

4. RESULTS
LIVE FOLIAGE AND LITTER CONTENT—significant differences existed in
foliar element concentrations between species in green-leaf and litterfall nutrient
content, particularly for Na, Ca, and Al (Table 1). The most limiting
macronutrients, N and P, however, showed less interspecies variability.
Phosphorus and Na showed low nutrient content variability. The high foliar Al
concentration of Conostegia is characteristic of the family Melastomataceae
(Delhaize and Ryan 1995, Jansen et al. 2002). As hypothesized, the values for live
foliage and leaf litter for both N and P were in the upper range when compared to
other cloud forests (Table 2).

12

RETRANSLOCATION PROFICIENCY—Species demonstrated varying degrees
of nutrient proficiency (Table 3). Meliosma had the lowest observed litterfall
nutrient content. Conostegia consistently showed the lowest mean proficiency,
although other species showed a lower absolute nutrient proficiency in litterfall
(Table 3).
RETRANSLOCATION EFFICICENCY—For all species combined, the order of
retranslocation efficiency was K>P>N>Na>Al>Ca (Fig. 5). As hypothesized,
elements useful for tree growth were retranslocated with relatively high rates for P
and K, whereas all other elements were transferred in the litterfall. Mean
retranslocation efficiency differed by element (p<0.001; Fig. 5). The high
standard errors for Na and Al (Fig. 5) resulted from isolated pulses of mineral
deposition in the litterfall, relative to green-leaf nutrient supply. These outliers
were included in the analysis because each data point is an average of multiple
leaf samples and not an individual sample susceptible to human input error. The
values for the retranslocation efficiency of both N and P were in the lower range
when compared to other cloud forests (Table 2).
Mean interspecific retranslocation efficiency was highly variable (Table
4). Among the macronutrients, K was the most variable. Variation was mostly
higher for the other elements. Ficus exhibited high Na variation; Al variation was
consistently high, except for Conostegia. Significant interspecific differences
existed in N (p=0.001), P (p<0.001), K (p<0.001), Na (p=0.002), but no
significant differences existed for Ca and Al.

13

Species also showed markedly different patterns of mean retranslocation
efficiency over time (Fig. 6). There were strong temporal correlations in N, P, and
K retranslocation efficiency, with a significant relationship (Spearman r>0.5)
between N and P (Spearman r=0.64; p<0.001), and K and P (spearman r=0.80;
p<0.001), but not K and N (Spearman r=0.46; p<0.001).
The effect of seasonality on retranslocation efficiency was negligible
when all three samples were combined (Fig. 7) and at the species level (Table 5).
There were no significant seasonal differences (alpha<0.05) in retranslocation
efficiency when sampling was combined for N, P, and K, or among any of the tree
species in this study. However, retranslocation efficiency did increase from the
dry to wet seasons (N, +3.6%; P, +3.7%; K, +1.6%) with the exception of K,
which peaked in the misty season and was greater than any other seasonal
macronutrient retranslocation.

5. DISCUSSION
At the ecosystem level, this study showed that the minerals most needed
for tree nutrient requirements (N, P, K) are those that are most readily
retranslocated, while all other minerals are transferred from the tree at higher rates
via litterfall. This is because most soil nutrients taken up by trees are used in
annual production of foliage, which serves as a reservoir of reusable nutrients
(Fife et al. 2008). Potassium is the most mobile nutrient and was the most readily
retranslocated. Phosphorus was retranslocated almost as much as K, and

14

significantly more than N, perhaps indicating P is a limiting nutrient at this site.
For all other elements, no retranslocation was observed.
Many studies have suggested P is more limiting than N in tropical
environments (Vitousek 1982, 1984, Vitousek and Sanford 1986), but these
studies are generalizations based on lowland tropical forests growing on relatively
older, highly weathered clay soils, e.g., oxisols and ultisols (Tanner et al. 1998b).
In contrast, the montane soils at Monteverde are considered to be mostly alfisols
and recently deposited andisols (Vance and Nadkarni 1990). In such soils, N is
thought to limit net primary productivity due to an elevationally constrained lower
mineralization rate. In contrast, P availability is high as fresh minerals weather,
but decreases over time due to leaching, occlusion by secondary minerals, and the
formation of recalcitrant soil organic matter (Tanner et al. 1998b, Harrington et al.
2001). I found higher foliar N (a potential index of nutrient availability) than
several other sites, although the N:P ratio was not higher than other similar
tropical localities (Fassbender and Grimm 1981, Grubb and Edwards 1982,
Veneklaas 1991, Nadkarni and Matelson 1992); Table 2). The contradiction
between the high P retranslocation efficiency found in this study and potentially
high P content in the soil at Monteverde may indicate retranslocation efficiency is
not a perfect measure of soil fertility (Killingbeck 1996) and does not differ
according to site nutrient status (Delarco et al. 1991, Aerts 1996, Wright and
Westoby 2003), thus making it a poor indicator of nutrient availability.
Senesced-leaf nutrient concentrations (retranslocation proficiency) are a
much more accurate indicator of site fertility (Killingbeck 1996, Wright and

15

Westoby 2003). The relatively high values of N and P in this study suggest that
trees at this site are relatively nutrient-rich and have access to a larger nutrient
supply (partially from allochthonous sources) than other cloud forests (Nadkarni
and Matelson 1992, Nadkarni and Solano 2002). These trees are less nutrientconservative, and as such, nutrients are not held as tightly by the trees. This would
allow the nutrients to recirculate into the environment (Harrington et al. 2001,
Townsend et al. 2007).
Differences in species retranslocation may have a functional significance
that helps determine the present performance, and likely the future species
composition, of a community. There exists a two-way connection between the
species diversity present in a community and the interactions occurring among
those species (Hacker and Gaines 1997). The relatively low overall mean
retranslocation rates may indicate trees in this study were not as overall
productive as those in other comparable studies. Retranslocation has been linked
to enhancing tree productivity by providing nutrient supply to apical growing
points in shoot growth, rather than linked to nutrient supply (Nambiar and Fife
1991, Fife et al. 2008). Younger gap-colonizing tree species in rapid production
of biomass tend to have higher retranslocation rates than mature forests with
competitively suppressed trees (Nambiar and Fife 1991). Although tree age was
not measured in this study, trees in the primary plot were all >80 cm dbh and
considered mature. In contrast, retranslocation rates of N, P, and K in Conostegia
were much higher than those in the primary plot (Table 3). The quick-growing
Conostegia is Monteverde’s dominant gap-colonizing tree species at this altitude.

16

The Conostegia trees that were sampled were much younger (<50 years old) than
those in the primary plot, and presumably were not as suppressed by competition
and limited in productivity by age. This implies that retranslocation efficiency is a
better indicator of productivity than nutrient conservation efficiency and may
explain differences in site results.
In this study, high variation in leaf nutrient proficiency and retranslocation
efficiency over time and between species hindered the detection of uniform
temporal patterns of retranslocation. Nutrient retranslocation tended to be greatest
during the wet and misty seasons (Fig. 7), but no significant seasonal differences
were found when all three samples were combined. No seasonal signal is
qualitatively discernable, although there is an apparent coupling between N, P and
K (Fig. 6). The lack of a seasonal signal may have been due to the length of this
study and short-term variance in weather patterns. However, at least one review
study has found seasonal controls on N:P values differing by 25% between wet
and dry seasons (Townsend et al. 2007), and that study also concluded that the
most striking feature of the 150-tree species data set was high variation at the
species level. This study also showed a 25% difference (N:P=0.48 in wet season;
N:P=.36 in dry season) between seasons and high interspecific variation,
indicating that some seasonal effect may be present but was not statistically
significant.
The variance in retranslocation proficiency may be due to multiple
constraints in biochemical and biophysical processes during leaf aging, nutrient
transformation, and phloem loading (Hattenshwiler et al. 2008). The existence of

17

high interspecific variation in this study and others also suggest that tree species
use a wide range of nutrient-conservation strategies at the levels of green-leaf
functioning, plant nutrient acquisition, and nutrient retranslocation physiology,
and to overcome environmental constraints (Hattenshwiler et al. 2008). This
indicates nutrient retranslocation is but one adaptive strategy of nutrient
conservation and can be influenced by other species-specific adaptations that
conserve nutrients by other means. This contradicts commonly held
generalizations on plant nutrient economies based upon broad functional groups,
and indicates that ecosystem-scale selection is of minor relevance for evolution of
plant nutrient-use strategies (Hattenshwiler et al. 2008).
Cloud forests are complex in their nutrient cycling regimes and need
further study. Relative to other cloud forests, the high contents of nutrients in tree
foliage found in this study suggest nutrients at Monteverde may not be as limiting
as those in other cloud forests, and the low rate of retranslocation suggests these
nutrients are being recycled into the soil. However, the feedback loop between
nutrient deposition and site fertility is still not fully understood (Binkley and
Giardina 1998, Townsend et al. 2007). Interactions between trees and their soil
may increase tree fitness, may indirectly benefit the tree’s fitness, or may not
optimize a tree’s fitness at all. The broad generalizations that often characterize
tropical forests as N-rich, P-poor environments still rely upon relatively small
amounts of data and may mask critical variation in the extent and nature of
nutrient limitation at multiple scales. In the future, longer-term surveys of the

18

foliage nutrient content and their associated retranslocation patterns of multiple
species will be useful.
Nutrient cycling studies have many implications for understanding
nutrient dynamics. The observed high variation in litter chemistry implies a highly
heterogeneous litter input to the soil at small spatial scales (Hattenshwiler et al.
2008). The wide range of variation in litter (as well as differences in carbon
quality) may affect higher trophic levels of the decomposer community with
varying constraints, depending on the local species composition of the litter. In
the future, broad generalizations that are rooted in sufficient data will result in a
more complete understanding of cloud forest nutrient cycling across multiple
spatial scales.

6. CONCLUSIONS
In this study, the amounts of nutrients being retranslocated in tree species
in Monteverde were complex at multiple scales, but they did follow some
predictable patterns. As hypothesized, at the stand level, foliage and litterfall N
and P content were higher than those in other cloud forests, indicating
Monteverde is a relatively nutrient-rich cloud forest. Nutrients needed for growth
were retranslocated, while elements not useful for plant productivity were
deposited in litterfall. However, since Monteverde is relatively nutrient-rich, these
nutrients were not retranslocated as efficiently as in other cloud forests. At the
species level, retranslocation efficiency was highly variable, indicating litterfall
nutrient return is variable at small spatial scales. However, seasonality did not

19

have a great effect as hypothesized. This might be due to the shortness of the
study’s length or variation at the species level.

######################################

20

7. BIBLIOGRAPHY
Aerts, R. 1996. Nutrient resorption from senescing leaves of perennials: are there
general patterns? Journal of Ecology 84:597-608.
Aerts, R. and F. Berendse. 1989. Above-Ground Nutrient Turnover and Net
Primary Production of an Evergreen and a Deciduous Species in a
Heathland Ecosystem. The Journal of Ecology 77:343-356.
Aerts, R. and F. Chapin. 2000. The mineral nutrition of wild plants revisited: a reevaluation of processes and patterns. Advances in Ecological Research
30:1-67.
Berendse, F. and R. Aerts. 1987. Nitrogen-use-efficiency: a biologically
meaningful definition? Functional Ecology 1:293-296.
Binkley, D. and C. Giardina. 1998. Why do tree species affect soils? The warp
and woof of tree-soil interactions. Biogeochemistry 42:89-106.
Chapin, F. I. 1980. The mineral nutrition of wild plants. Annual Review of
Ecology and Systematics 11:233-260.
Clark, K., R. Lawton, and P. Butler. 2000. The Physical Environment. Pages 1538 in N. Nadkarni and N. Wheelwright, editors. Monteverde: Ecology and
conservation of a tropical cloud forest. Oxford University Press, New
York.
Clark, K., N. Nadkarni, D. Schaefer, and H. Gholz. 1997. Atmospheric deposition
and net retention of ions by the canopy in a tropical montane forest,
Monteverde, Costa Rica. Journal of Tropical Ecology 14:27-45.
Delarco, J., A. Escudero, and M. Garrido. 1991. Effects of site characteristics on
nitrogen retranslocation from senescing leaves. Ecology 72:701-708.
Delhaize, E. and P. Ryan. 1995. Aluminum toxicity and tolerance in plants. Plant
Physiology 107:315-321.
Fassbender, H. and V. Grimm. 1981. Ciclos bioquimicos en un ecosistema
forestal de los Andes Occidentales de Venezuela. II. Produccion y
descompsicion de los residuos vegetales. Turrialba 31:27-37.
Fife, D., E. Nambiar, and E. Saur. 2008. Retranslocation of foliar nutrients in
evergreen tree species planted in a Mediterranean environment. Tree
Physiology 28:187-196.
Grubb, P. and P. Edwards. 1982. Studies of mineral cycling in a montane rain
forest in New Guinea. III. The distribution of mineral elements in the
above-ground material. Journal of Ecology 70:623-648.
Haber, A. 2000. Plants and Vegetation. Pages 39-94 in N. Nadkarni and N.
Wheelwright, editors. Monteverde: Ecology and conservation of a tropical
cloud forest. Oxford University Press, New York.
Hacker, S. and S. Gaines. 1997. Some implications of direct positive interactions
for community species diversity. Ecology 78:1990-2003.
Harrington, R., J. Fownes, and P. Vitousek. 2001. Production and resource use
efficiencies in N- and P-limited tropical forests: a comparison of responses
to long-term fertilization. Ecosystems 4:646-657.

21

Hattenshwiler, S., B. Aeschlimann, M. Couteaux, J. Roy, and D. Bonal. 2008.
High variation in foliage and leaf litter chemistry among 45 tree species of
a neotropical rainforest community. New Phytologist 179:165-175.
Helmisaari, H. 1992. Nutrient retranslocation in three Pinus sylvestris stands.
Forest Ecology and Management 51:347-367.
Jansen, S., T. Watanabe, and E. Smets. 2002. Aluminium accumulation in leaves
of 127 species in Melastomataceae, with comments on the order Myrtales.
Annals of Botany 90:53-64.
Jordan, C. 1985. Nutrient cycling in tropical forest ecosystems. Wiley & Sons,
New York.
Killingbeck, K. 1996. Nutrients in senesced leaves: keys to the search for
potential resorption and resorption proficiency. Ecology 77:1716-1727.
Knops, J., W. Koenig, and T. Nash. 1997. On the relationship between nutrient
use efficiency and fertility in forest ecosystems. Oecologia 110:550-556.
Lawton, R. and V. Dryer. 1980. The vegetation of the Monteverde cloud forest
reserve. Brensia 18:101-116.
Mai, C. 1986. Cloud forests in the humid tropics: distribution and hydrologic
characteristics. International Symposium on Tropical Forest Hydrology
and Application. Tropical Agricultural Research and Training Center,
Turrialba, Costa Rica.
Nadkarni, N. 1986. The nutritional effects of epiphytes on host trees with special
reference to alteration of precipitation chemistry. Selbyana 9:44-51.
Nadkarni, N. 1988. Tropical rainforest ecology from a canopy perspective. Mem.
Cal. Acad. of Sciences 12:57-67.
Nadkarni, N., R. Lawton, K. Clark, T. Matelson, and D. Schaefer. 2000.
Ecosystem Ecology and Forest Dynamics. Pages 303-350 in N. Nadkarni
and N. Wheelwright, editors. Monteverde: Ecology and Conservation of a
Tropical Cloud Forest. Oxford University Press, New York.
Nadkarni, N. and T. Matelson. 1992. Biomass and nutrient dynamics of fine litter
of terrestrially rooted material in a neotropical montane forest, Costa Rica.
Biotropica 24:113-120.
Nadkarni, N., T. Matelson, and W. Haber. 1995. Structural characteristics and
floristic composition of a neotropical cloud forest, Monteverde, Costa
Rica. Journal of Tropical Ecology 11:481-495.
Nadkarni, N., D. Schaefer, T. Matelson, and R. Solano. 2004. Biomass and
nutrient pools of canopy and terrestrial components in a primary and a
secondary montane cloud forest, Costa Rica. Forest Ecology and
Management 198:223-236.
Nadkarni, N. and R. Solano. 2002. Potential effects of climate change on canopy
communities in a tropical cloud forest: an experimental approach.
Oecologia 131:580-586.
Nadkarni, N. and N. Wheelwright. 2000. Scope of past work. . Pages 11-13 in N.
Nadkarni and N. Wheelwright, editors. Monteverde: Ecology and
conservation of a tropical cloud forest. Oxford University Press, New
York.

22

Nambiar, E. and D. Fife. 1991. Nutrient retranslocation in temperate conifers.
Tree Physiology 9:185-207.
Perry, D. 1978. A method of access into the crowns of emergent and canopy trees.
. Biotropica 12:155-157.
Salisbury, F. and C. Ross. 1992. Plant Physiology. 4th edition. Wadsworth
Publishing Co.
Tanner, E., P. Vitousek, and E. Cuevas. 1998a. Experiemental investigation of
nutrient limitation of forest growht on wet tropical mountains. Ecology
79:10-22.
Tanner, E., P. Vitousek, and E. Cuevas. 1998b. Experimental investigation of
nutrient limitation of forest growth on wet tropical mountains. Ecology
79:10-22.
Thompson, J., N. Brokaw, J. Zimmerman, R. Waide, E. Everham, D. Lodge, C.
Taylor, D. Garcia-Montiel, and M. Fluet. 2002. Land use history,
environment and tree composition in a tropical forest. Ecological
Applications 12:1344-1363.
Townsend, A., C. Cleveland, G. Asner, and M. Bustamante. 2007. Controls over
foliar N:P ratios in tropical rain forests. Ecology 88:107-118.
Vance, E. and N. Nadkarni. 1990. Microbial biomass and activity in canopy of
organic matter and the forest floor of a tropical cloud forest. Soil Biol.
Biochem. 22:677-684.
Veneklaas, E. 1991. Litterfall and nutrient fluxes in two montane tropical rain
forests, Colombia. Journal of Tropical Ecology 7:319-336.
Villela, D. and J. Proctor. 1999. Litterfall mass, chemistry, and nutrient
retranslocation in a monodominant forest on maraca island, Roraima,
Brazil. Biotropica 31:198-211.
Vitousek, P. 1982. Nutrient cycling and nutrient use efficiency. American
Naturalist 119:553-572.
Vitousek, P. 1984. Litterfall, nutrient cycling and nutrient limitation in tropical
forests. Ecology 65:285-298.
Vitousek, P. and R. Sanford. 1986. Nutrient cycling in moist tropical forests. Ann.
Rev. Ecol. Syst. 17:137-167.
Wright, I. and M. Westoby. 2003. Nutrient concentration, resorption and lifespan:
leaf traits of Australian sclerophyll species. Functional Ecology 17:10-19.
Zou, X., C. Zucca, R. Waide, and W. McDowell. 1995. Long-term influence of
deforestation on tree species composition and litter dynamics of a tropical
rain forest in Puerto Rico. Forest Ecology and Management 78:147-157.

23

8. TABLES AND FIGURES

Table 1: Mean percent of total leaf dry weight (+/-SE) in live leaves and leaf
litter. Al is expressed in ppm because it is a micronutrient. Values were compared
by species (vertically) and different letters do differ at alpha=0.05.

N

P

live foliage
K

Na

Ca

Al

1.73
(0.09)a
1.81
(0.10)a

0.15
(0.01)a
0.12
(0.01)b

2.18
(0.14)a
0.97
(0.15)b

0.07
(0.01)a
0.38
(0.03)b

1.57
(0.10)a
0.77
(0.04)b

130.45
(24.46)a
65.73
(6.83)b

Ocotea

2.22
(0.03)b

0.12
(0.01)b

1.01
(0.05)b

0.27
(0.01)c

0.77
(0.04)b

83.74
(8.06)b

Conostegia

2.01
(0.06)c

0.14
(0.01)a

1.03
(0.05)b

0.14
(0.01)d

2.02
(0.06)c

7157.91
(227.06)c

Genre
Ficus
Meliosma

leaf litter
Genre

N

P

K

Na

Ca

Al

Ficus

1.65
(0.09)a

0.11
(0.01)a

1.27
(0.08)a

0.08
(0.01)a

1.88
(0.06)a

153.18
(35.80)a

1.73
(0.11)a
2.01
(0.04)b

0.11
(0.01)a
0.11
(<0.01)a

0.82
(0.15)b
0.86
(0.04)b

0.33
(0.03)b
0.25
(0.01)c

0.88
(0.03)b
0.94
(0.03)c

61.64
(12.36)b
65.30
(16.51)b

1.59
(0.05)a

0.08
(<0.01)b

0.60
(0.02)c

0.14
(0.01)d

2.13
(0.05)d

6106.26
(95.07)c

Meliosma
Ocotea
Conostegia

24

Table 2: Cross-study comparison of mean percentage of nutrients in green-leaf
and litterfall content for multiple cloud forests; a) Fassbender and Grimm 1981; b)
Grubb and Edwards 1982; c) Veneklaas 1991; d) Nadkarni and Matelson 1992; e)
the current study.
Site

Elements (%)
N

P
Live foliage Leaf Litter
0.08
0.06

Live foliage
1.74

Leaf Litter
1.20

b) Papua
New Guinea

1.32

1.30

0.09

c) Colombia

1.8

1.10

d) Costa
Rica

1.97

e) This study

1.94

a) Venezuela

25

Retranslocation
efficiency (%)
N
P
31

25

0.07

2

22

0.10

0.06

38

40

1.47

0.14

0.08

25

42

1.75

0.13

0.10

10

25

Table 3: Mean retranslocation proficiency for all sampling dates of tree species
(mean) and the lowest observed litterfall nutrient content (low). Standard errors of
the nutrient content mean are also provided. Quantities are mean percent of total
leaf dry weight.
N
Genre

P

K

Mean

Low

SE

Mean

Low

SE

Mean

Low

SE

Ficus

1.65

1.13

0.09

0.11

0.06

0.01

1.27

0.81

0.08

Meliosma

1.73

1.02

0.10

0.11

0.04

0.01

0.82

0.24

0.15

Ocotea

2.01

1.62

0.04

0.11

0.07

<0.01

0.86

0.40

0.04

Conostegia

1.59

1.11

0.05

0.08

0.04

<0.01

0.06

0.040

0.02

26

Table 4: Mean percent retranslocation efficiency by species (+/- SE) for all
sampling dates combined. Values were compared by species (vertically) and same
letters do not differ at alpha=0.05.

Genre

N

P

K

Na

Ca

Al

Ficus

4.15
(2.36)a

26.17
(2.22)a

40.30
(2.50)a

-50.66
(23.38)a

-21.62
(3.77)a

-70.63
(39.46)a

Meliosma

4.10
(2.80)a

12.96
(3.31)b

13.13
(5.50)b

11.23
(6.79)b

-16.79
(5.47)a

-9.44
(24.44)a

Ocotea

8.90
(2.08)a

10.52
(4.17)b

10.92
(5.44)b

8.61
(3.44)b

-28.12
(7.7)a

4.17
(27.07)a

Conostegia

19.52
(3.09)b

40.56
(4.06)c

37.74
(4.21)a

1.33
(4.32)c

-7.29
(4.00)a

13.24
(2.42)a

27

Table 5: Mean retranslocation efficiency (+/- SE) of nutrients grouped by season.
No significant seasonal differences of means were found among species.
N

P

K

W

7.71 (2.52)

28.92 (3.40)

38.33 (3.96)

M

-0.24 (7.07)

20.57 (4.82)

47.94 (2.75)

D

-0.04 (5.19)

24.39 (2.81)

39.14 (3.76)

W

9.09 (3.47)

12.72 (4.18)

9.77 (6.96)

M

-4.71 (11.47)

25.69 (13.75)

45.52 (16.30)

D

-2.32 (2.10)

7.12 (4.12)

4.23 (5.96)

W

9.39 (3.19)

13.49 (6.15)

14.07 (8.75)

M

14.19 (4.60)

16.54 (10.77)

17.87 (9.78)

D

4.30 (1.96)

0.05 (4.41)

-0.53 (5.12)

W

17.27 (4.52)

39.86 (6.24)

38.42 (6.30)

M

18.92 (5.32)

35.36 (8.04)

23.67 (9.98)

D

24.81 (6.01)

45.54 (6.38)

45.63 (3.70)

Genus

Ficus

Meliosma

Ocotea

Conostegia

28

Figure 1. Mean annual rainfall (mm) with standard error margin (SE) at
Monteverde Forest Preserve from 1956-1995. Seasons: misty (M), dry (D), wet,
(W). Data collected by John Campbell.

29

Figure 2: Map of Costa Rica (opposite page) and study site, Monteverde Cloud
Forest Reserve (MVCFR). Small circle indicates the field station of the Tropical
Science Center. Black square represents the 4-ha study site.

30

Figure 3: The primary forest research plot located in a tropical lower montane
forest (1480 m)

31

Figure 4: Conostegia oersediana in the secondary forest research plot located in
tropical lower montane forest.

32

Figure 5: Mean mineral percent retranslocation efficiency (+/- SE) for all tree
species and all sample dates combined. Values were compared and same letters
do not differ at alpha=0.05.

33

Figure 6: Temporal changes in mean retranslocation efficiency for Nitrogen,
Phosphorus, and Potassium for four species A) Ficus tuerckheimii, B) Meliosma
vernicosa, C) Ocotea tonduzii, D) Conostegia oerstediana; W=wet season,
M=misty-windy season, D=dry season.

34

Figure 7: Seasonal efficiency (+/-SE) for all species combined. No significant
seasonal differences of means were found among nutrients.

35