<%BANNER%>

Leaf Traits of Canopy Trees on a Precipitation Gradient in Panama: Integrating Plant Physiological Ecology and Ecosystem...


PAGE 1

LEAF TRAITS OF CANOPY TREES ON A PRECIPITATION GRADIENT IN PANAMA: INTEGRATING PLANT PHYSIOLOGICAL ECOLOGY AND ECOSYSTEM SCIENCE By LOUIS STEPHEN SANTIAGO A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2003

PAGE 2

ACKNOWLEDGMENTS Many people and organizations contributed to the work presented in this dissertation, and I am grateful for all of their help. I thank my committee members Steve Mulkey, Kaoru Kitajima, Jack Putz, Tim Martin and Joe Wright, and everyone in the Plant Ecology group for five years of stimulating interactions and discussions leading to the ideas presented here. I am grateful to Guillermo Goldstein, Rick Meinzer, Ted Schuur, Michelle Mack, Tom Kursar, Lissy Coley, Klaus Winter, Steve Hubbell, and Allen Herre for comments and discussions on this project at various stages of its development. Tim Jones, Dave Woodruff, Katy Balatero, Mirna Sarmaniego, Aurelio Virgo, Augustin Somoza, Elizabeth Osorio, Steve Davis, Eric Graham, Aurelio Virgo, Kate Moran and Sarah Bouchard helped with lab and field procedures. I thank Edwin Andrade, Jose Herrera, Oscard and Pitti for their patience and skill in operating the canopy crane. Rick Condit and Suzanne Lao of CTFS provided generous use of their tree database. Funding was provided by the U.S. Environmental Protection Agency, the Smithsonian Tropical Research Institute (STRI), the Florida-Georgia Alliance, the Mellon Foundation, and the University of Florida. Katia Silvera, the Silvera family and my parents provided support throughout the project. ii

PAGE 3

TABLE OF CONTENTS Page ACKNOWLEDGMENTS..................................................................................................ii LIST OF TABLES.............................................................................................................vi LIST OF FIGURES..........................................................................................................vii ABSTRACT.......................................................................................................................ix CHAPTER 1 INTRODUCTION........................................................................................................1 2 LEAF PHOTOSYNTHETIC TRAITS OF CANOPY TREES ALONG A PRECIPITATION GRADIENT IN LOWLAND TROPICAL FOREST....................3 Introduction...................................................................................................................3 Materials and Methods.................................................................................................5 Site Characterization and Species.........................................................................5 Photosynthesis.......................................................................................................7 Leaf Structure and Chemistry................................................................................8 Two-Site Comparisons..........................................................................................9 Data Analysis.......................................................................................................10 Results.........................................................................................................................10 Soil Water Potential.............................................................................................10 Photosynthesis.....................................................................................................10 Leaf Structure and Chemistry..............................................................................11 Two-Site Comparisons........................................................................................12 Discussion...................................................................................................................12 3 NUTRIENT CYCLING ON A PRECIPITATION GRADIENT IN LOWLAND TROPICAL FOREST.................................................................................................25 Introduction.................................................................................................................25 Materials and Methods...............................................................................................27 Study Site.............................................................................................................27 Foliar Chemistry..................................................................................................28 Litterfall Collection and Processing....................................................................28 Leaf Litter Chemistry..........................................................................................29 Nutrient Availability............................................................................................29 iii

PAGE 4

Extractable nitrogen and phosphorus...........................................................29 Nitrogen mineralization and nitrification.....................................................30 Ion exchange resins......................................................................................30 Soil nutrient pools........................................................................................31 Results.........................................................................................................................31 Foliar Chemistry..................................................................................................31 Litter Production and Chemistry.........................................................................31 Nutrient Availability............................................................................................32 Extractable nitrogen and phosphorus...........................................................32 Ion exchange resins......................................................................................33 Soil nutrient pools........................................................................................33 Discussion...................................................................................................................33 Foliar Chemistry..................................................................................................33 Litter Production and Chemistry.........................................................................34 Nutrient Availability............................................................................................35 Conclusions.........................................................................................................36 4 HYDRAULIC CONDUCTIVITY AND WOOD DENSITY SCALE WITH LEAF PHOTOSYNTHETIC TRAITS IN PANAMANIAN FORES CANOPY TREES....46 Introduction.................................................................................................................46 Materials and Methods...............................................................................................48 Study Site and Species.........................................................................................48 Gas Exchange Measurements..............................................................................49 Hydraulic Conductivity.......................................................................................49 Wood Density......................................................................................................50 Foliar Analysis.....................................................................................................50 Results.........................................................................................................................51 Discussion...................................................................................................................52 5 LEAF DECOMPOSITION IN A WET TROPICAL FOREST: LINKING LEAF TRAITS WITH NUTRIENT CYCLING...................................................................59 Introduction.................................................................................................................59 Materials and Methods...............................................................................................61 Study Site and Species.........................................................................................61 Litter Collection and Decomposition..................................................................61 Initial Litter Quality.............................................................................................62 Gas Exchange and Leaf Chemistry.....................................................................62 Results.........................................................................................................................64 Discussion...................................................................................................................65 Plant Growth Forms and Ecosystem Processes...................................................65 Litter Quality and Decomposition.......................................................................66 Photosynthesis and Decomposition.....................................................................66 iv

PAGE 5

6 TEST OF GAS EXCHANGE MEASUREMENTS ON EXCISED BRANCHES OF TEN TROPICAL TREE SPECIES: A TECHNICAL REPORT...............................77 Introduction.................................................................................................................77 Materials and Methods...............................................................................................79 Study Sites and Species.......................................................................................79 Gas Exchange Measurements..............................................................................80 Results and Discussion...............................................................................................82 Time Course Measurements................................................................................82 Response to Light................................................................................................83 Response to CO 2 ..................................................................................................85 Conclusions.........................................................................................................87 7 CONCLUSIONS........................................................................................................94 LIST OF REFERENCES...................................................................................................95 BIOGRAPHICAL SKETCH...........................................................................................106 v

PAGE 6

LIST OF TABLES Table page 2-1. Site characteristics for 1 ha census plots....................................................................16 2-2. Study species from four sites along precipitation gradient.........................................17 2-3. Correlation coefficients between leaf defensive characteristics.................................18 3-1. Characteristics of sites along precipitation gradient...................................................38 3-2. Percentage of N and C in canopy sun leaves..............................................................38 3-3. Litterfall rates separated by component......................................................................39 3-4. Soil chemistry to a depth of 10 cm.............................................................................40 3-5. Bulk soil chemistry from the surface 10 cm of soil....................................................41 3-6. Soil bulk density and water content............................................................................41 4-1. Area-based maximum photosynthetic rate (A area ).......................................................55 5-1. Summary of study species and leaf litter decomposition rate (k)...............................68 5-2. Leaf litter decomposition rates and initial litter quality.............................................69 5-3. Regressions of litter chemical parameters and leaf litter decomposition...................70 5-4. Regressions of leaf chemical and structural parameters.............................................70 5-5. Results of general linear model..................................................................................70 6-1. Light-saturated rate of net CO 2 assimilation..............................................................88 6-2. Repeated measures analysis of variance.....................................................................88 6-3. Means of model parameters........................................................................................89 vi

PAGE 7

LIST OF FIGURES Figure page 2-1. Seasonal variation in soil water potential ( soil )........................................................19 2-2. Relationship between mean annual precipitation and photosynthesis........................20 2-3. Relationship between mean annual precipitation and leaf N concentration...............21 2-4. Relationship between mean annual precipitation and lamina thickness.....................22 2-5. Maximum photosynthetic O 2 evolution rate (P mass )...................................................23 2-6. Seasonal variation in leaf water potential ( leaf )........................................................24 3-1. The difference between seasonal maximum and minimum litterfall rates.................42 3-2. Relationship between carbon isotope composition of soil and litter..........................43 3-3. Soil N mineralization rates for the top 10 cm.............................................................44 3-4. Soil N concentration as a function of (A) litter lignin and (B) litter lignin:N............45 4-1. Leaf photosynthetic rate per unit area........................................................................56 4-2. Leaf nitrogen per unit area (N area )...............................................................................57 4-3. Stem saturated water content......................................................................................58 5-1. Relationship between leaf litter decomposition rate (k) and initial litter lignin.........71 5-2. Relationship between leaf litter decomposition rate (k) and initial litter carbon........72 5-3. Relationship between leaf litter decomposition rate (k) and nitrogen........................73 5-4. Relationship between leaf litter decomposition rate (k) and specific leaf area..........74 5-5. Relationship between A mass specific leaf area and leaf N...........................................75 5-6. Schematic diagram depicting how photosynthesis is related to decomposition.........76 6-1. Percent of maximum photosynthesis..........................................................................90 vii

PAGE 8

6-2. Percent of maximum stomatal conductance...............................................................91 6-3. Representative curves of net CO 2 assimilation as a function of PFD.........................92 6-4. Representative curves of net CO 2 assimilation as a function of internal CO 2 ............93 viii

PAGE 9

Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy LEAF TRAITS OF CANOPY TREES ON A PRECIPITATION GRADIENT IN PANAMA: INTEGRATING PLANT PHYSIOLOGICAL ECOLOGY AND ECOSYSTEM SCIENCE By Louis Stephen Santiago August 2003 Chair: Stephen S. Mulkey Major Department: Botany There is increasing awareness in ecology of the importance of species effects on processes at the ecosystem scale. This dissertation uses a comparative approach involving many species along a precipitation gradient (1800-3500 mm/yr) in lowland Panama to understand how species traits vary among different plant communities, and how these traits feed back into ecosystem processes such as decomposition and soil nutrient availability. As precipitation increases from South to North across the Isthmus of Panama, there is a gradual change in canopy leaf traits from short-lived leaves with high photosynthetic rates in seasonally dry forest, to relatively long-lived leaves with lower photosynthetic rates and increased allocation to structural defense in wet forest. Increases in leaf litter lignin:N also accompany increases in precipitation, indicating a decrease in potential decomposability of leaf litter in wetter sites. Leaf litter lignin:N was negatively correlated with soil N mineralization rates, and positively correlated with total soil N pools indicating that slowly decomposing litter reduces mineralization, but conserves N ix

PAGE 10

in the soil organic matter matrix. Leaf litter lignin:N was the strongest litter quality predictor of decomposition at the one site where decomposition was studied. Decomposition was positively related to specific leaf area, leaf N concentration and photosynthetic rate per unit mass suggesting that these traits may be useful predictors of the effects of species on ecosystem processes. Photosynthetic rate per unit area and stomatal conductance were positively related to leaf specific hydraulic conductivity and negatively related to branch wood density indicating that leaf traits controlling gas exchange correlate with processes at the branch and whole plant levels of organization. Overall this dissertation provides evidence that many plant traits are correlated along a minimal number of axes, and that these traits can be used to predict the movement of matter and energy between plants and their environments. x

PAGE 11

CHAPTER 1 INTRODUCTION This dissertation is an attempt to understand how changes in plant community composition along a precipitation gradient feed back into nutrient cycling and ecosystem processes. This research area integrates plant physiology, community ecology, and ecosystem science. The ideas in this dissertation are also based on studies of how species respond to environmental factors, and the effects that species have on those factors (Cornelissen 1996; Hobbie 1992; Wardle et al. 1998). Several studies have demonstrated that many plant species traits are correlated, and that suites of traits appear to have evolved in response to specific environmental regimes vary along a minimal number of axes (Chapin 1980; Grime 1977; Tilman 1988). More recently, it has been shown that leaf physiological traits governing the carbon economy of the leaf are interrelated and reflect fundamental evolutionary tradeoffs and biochemical constraints (Reich et al. 1992). Together, these studies suggest that if we can link the effects of plant species on ecosystem processes with ecophysiological or life history characteristics, then we may augment our understanding of ecosystem functioning, by drawing upon evolutionary and ecological principles. In this dissertation, I take two main approaches to understanding how species traits feed back into ecosystem processes. The first is along a precipitation gradient in lowland Panama. As precipitation increases in the lowland tropics, there is a gradual change in canopy species from deciduous species in seasonally dry forest, to evergreen species in wet, aseasonal forest. Leaf turnover represents a major pathway of energy and 1

PAGE 12

2 matter between the plant and soil components of the ecosystem. Therefore, along this precipitation gradient, I present data on how patterns of canopy phenology are related to photosynthesis, and leaf life span. In turn, I relate leaf traits to indices of litter quality and discuss ways in which variation in species composition along this precipitation gradient can influence the cycling of nutrients and the size of soil nutrient pools. The second approach I take to understanding how species traits feed back into ecosystem processes is at one wet forest site along the precipitation gradient. At this site I present data on how leaf physiological traits that control carbon assimilation and water loss are related to hydraulic and biophysical characteristics at the branch and whole plant scale, and discuss how relationships between leaf and branch physiology reflect evolutionary tradeoffs. I also compare how plant growth forms vary in the litter quality and decomposition rates of their leaves, and whether leaf litter decomposition can be predicted from leaf physiological characteristics. Together, this collection of studies provides evidence that changes in plant community composition with precipitation in lowland Panama can influence nutrient cycling, and that leaf physiological characteristics provide information about the potential effects of species on ecosystem processes.

PAGE 13

CHAPTER 2 LEAF PHOTOSYNTHETIC TRAITS OF CANOPY TREES ALONG A PRECIPITATION GRADIENT IN LOWLAND TROPICAL FOREST Introduction Tropical forests often exhibit gradients of vegetation structure and species composition in relation to precipitation (Gentry 1988; Schimper 1898; Wright 1992). Forests on the dry side of this gradient commonly have a higher proportion of deciduous canopy species that lose their leaves during seasonal dry periods when low soil water availability may limit physiological activity. Deciduous leaves of tropical forest reduce whole-plant transpiration and respiration during drought; and often have higher rates of photosynthesis per unit leaf mass (P mass ) than evergreen species (Chabot and Hicks 1982; Eamus and Prior 2001; Prado and DeMoraes 1997; Reich et al. 1992). In contrast, evergreen leaves have lower P mass but exhibit a potentially longer photosynthetic season; and appear to reduce demand on soil nutrients required to replace leaves (Chabot and Hicks 1982; Cunningham et al. 1999; Vzquez and Givnish 1998). These respective costs and benefits partially explain why deciduous trees dominate seasonally dry tropical forest; why evergreen species dominate wet, aseasonal forest; and suggest that P mass of canopy species should decrease with increasing precipitation in the tropics. There are few data available for photosynthetic traits on broad precipitation gradients in lowland tropical forest. On a global scale, however, P mass and specific leaf area are reported to increase with increasing moisture availability (Niinemets 2001; Reich et al. 1999). The purpose of our study is to determine how patterns of canopy phenology are related to 3

PAGE 14

4 photosynthesis and leaf life span along a regional precipitation gradient in lowland Panama. Interest in leaf trait variation along climate gradients dates back to the time of Theophrastus and formed some of the earliest ecological works. Recent studies have focused on precipitation as a driver of resource availability with direct and indirect effects on plant processes. For example, plants of relatively dry environments in Australia exhibited more sclerified vasculature than did species in high rainfall sites, possibly reflecting adaptation to resist wilting and minimize cell damage when water availability is low (Cunningham et al. 1999). Relatively high leaf N per unit area in dry habitats in Australia may represent a mechanism by which plants capitalize on higher light availability in dry habitats (Cunningham et al. 1999; Mooney et al. 1978). Leaf N of montane forest species in Hawaii was also reported to decrease with increasing precipitation and was related to decreasing soil N availability with increasing precipitation (Schuur and Matson 2001), suggesting that evergreenness in wet tropical forest may be a response to relatively low nutrient availability (Monk 1966). Precipitation may therefore directly affect vegetation structure and community composition through constraints imposed by water deficit, or indirectly through effects on availability of light and/or nutrients (Schuur and Matson 2001). Most studies of leaf trait variation with climate have been conducted along gradients with a maximum precipitation below 2500 mm yr -1 (Cunningham et al. 1999; Mooney et al. 1978; Werger and Morris 1991). Other studies on precipitation gradients with maximum precipitation above 5000 mm yr -1 have focused on phenotypic changes within individual plant species (Austin and Vitousek 1998; Schuur and Matson 2001). In

PAGE 15

5 contrast, our study addresses variation in leaf traits caused by large changes in species community composition over short (10 km) distances (Condit et al. 2002). I measured leaf physiological and structural traits to understand how patterns of resource allocation to leaves are related to climate on this regional gradient, which lies at the high end of the global precipitation range (1800 to 3500 mm yr -1 ). Specifically I wanted to link the leaf functional traits of dominant species to shifts in community phenology and leaf longevity characteristics. Materials and Methods Site Characterization and Species Our study was conducted in lowland tropical forest along a precipitation gradient in the Panama Canal Watershed. Mean annual precipitation (MAP) across this part of the Panamanian Isthmus ranges from 1800 mm yr -1 on the Pacific Coast to 4000 mm yr -1 on the Caribbean Coast (Condit 1998). I selected four 1-ha lowland forest study plots established by the Center for Tropical Forest Science (CTFS) over a range of precipitation with minimal changes in altitude and therefore temperature (Table 2-1). All sites have a mean monthly precipitation of >100 mm during the wet season (between May and December) but dry season length (mean number of 30-day periods with <100 mm precipitation) varies between 129 days at the 1800 mm site and 67 days near the 3500 mm site (ACP 2002) Between December and May, the probability of a site receiving <100 mm of monthly precipitation varies from 80% at the 1800 mm site to 22% at the 3500 mm site (Paton and Wright 2003). Variation in rainfall during the study period (2000-2002) was within 15% of MAP at all sites. Variation in the amount and distribution of annual precipitation has the potential to influence light availability and relative humidity. Average daily photosynthetically active radiation (PAR) decreases

PAGE 16

6 with increasing precipitation from 32.3 mol m -2 at the 1800 mm site (Juan Posada, unpublished data); to 31.9 mol m -2 at Barro Colorado Island in the middle of the isthmus and to 31.4 mol m -2 at the 3100 mm site (Paton and Wright 2003). Mean relative humidity appears to be highest on the wet Caribbean coast (94.6% at the 3100 mm site); the 1800 mm site and Barro Colorado Island maintain values of 87.9 and 84.0%, respectively (Paton and Wright 2003). The soils in the Panama Canal forests are well-drained clays high in Ca, Mg and N; and low in K and P relative to other tropical soils (Dietrich et al. 1982; Kursar et al. 1995; Yavitt et al. 1993). All of the study sites are on volcanic substrate (except Ft. Sherman, which lies on sedimentary substrate). However, similarities in soil characteristics between sedimentary and volcanic substrates on Barro Colorado Island (which lies in the middle of the isthmus) suggest that in this area nutrient availability is determined more by weathering and nutrient cycling by vegetation than by parent material (Yavitt 2000). Soil water potential ( soil ) was measured with the filter paper technique (Deka et al. 1995) at six randomly selected locations in each 1-ha study plot, nine times over a 14-month period (February 2001 to March 2002). One 42.5 mm diameter filter paper (Whatman no. 42, batch #711492, Whatman, Kent, UK) was equilibrated for 6 days with a fresh soil sample taken from 15 to 20 cm depth. Then the gravimetric moisture content of the filter paper was used to predict water potential using the regression equation of Deka et al. (1995). Assuming that gravitational and solute potential are negligible, the resulting values represent soil matric potential. I also determined the gravimetric soil water content on a 5 g subsample by drying at 105C for 24 h.

PAGE 17

7 At each site, I mesaured the eight canopy tree species with the largest relative proportion of basal area. Species composition and phenological habit of the most common canopy trees change rapidly across this gradient, as species richness increases steeply with mean annual precipitation (Table 2-2) (Pyke et al. 2001). One recent study comparing beta-diversity (how species composition changes with distance) of tree communities between lowland forests in Ecuador and Peru found that distant forests (>1000 km) with similar climate shared a much larger proportion of the most common species than would be expected by chance (Pitman et al. 2001). In contrast, tree community composition varies substantially even over 10 to 20 km distances in Panama; and such variation appears to be regulated largely by climate variation (Condit et al. 2002). All study plots are located in mature forest (>500 yr), except the 1800 mm site, which is a forest of mixed age (70-100 yr). Measurements on the largest canopy trees at the 1800 mm site are comparable to other sites even though this forest is younger, because canopy composition was representative of mature moist and dry forest (Croat 1978). Photosynthesis At the 1800 and 3100 mm sites, canopy leaves were collected using canopy cranes maintained by the Smithsonian Tropical Research Institute. The youngest fully expanded mature leaves were cut at the petiole; and immediately sealed in a darkened humidified container. At the 2300 and 3500 mm sites, leaves were collected from the upper canopy using a shotgun and then treated the same. Canopy trees were defined as individuals with approximately 80% of the crown exposed to full sun. All leaves were collected before 1030 h the day after rainfall and were transported to the laboratory within 2.5 h where photosynthesis was measured immediately.

PAGE 18

8 Photosynthetic capacity was measured as the maximum rate of O 2 evolution with a Clark electrode (Model LD2, Hansatech, Norfolk, UK) (Delieu and Walker 1981) on 3 to 8 leaves of each of 2 to 4 individuals for a total of 15 leaves per species from a site. A 10-cm 2 leaf disk was placed in a sealed chamber containing 10% CO 2 and maintained at 29C. Leaves were first induced with 250 and 560 mol photosynthetic photon flux density (PPFD) before the maximum rate of O 2 evolution at 2000 mol was measured. Light was provided by a quartz-halogen lamp (Bjrkman and Demmig 1987) with attenuation achieved by inserting neutral density filters and verified with a quantum sensor (Model LI 190SB, Li-Cor Inc., Lincoln, NE). At each light level, a stable signal was usually obtained in 5 to 7 min. The chamber was flushed with 10% CO 2 for 2 min between changes in light intensity. The high concentration of CO 2 in the O 2 electrode chamber bypasses all stomatal and cuticular resistance, so that the measured maximum rate reflects the Rubisco-limited rate of photosynthesis; and thus is a good index of enzyme allocation to photosynthetic capacity. Maximum rates of photosynthetic O 2 evolution were correlated with measurements of CO 2 assimilation conducted with an infrared gas analyzer (Model 6400, Li-Cor, Inc.) on a subset of study species (r 2 =0.82; P<0.0001; n=12). Leaf Structure and Chemistry After photosynthetic measurements, I measured lamina thickness between primary and secondary veins with a digital caliper (Mitutoyo Inc., Japan). Leaf discs were dried overnight at 65C; weighed for determination of specific leaf area (SLA); and ball milled. All leaf discs from the same tree were pooled for chemical analysis. One composite sample from each tree was analyzed for N using an elemental analyzer (Model NCS

PAGE 19

9 2500, Carlo Erba, Milano, Italy). A separate subsample was analyzed for 15 N at the University of Georgia Institute of Ecology. Leaf carbon fraction analyses were performed on a single composite sample per species at a site using forest-product techniques (Ryan et al. 1989). Dried leaf samples were digested in a detergent solution by which soluble and nutritionally available cell contents were separated from neutral detergent fiber (NDF), which includes all cell wall constituents and is not immediately nutritionally available. A dilute acid detergent solution was then used to determine acid detergent fiber (ADF, lignocellulose) before lignin was separated from cellulose in 72% H 2 SO 4 Leaf toughness was measured in the field on freshly collected leaves with a penetrometer (Pesola, Switzerland), which measured the maximum force during punching through the lamina between primary and secondary veins at a steady, slow rate with a 1 mm diameter plunger. Such measurements are not equivalent to the material property of fracture toughness, which is a more theoretically relevant measure of toughness against herbivore action. However, leaf toughness was correlated with fracture toughness for species at the 1800 mm site (Kaoru Kitajima, unpublished data). Two-Site Comparisons Canopy access at the two crane sites allowed for ease in data collection of leaf water potential and leaf life span of study species. Leaf water potentials ( leaf ) were measured at predawn (0600 h) and midday (1200 h) for the study species at the 1800 and 3100 mm sites using a pressure chamber (PMS Instruments, Corvallis, Oregon, USA). At both sites, canopy cranes were used to collect three terminal shoots from two individuals per species from the upper canopy in July 2001, October 2001, and February 2002 representing the early and late wet season and dry season, respectively. Leaf life span was

PAGE 20

10 measured on marked sun leaves that were censused every 6 weeks for 5 years as part of a larger study on canopy leaf dynamics (S. Joseph Wright, unpublished data). Data Analysis Analyses of the statistical effect of climate on leaf traits were conducted with one-way general linear models, to test whether leaf traits varied significantly with precipitation. Analyses of the relationships between leaf traits were performed with one-way analyses of covariance to test for heterogeneity of means around regression slopes. Each analysis of covariance was conducted independently with P mass as the dependent variable; MAP as the grouping factor; and leaf N, SLA or leaf life span as the covariate. Mean leaf between sites was compared by repeated measure analysis of variance and profile analysis with site as a between-subject factor; and measurement date and time of day as multiple within-subject factors (von Ende 1993). All analyses were conducted using SAS version 6.12 (SAS 1985). Results Soil Water Potential Soil water potential reinforced precipitation data, demonstrating that the main difference between sites is the length and intensity of the dry season and not differences during the wet season (Figure 2-1). Furthermore, maximum gravimetric soil moisture, an index of soil water-holding capacity, increased with precipitation from a maximum of 49.4 to 89.1% of dry soil mass from the driest to wettest site. Photosynthesis Mean annual precipitation (MAP) explained 13% of the variation in area-based maximum photosynthetic rates (P area ) of the eight most abundant canopy tree species at each of the sites (Figure 2-2A). The P area showed a marginally significant decrease with

PAGE 21

11 increasing MAP. When photosynthesis was expressed on a mass basis (P mass ), MAP explained 42% of the variation and P mass decreased significantly with increasing precipitation (Figure 2-2B). There was no clear relationship between leaf N per unit area and MAP (r 2 =0.00, P=0.95). However, MAP explained 21% of the variation in leaf N per unit mass and 34% of the variation in 15 N, both of which decreased significantly with increasing precipitation (Figure 2-3A,B). Both photosynthesis and leaf N showed stronger correlations with precipitation when expressed on a mass basis than on an area basis because there was an increase in leaf thickness with increasing MAP (Figure 2-4A). In a related index of leaf structure, SLA decreased with increasing MAP (Figure 2-4B). Analysis of covariance revealed no heterogeneity of means around regression slopes caused by MAP in the relationship between P mass and leaf life span (F=0.23; P=0.58) or leaf N (F=0.51; P=0.68), and only a marginally significant heterogeneity of slope caused by MAP in the relationship between P mass and SLA (F=2.95; P=0.05). Therefore, species from all sites were regressed together in the same predictive relationships, regardless of site, although species from different sites tended to occupy different ranges of the relationship (Figure 2-5). The P mass was positively correlated with leaf N across all species and sites; and this predictive relationship improved from an r 2 of 0.56 to 0.78 by removing the statistical outlier (Figure 2-5A). The P mass and SLA were positively correlated across all species and sites; and were expressed as a log-linear function (Figure 2-5B). Leaf Structure and Chemistry Toughness of canopy leaves significantly increased with MAP and lamina thickness (Table 2-3). Fiber (NDF) and cellulose concentration were positively related to

PAGE 22

12 lamina thickness; thus the proportion of leaf cell-wall material increases as leaves increase in thickness. With increasing precipitation, I also noted trends toward reduced nutritional content and increased antiherbivore defense (fiber and lignin per unit leaf N) (Table 2-3) (Cunningham et al. 1999). Two-Site Comparisons Median leaf life spans were longer at the 3100 mm site than they were at the 1800 mm site (t=-6.25; P=0.003; df=4); and were negatively correlated with P mass (Figure 2-5C). Midday leaf water potential ( leaf ) was lower at the 1800 mm site than at the 3100 mm site during the early (F 1,14 =6.11; P<0.05) and late wet season (F 1,14 =3.17; P<0.1; Figure 2-6); but not during the dry season. There were no significant differences in predawn leaf among sites during any measurement period. Discussion As precipitation increases from South to North across the Isthmus of Panama, there is a gradual change in canopy leaf traits from short-lived leaves with high P mass in seasonally dry forest, to relatively long-lived leaves with lower P mass and increased allocation to structural defense in wet forest. Relatively high P mass and short leaf life spans of canopy leaves in seasonally dry forest may allow canopy species to take advantage of high light availability when water is available; and to minimize water loss and respiration costs during rainless periods. Longer leaf life spans exhibited by canopy trees in wetter forest appear to place constraints on P mass by necessitating increased allocation to structural defenses. Although substantial variation in leaf traits exist along this precipitation gradient, it is important to note that precipitation drives the expression of leaf traits through soil moisture and tree water status (Reich 1995; Reich and Borchert

PAGE 23

13 1984); and potentially by influencing light and nutrient availability, although these indirect effects are less documented (Schuur and Matson 2001). Our data show coupling between soil and seasonality in precipitation and show how variation in dry season intensity among study sites appears to be the greatest source of variation in soil water availability among sites (Figure 2-1). The range of soil values observed are consistent with other studies of lowland Neotropical and West African forest reporting upper soil values near 0 MPa during the wet season to .5 MPa or lower during dry periods (Goldstein et al. 1986; Holbrook et al. 1995; Veenendaal et al. 1996). Our values of predawn and midday leaf are also comparable to studies conducted in seasonally dry tropical forest in Venezuela, Australia, and Costa Rica (Borchert 1994; Eamus and Prior 2001; Goldstein et al. 1986; Medina and Francisco 1994; Sobrado 1986). Differences in midday leaf between the 1800 and 3100 mm sites during the wet season but not during the dry season (despite strong differences in soil in the dry season) indicate that differences in atmospheric water content and stomatal function can affect leaf independently of soil Additionally, two of the deciduous species at the 1800 mm site had no leaves during February 2002 and were therefore not measured; whereas the remaining species maintained leaf values between .0 and .5 MPa and thus minimized variation between the 1800 and 3100 mm sites, resulting in no statistical difference among the measured species. Sobrado (1986) and Medina and Francisco (1994) also found a stronger decrease in leaf in deciduous than in evergreen species during the dry season, suggesting that maintaining functional leaves during the dry season requires greater resistance to desiccation, possibly through allocation to deeper roots.

PAGE 24

14 Several studies have addressed the effects of precipitation on nutrient availability as a factor governing the expression of leaf traits at the community scale (Austin and Vitousek 1998; Givnish 2002; Schuur and Matson 2001). Evergreen vegetation tends to dominate nutrient-poor habitats (Monk 1966). Dominance of evergreen species in wet forest is consistent with the notion of reduced nutrient availability as precipitation increases. Both Schuur and Matson (2001) and Austin and Vitousek (1998) found the lowest foliar N concentrations at highest precipitation sites along precipitation gradients in Hawaiian montane forest. Decreasing leaf N per unit mass in lowland Panamanian forest is consistent with this line of reasoning. Decreasing foliar 15 N with increasing precipitation has been interpreted as signifying increasing N-limited conditions (Austin and Vitousek 1998; Schuur and Matson 2001; Shearer and Kohl 1986). However, our bulk soil 15 N did not match this pattern (Chapter 3). Thus decreasing foliar 15 N with increasing precipitation in lowland Panamanian canopy trees may reflect the fact that species with shorter leaf life spans retranslocate leaf N more frequently than evergreen species and that 15 N becomes enriched during re-assimilation of nitrate and leaf N re-metabolism (Evans 2001). Decreasing leaf N with increasing precipitation may also reflect diminishing returns for N allocation to canopy leaves if the decrease in light availability with increasing precipitation is sufficiently strong to drive such a pattern. A difference of 0.9 mol m -2 for average daily PAR exists across this precipitation gradient. It is unknown whether this amount could contribute to patterns of N allocation to canopy leaves. Reduced N allocation to leaves with lower light availability is consistent with the functional convergence hypothesis (which predicts that plants should only allocate

PAGE 25

15 resources to photosynthetic capacity that can be used despite constraints imposed by limiting resources such as light and water) (Field 1991). One species at the 1800 mm site, Luehea seemannii, increased rates of CO 2 assimilation, branch growth, and fruit production in response to experimental light enrichment; supporting the notion that canopy trees of lowland forest have the potential to acclimate to higher light availability (Graham et al. 2003). Therefore, decreasing light availability with increasing precipitation may contribute to relatively low photosynthetic rates in wet tropical forest. Our pattern of decreasing P mass and SLA with increasing moisture availability on a regional precipitation gradient is consistent with the notion of a bimodal distribution of leaf longevity (Chabot and Hicks 1982; Kikuzawa 1991). On a global scale, deciduousness is highest in mid-precipitation and mid-latitude ecosystems. Our relationships between P mass and leaf life span and between leaf N and SLA show slopes similar to those of global comparisons (Reich et al. 1992). This suggests that fundamental tradeoffs between leaf traits are constant; but at the high end of global precipitation range, where water availability may exceed biological demand for much of the annual cycle, P mass and SLA decrease with increasing precipitation, in contrast to the pattern of increasing P mass and SLA with increasing precipitation common in other biomes (Reich et al. 1999).

PAGE 26

16 Table 2-1. Site characteristics for 1 ha census plots in the Panama Canal Watershed. Basal area, species richness, and tree density represent species with stems >10 cm in diameter. Site Mean annual precipitation (mm yr -1 ) Elevation (m) Basal area (m 2 ha -1 ) Species richness (no. ha -1 ) Tree density (no. ha -1 ) Parque Metropolitano 1800 60 25.39 36 318 Pipeline Road 2300 210 26.59 95 560 Fort Sherman 3100 140 32.50 87 569 Santa Rita 3500 282 25.74 162 497

PAGE 27

17 Table 2-2. Study species from four sites along precipitation gradient in Panamanian lowland tropical forest, including phenological classification into deciduous (losing leaves for more than a few weeks), brevi-deciduous (losing leaves once per year and immediately flushing a new set) and evergreen. Classification based on field observations and the Flora of Barro Colorado Island (Croat 1978). Species Family Phenology 1800 mm Anacardium excelsum Anacardiaceae Brevi-deciduous Astronium graveolens Anacardiaceae Deciduous Calycophyllum candidissimum Rubiaceae Evergreen Chrysophyllum cainito Sapotaceae Evergreen Enterolobium cyclocarpum Fabaceae Deciduous Luehea seemannii Tiliaceae Brevi-deciduous Pseudobombax septenatum Bombacaeae Deciduous Spondias mombin Anacardiaceae Deciduous 2300 mm Poulsenia armata Moraceae Brevi-deciduous Pourouma bicolor Moraceae Evergreen Sterculia apetala Sterculiaceae Deciduous Tabebuia guayacan Bignoniaceae Deciduous Tapirira guianensis Anacardiaceae Evergreen Terminalia amazonica Combretaceae Brevi-deciduous Trattinickia aspera Burseraceae Deciduous Virola sebifera Myristicaceae Evergreen 3100 mm Aspidosperma cruenta Apocynaceae Evergreen Brosimum utile Moraceae Evergreen Calophyllum longifolium Clusiaceae Evergreen Dussia mundia Fabaceae Evergreen Manilkara bidentata Sapotaceae Evergreen Marila laxiflora Clusiaceae Evergreen Poulsenia armata Moraceae Brevi-deciduous Tapirira guianensis Anacardiaceae Evergreen 3500 mm Aspidosperma cruenta Apocynaceae Evergreen Carapa guianensis Meliaceae Evergreen Cassipourea eliptica Rhizophoraceae Evergreen Erisma blancoa Vochysiaceae Evergreen Sacaglottis trygynum Hernandiaceae Evergreen Sterculia costaricana Sterculiaceae Brevi-deciduous Virola koschnyi Myrsticaceae Evergreen Zygia ramiflora Fabaceae Evergreen

PAGE 28

18 Table 2-3. Correlation coefficients between leaf defensive characteristics and mean annual precipitation and lamina thickness for 32 species of lowland tropical forest canopy trees at four levels of mean annual precipitation. Mean annual precipitation (mm yr -1 ) Lamina thickness (mm) r P r P Toughness (kg) 0.535 0.0016 0.843 <0.0001 FiberNDF (% dry mass) 0.127 0.490 0.356 0.0454 FiberADF (% dry mass) 0.124 0.500 0.276 0.1270 Lignin (% dry mass) 0.077 0.674 0.091 0.6223 Cellulose (% dry mass) 0.168 0.358 0.434 0.0130 NDF:N 0.470 0.0067 0.674 <0.0001 ADF:N 0.438 0.0121 0.600 0.0003 Lignin:N 0.376 0.0339 0.442 0.0113 Note: Bold type indicates significant correlation.

PAGE 29

19 JFMAMJJASONDJF soil (MPa) -4-3-2-10 3500 mm 3100 mm 2300 mm 1800 mm 20012002 Date JFMAMJJASONDJF Soil Gravimetric Water Content (% of wet mass) 2030405060708090100 Wet seasonDry seasonDry season Wet seasonDry seasonDry seasonAB Figure 2-1. Seasonal variation in (A) soil water potential ( soil ) and (B) gravimetric water content determined between 15-20 cm depth in four 1-ha lowland tropical forest study plots in Panama. Points represent mean (SE) (n=6).

PAGE 30

20 150020002500300035004000 Parea (mol O2 m-2 s-1) 51015202530 r 2 = 0.13p < 0.05 Mean annual precipitation (mm) 150020002500300035004000 Pmass (nmol O2 g-1 s-1) 050100150200250300350 r 2 = 0.42p < 0.0001AB Figure 2-2. Relationship between mean annual precipitation and maximum photosynthetic oxygen evolution rate (A) per unit area (P area ) and (B) per unit mass (P mass ) for the eight most common canopy species at four sites along a precipitation gradient in lowland Panamanian forest.

PAGE 31

21 150020002500300035004000 Leaf N (mg g-1) 510152025303540 r 2 = 0.21p < 0.01 Mean annual precipitation (mm) 150020002500300035004000 15N () -2-10123456 ABr 2 = 0.34p < 0.001 Figure 2-3. Relationship between mean annual precipitation and (A) leaf N concentration per unit mass and (B) N isotopic composition ( 15 N) for the most common canopy species (excluding legumes) at four sites along a precipitation gradient in Panama.

PAGE 32

22 Mean annual precipitation (mm) 150020002500300035004000 SLA (cm2 g-1) 04080120160200240 r 2 = 0.23p = 0.005 150020002500300035004000 Lamina thickness (mm) 0.00.10.20.30.40.50.6 r 2 = 0.30p = 0.001AB Figure 2-4. Relationship between mean annual precipitation and (A) lamina thickness and (B) specific leaf area (SLA) for the eight most common canopy species at four sites along a precipitation gradient in Panama.

PAGE 33

23 Leaf N (mg g-1) 510152025303540 Pmass (nmol O2 g-1 s-1) 050100150200250300350 r 2 = 0.78p < 0.0001 Log (SLA) 1.41.61.82.02.22.4 Pmass (nmol O2 g-1 s-1) 050100150200250300350 r 2 = 0.66p < 0.0001 Median leaf life span (days) 100200300400500600 Pmass (nmol O2 g-1 s-1) 050100150200250300350 r 2 = 0.52p = 0.005ABC Figure 2-5. Maximum photosynthetic O 2 evolution rate (P mass ) as a function of (A) leaf N from study species at the 1800 mm (open circles), 2300 mm (open triangles), 3100 mm (closed circles), and 3500 mm (closed triangles) sites. The outlier from the 3500 mm site (Zygia ramiflora) was not included in the regression due to large studentized residuals; P mass =95.56(leaf N).81. (B) P mass as a function of log-transformed specific leaf area (SLA); P mass =301log(SLA)429.7. (C) P mass as a function of median leaf life span for study species at the 1800 mm (open circles) and 3100 mm (closed circles) canopy crane sites P mass =0.53(leaf life span).93.

PAGE 34

24 Month JJASONDJFM leaf (MPa) -1.5-1.0-0.50.0 3100 mm predawn 1800 mm predawn 3100 mm midday 1800 mm midday Wet season Dry season20012002 Figure 2-6. Seasonal variation in leaf water potential ( leaf ) for canopy trees at two sites with contrasting precipitation in lowland Panama. Each point represents the mean (SE) of two individuals from 6-8 species at each site.

PAGE 35

CHAPTER 3 NUTRIENT CYCLING ON A PRECIPITATION GRADIENT IN LOWLAND TROPICAL FOREST Introduction Water availability has the potential to control components of nutrient cycles, such as nutrient pool sizes and fluxes between pools in terrestrial ecosystems. Water may affect nutrient cycling directly through soil processes such as leaching, weathering and decomposition of organic matter. Water may also influence nutrient cycling indirectly, through effects on plant community composition, since many of the plant characteristics that influence nutrient cycles, such as litter quality and productivity, vary with precipitation. Lowland tropical forest is important in global nutrient cycles and therefore, understanding how nutrient pool sizes and cycling respond to water availability is crucial. Studies of nutrient cycling in humid montane tropical forest (>2000 mm precipitation yr -1 ) suggest that C pool sizes increase whereas N availability decreases with increasing precipitation (Austin and Vitousek 1998; Schuur et al. 2001; Schuur and Matson 2001). These montane forest studies were conducted in Hawaii, where precipitation may vary while other state factors such as parent material, species composition and temperature remain constant (Vitousek 1995). The purpose of this study is to use a precipitation gradient to sample patterns of nutrient cycling in lowland tropical forests that vary in soil parent material and plant community composition. As precipitation inputs exceed biological demand, such as in the wet tropics, several key soil processes that influence soil nutrient cycling and storage are likely to 25

PAGE 36

26 occur. Removal of highly mobile nutrients in the soil solution may occur via leaching, a process that may contribute to reduced nutrient availability under conditions of high water inputs (Radulovich and Sollins 1991; Schuur and Matson 2001). Where poorly drained soils are prevalent, oxygen availability may limit microbial activity that mineralizes N. Several lines of evidence also suggest that water availability may affect soil nutrient pools by weathering soil to produce secondary minerals with a higher surface area and thus higher capacity to adsorb organic matter (Torn et al. 1997). Therefore, water availability has the potential to shape ecosystem nutrient dynamics through effects on the size of soil nutrient pools and the rate at which nutrients become available for plant uptake. Differences in water availability also result in distinct species assemblages that have the potential to influence nutrient cycling through effects on litter productivity and quality. Species in low resource environments tend to minimize tissue turnover and produce long-lived leaves with high concentrations of carbon-based defenses and slow rates of decomposition (Chapin 1980). In contrast, species from high resource habitats produce relatively short-lived leaves with higher photosynthetic rates and higher N concentrations, resulting in faster decomposition rates (Chapin 1980). The influence of litter quality in determining nutrient availability is thought to increase with actual evapotranspiration (AET) (Meentemeyer 1978). Tropical forest has high AET, so litter quality is expected to have a strong effect on nutrient availability. I present data on nutrient cycling from a precipitation gradient in lowland tropical forest in Panama. Substantial changes in species composition of canopy trees exist along this precipitation gradient (Chapter 2). The objective was to use the precipitation gradient to sample

PAGE 37

27 diversity of leaf phenological patterns across plant communities, which are also affected by soil properties. Materials and Methods Study Site This study was conducted in lowland tropical forest along a rainfall gradient in Central Panama (Chapter 2). Mean annual precipitation (MAP) across these sites varied from 1800 mm yr -1 at the driest site to approximately 3500 mm yr -1 at the wettest site (ACP 2002). Four 1-ha lowland forest study plots established by the Center for Tropical Forest Science (CTFS) over a range of precipitation with minimal changes in altitude and temperature were used as study sites (Table 3-1). All sites have a mean monthly precipitation >100 mm during the wet season between May and December, but dry season length (mean number of 30-day periods with <100 mm precipitation) varies between 129 days at the 1800 mm site and 67 days near the 3500 mm site (Condit 1998). Variation in rainfall during the study period (2000-2002) was within 15% of MAP at all sites. The soils in the Panama Canal forests are generally well-drained and rich in clay, Ca, Mg and N, but poor in K and P relative to other tropical soils (Dietrich et al. 1982; Kursar et al. 1995; Yavitt et al. 1993). All of the study sites lie on volcanic substrate except the 3100 mm site, which lies on sedimentary substrate (Table 3-1). The 1800 mm site is derived from the early to late Oligocene, principally agglomerate, generally andesitic in fine-grained tuff and includes stream-deposited conglomerate (Woodring et al. 1980). Soils of the 2300 and 3500 mm sites are described as derived from altered basaltic and andesitic lavas and tuff, including dioritic and dacitic intrusive rocks. The

PAGE 38

28 3100 mm site is derived from the late Miocene or early Pliocene with massive, generally fine-grained sandstone (Woodring et al. 1980). Species composition and phenological habit of the most common canopy trees change rapidly across this gradient as species richness increases steeply with mean annual precipitation (Pyke et al. 2001). The dominant canopy tree species in drier forest tend to exhibit shorter leaf life spans and a dry season deciduous leaf phenology (S.J. Wright, unpublished data). Foliar Chemistry As part of a larger study on photosynthetic leaf traits of canopy tree species along this precipitation gradient (Chapter 2), the eight canopy tree species at each site with the largest relative proportion of basal area were studied. Young, fully-expanded mature sun leaves were collected from 2-4 individuals for a total of 15 leaves per species from a site. Leaves were collected using canopy cranes maintained by the Smithsonian Tropical Research Institute at the 1800 and 3100 mm sites. At the 2300 and 3500 mm sites, leaves were collected from the upper canopy using a shotgun. Leaf material was dried for 48 h. at 65C. Leaf samples from the same tree were pooled for chemical analysis. One composite sample from each tree was analyzed for C and N using an elemental analyzer (Model NCS 2500, Carlo Erba, Milano, Italy). Litterfall Collection and Processing Litter was collected in 0.25 m 2 traps randomly located at 20X20-m grid points in each 1 ha plot. Large items, such as palm fronds often fell across traps and only material that fell in the area above the trap was collected. Litter was collected 11 times at intervals ranging from 1 to 12 weeks for the period between February 2001 and February 2002. Litter used in chemical analyses was collected at intervals of 7-10 days in February-June

PAGE 39

29 2001, October 2001, and February 2002. Litter collected at intervals >2 weeks were adjusted for mass loss within traps using decomposition data from the 3100 mm site (Chapter 5), and were not used in chemical analyses. Litter was sorted into four classes: (1) fine woody debris 1 cm in diameter; (2) leaves; (3) reproductive structures including fruits, flowers and seeds; and (4) other components of litterfall including insects, frass, canopy soil, and items too decomposed to identify. Litter was dried for 48 hours at 65C and the separate classes were weighed. Entire samples from each site and pickup date were ground in a Wiley mill (mesh size 40) and homogenized. A 120-ml subsample was then retained for further chemical analyses. Leaf Litter Chemistry Leaf carbon fraction analyses were performed using a series of increasingly aggressive extractants (Ryan et al. 1989). Dried, ground litter samples were digested in a detergent solution to separate labile cell contents from neutral detergent fiber (NDF), which includes all cell wall components. A dilute acid detergent solution was then used to determine acid detergent fiber (ADF, lignocellulose) before cellulose was separated from lignin and insoluble ash in 72% H 2 SO 4 Litter C and N concentrations were determined with an elemental analyzer (Model ECS 4010, Costech, Valencia, CA). Isotopic ratios of C ( 13 C) and N ( 15 N) were determined with a continuous flow isotope ratio mass spectrometer (Model Delta plus XL, Thermo Finnigan, Germany). Nutrient Availability Extractable nitrogen and phosphorus Soil cores were taken at each of the four sites in July 2002, approximately 2 months after the beginning of the 8-month wet season when nutrient availability is expected to be

PAGE 40

30 highest. Six 10 cm deep soil samples were taken with a slide hammer corer at random locations in each 1 ha plot. Samples were returned to the lab and hand sorted to remove roots and rocks, and three subsamples were taken. Weakly sorbed P from the first 10 g subsample was extracted in 50 ml 0.5 M NaHCO 3 (Crews et al. 1995; Miller et al. 2001). Extracts were shaken for one minute and after 24 h a 10-ml sample of the supernatant was removed and frozen until transported to the University of Florida for analysis. I extracted NH 4 and NO 3 from the second 10-g subsample for 24 h in 50 ml 2 M KCl. Nitrogen mineralization and nitrification The third 10-g subsample of each core was weighed into a 100-ml sample cup, covered, and allowed to incubate aerobically for 10 days in a dark cabinet at 24C. After 10 days, incubated samples were extracted in 2M KCl as described above and the difference in NO 3 and NH 4 was recorded as net mineralization (Riley and Vitousek 1995). Ion exchange resins Soil nutrient availability was determined with ion exchange resins (Binkley and Matson 1983). Three grams of anion exchange resin (Biorad, AG 1-X8, 20-50 mesh, Cl form) and the same amount of cation exchange resin (Biorad, AG 50W-X8 20-50 mesh, H + form) were weighed into separate 5X6 cm undyed monopolyester bags (approx. 190 m mesh size). At each of the four sites, resin bags were placed vertically 4 to 7 cm deep. One anion and one cation resin bag were secured to a plastic stake with monofilament line at 6 random locations in each 1 ha plot. Resin bags were collected from the field after 21 days and rinsed with DI water to remove soil particles. Ions were extracted with 20 ml of 0.5 M HCl and then neutralized with 20 ml of 0.5 M NaOH. PO 4 and NO 3 from anion

PAGE 41

31 extracts and NH 4 K, Ca and Mg from cation extracts were measured colorimetrically using an autoanalyzer at the University of Florida, Food and Agricultural Sciences Soils Testing Lab to determine nutrient content per bag. Soil nutrient pools The remaining sample of each soil core was dried at 50C for 48 h, and sifted to pass through a 2-mm sieve. Total soil C and N as well as 13 C and 15 N were analyzed on an elemental analyzer (Model ECS 4010, Costech) connected to a continuous flow isotope ratio mass spectrometer (Model Delta plus XL, Thermo Finnigan, Germany). Bulk density of surface 10 cm was estimated using 5 cm diameter volumetric cores dried at 105C. Soil moisture was determined gravimetrically on bulk density samples and expressed as grams of water per unit mass of dry soil. Results Foliar Chemistry Foliar N of the eight most common canopy tree species was highest at the 1800 mm site and lowest at the 3100 mm site, showing a general decrease in foliar N as precipitation increases (Table 3-2). In contrast, foliar C values were similar at all sites. Foliar C:N of the eight most common canopy tree species decreased significantly with increasing precipitation (r 2 =0.32; P<0.001), and was largely driven by changes in N. Litter Production and Chemistry Total litter production varied from 12.47 Mg ha -1 yr -1 at the 1800 mm site to 9.80 Mg ha -1 yr -1 at the 3500 mm site. Fine woody litterfall represented 11-17% of total litter and was about 20% higher at the two wetter sites (Table 3-3). Leaf litter represented 60-75% of total litter and was 45% higher at the 1800 mm site than the other three sites (Table 3-3). Reproductive structures comprised 7-15% of total litter fall and showed no

PAGE 42

32 clear pattern with precipitation (Table 3-3). Other litter components were 5-9% of total litter production (Table 3-3). Litterfall rates were seasonal with more litter falling during the dry season, and seasonal differences between maximum and minimum litterfall rates decreased with increasing precipitation (Figure 3-1). There was significant variation in mean litter N concentration of leaf litter with precipitation (F=5.07; P<0.05) and values tended to increase with increasing precipitation (Table 3-3). There was also significant variation in litter 15 N among sites with different soil parent material (Table 3-3; F=28.05; P<0.0001), but values did not vary in any predictable manner with precipitation. Litter 13 C decreased linearly with increasing precipitation indicating greater integrated water use efficiency of leaves in drier forest (Figure 3-2). Litter C increased (F=33.01; P<0.0001) with increasing precipitation and may be related to variation in litter lignin concentration among sites (Table 3-3; F=11.76; P<0.001). Litter lignin:N significantly increased with increasing precipitation (F=3.74; P<0.05) whereas cellulose concentration showed no significant variation with precipitation (Table 3-3). Nutrient Availability Extractable nitrogen and phosphorus Extractable P was significantly higher at the driest site than at the three wetter sites (Table 3-4; F=3.50; P<0.05). Extractable NO 3 was significantly higher at the wettest site than at the three drier sites (Table 3-4; F=4.80; P<0.05). Extractable NH 4 net N mineralization and nitrification showed no clear pattern with precipitation (Table 3-4), but N mineralization decreased linearly with increasing litter lignin:N (Figure 3-3) indicating an effect of litter quality on N availability.

PAGE 43

33 Ion exchange resins Resin exchangeable NO 3 was about 420% higher at the two sites on pre-tertiary basalt (2300 and 3500 mm sites) compared to the other two sites (Table 3-4; F=5.81, P<0.01). There were no clear patterns in exchangeable NH 4 or K, but exchangeable Ca decreased 76% from the driest to wettest sites (Table 3-4; F=11.75; P<0.0005) and measurable quantities of exchangeable Mg were only detected at the driest site (Table 3-4). Soil nutrient pools Total soil N and C pools of the top 10 cm were higher at the two wetter sites than at the two drier sites (Table 3-5). Soil N showed an exponential increase in relation to litter lignin concentration (Figure 3-4A) and a strong linear increase in relation to litter lignin:N (Figure 3-4B) suggesting that lignin-bound proteins in the soil organic matter matrix comprise a large part of the total soil N pool. Soil 13 C and 15 N were more enriched than litter inputs, possibly due to faster turnover of lighter isotopes (Figure 3-2; Table 3-5) (Nadelhoffer and Fry 1988). Soil bulk density decreased with increasing precipitation (Table 3-6). Discussion Foliar Chemistry Our results suggest that patterns of foliar C per unit N reflect changes in species composition across this precipitation gradient. Decreasing foliar N of canopy species may reflect increased allocation of carbon-based leaf defenses to canopy leaves in association with longer leaf life spans in wetter forest. Decreasing foliar N with increasing precipitation does not appear to be caused by reduced N availability with increasing precipitation as has been observed on rainfall gradients in Hawaii given that soil N

PAGE 44

34 availability did not decrease with increasing precipitation. This pattern, whether driven by photosynthetic or anti-herbivore allocation patterns, may provide an important feedback to nutrient availability by affecting litter quality since litter C:N is negatively correlated with decomposition rate (Chapter 5). Litter Production and Chemistry Patterns of litterfall suggest that the driest site may be slightly more productive than wetter forests. The primary differences are greater leaf litter production and stronger seasonality from the 1800 mm site, which has the highest proportion of dry season deciduous tree species. Therefore increased leaf litter production at the driest site may be the result of increased leaf turnover. However, in order to understand the extent to which productivity is regulated by water availability, other components of productivity such as root growth, trunk growth and respiration need to be incorporated. Nonetheless, litterfall is often the greatest fraction of productivity (Clark et al. 2001a). Several recent studies have also revealed that productivity in humid tropical forest may actually decline at high annual precipitation (>2500 mm) (Clark et al. 2001b; Schuur and Matson 2001). Therefore, assuming that all sites are at steady state, the result reported here is consistent with reduced ecosystem productivity at extremely high precipitation in the tropics due to light or nutrient limitation (Schuur and Matson 2001). The reduction in leaf litter quality with increasing precipitation suggests that nutrient mineralization slows with increasing precipitation. Because fine root litter quality may respond similarly to leaves (Ostertag 2001), belowground litter may contribute to reduced nutrient mineralization in sites with lower litter quality (Figure 3-3). Decomposition rates at the 3100 mm site decrease substantially with increases in litter lignin:N (r 2 =0.38; P<0.05; Chapter 5). Changes in litter lignin and lignin:N suggest that

PAGE 45

35 compound specific changes in litter quality may be important across this gradient. The increase in lignin suggests that a higher proportion of litter will enter directly into the slow decomposing pool or organic matter (Vitousek et al. 1994). This is a potential explanation for the increase in soil C and N storage with increasing precipitation and decreasing litter quality. Nutrient Availability The availability of P, Ca and Mg varied in relation to precipitation, whereas N appeared more responsive to parent material. Decreasing available P, Ca and Mg with increasing precipitation is consistent with the observation that weathering can leach these elements from the soil profile and reduce availability to plants. Relatively high P availability at the 1800 mm site may contribute to higher litterfall productivity, since P is often considered to be the most limiting nutrient in lowland tropical forest, and both P availability and litterfall productivity are highest at the 1800 mm site and relatively low at all other sites. However, litterfall is extremely variable and measurements over several years are needed to determine if this trend is robust. In contrast to P, N appeared to be under stronger regulation of parent material with both sites on tertiary basalt exhibiting high exchangeable NO 3 Both Austin and Vitousek (1998) and Schuur and Matson (2001) found decreasing N availability with increasing precipitation in Hawaiian montane forests in sites with consistent parent material. Our results suggest that parent material may alter the relationship between precipitation and N availability in tropical forest. The results of increasing total soil N and C with increasing precipitation support the notion that soil organic matter increases with increasing precipitation in humid forest (Schuur et al. 2001). Soil N pools in this part of Panama are high relative to both

PAGE 46

36 temperate and montane tropical forest (Austin and Vitousek 1998; Schuur and Matson 2001; Vitousek and Sanford 1986). The increases in soil N and C pools with increasing litter lignin and litter lignin:N suggest that lower quality litter may function to increase soil organic matter accumulation. Since soil waterlogging appears to be of minimal importance in these study sites, the observed increase in soil organic matter with increased precipitation may reflect an indirect effect of precipitation on species composition with more evergreen species producing lower quality litter as precipitation increases. Total soil N increases as a function of litter lignin:N, whereas soil N mineralization rates decrease with increasing litter lignin:N (Figure 3-3) (Scott and Binkley 1997). Therefore, as lignin:N increases, it appears that a higher proportion of soil N is tightly held in the organic matter matrix and the mineralization rate of that N is slower. Lower litter quality may function as an N conservation mechanism to prevent N losses through leaching by decreasing the rate of organic N reactivity. Conclusions Overall, the results suggest that variation in plant community composition along this precipitation gradient can have substantial effects on soil nutrient pools and on how nutrients are cycled by vegetation. Several general patterns, such as decreasing litter quality and decreasing soil availability of PO 4 Ca and Mg with increasing precipitation appear to corroborate patterns described for island ecosystems. Despite the diversity of soil substrates and plant community compositional changes across even short distances in Panama, there appear to be some patterns of C and N cycling and accumulation that are consistent with previous findings in Hawaiian ecosystems, where greater control over soil and species has generated theoretical predictions. Clearly, reduced litter quality and

PAGE 47

37 increased soil nutrient pools with increasing precipitation, which I have shown to be related, are two components of nutrient cycling that vary similarly in Panamanian lowland forest and Hawaiian montane forest. However, other patterns of N cycling are more strongly related to soil parent material than precipitation, making some predictions developed in relatively homogeneous island systems more difficult to apply in more heterogeneous landscapes. Further studies in lowland continental tropical forest are likely to contribute to our understanding of nutrient cycling in humid ecosystems and dispel myths or corroborate patterns observed in model island systems.

PAGE 48

38 Table 3-1. Characteristics of sites along precipitation gradient across the Isthmus of Panama. Site CTFS plot code a Mean annual precipitation (mm) Parent material b Order Suborder Parque Metropolitano PM 1800 Panama formation Ultisol Ustult Pipeline Road 8 2300 Pre-Tertiary basalt Ultisol Humult Fort Sherman S3 3100 Chagres Sandstone Histosol Saprist Santa Rita 31 3500 Pre-Tertiary basalt Ultisol Humult a Pike et al. (2001) b Woodring et al. (1980) Table 3-2. Percentage of N and C in canopy sun leaves from the eight most common canopy tree species across a precipitation gradient in lowland forest in Panama (Chapter 2). Values are means 1 SE. (n=8) Site Foliar N (%) Foliar C (%) 1800 mm 2.26 a 0.24 49.02 a 0.83 2300 mm 2.01 ab 0.20 47.72 a 1.57 3100 mm 1.43 b 0.12 49.24 a 1.10 3500 mm 1.64 ab 0.28 49.42 a 1.08

PAGE 49

39 Table 3-3. Litterfall rates separated by component for the year between February 2001-February 2002 and litter quality of leaf litterfall from 4 sites along a precipitation gradient in lowland Panamanian forest. Values for leaf litter quality are means 1 SE (n=4). Site 1800 mm 2300 mm 3100 mm 3500 mm Litterfall (Mg ha -1 yr -1 ) Fine woody debris 1.37 1.37 1.65 1.58 Leaf 9.47 6.33 6.45 6.74 Reproductive structures 0.94 1.40 1.79 0.64 Other 0.69 0.93 0.62 0.83 Total 12.47 10.03 10.51 9.79 Litter chemistry N (%) 0.96 a 0.04 1.34 b 0.06 1.13 ab 0.06 1.28 b 0.09 15 N () -0.46 a 0.24 1.24 b 0.09 0.25 c 0.10 1.23 b 0.06 C (%) 39.7 a 0.6 44.3 b 0.5 47.1 c 0.3 45.5 bc 0.5 Lignin (%) 15.4 a 1.9 20.9 b 1.6 23.8 b 1.6 23.7 b 1.2 Lignin:N 16.0 a 1.5 16.0 a 0.9 21.1 b 2.0 18.5 ab 1.7 Cellulose (%) 20.0 a 0.7 18.4 a 2.5 22.6 a 1.8 21.6 a 1.2

PAGE 50

40 Table 3-4. Soil chemistry to a depth of 10 cm. Values with different letters are significantly different at a P-value of 0.1. Site 1800 mm 2300 mm 3100 mm 3500 mm Extractable nutrients (mg kg -1 ) P 5.26 a 0.78 3.21 b 0.53 2.85 b 0.77 3.33 b 0.36 NO 3 0.07 a 0.08 0.73 a 0.61 0.23 a 0.25 3.73 b 1.59 NH 4 2.73 a 1.19 2.95 a 0.28 3.91 a 0.40 4.73 a 0.85 Net N nitrification (mg kg -1 d -1 ) 0.71 a 0.20 1.13 a 0.21 0.57 a 0.20 0.94 a 0.43 Net N mineralization (mg kg -1 d -1 ) 0.74 a 0.18 1.17 a 0.17 0.35 b 0.15 0.74 a 0.38 Exchangeable nutrients (mg kg resin -1 21 d -1 ) NO 3 2.34 a 1.46 14.75 b 1.80 3.82 a 2.10 17.4 b 5.20 NH 4 0.47 a 0.08 0.55 a 0.10 0.50 a 0.02 0.54 a 0.01 PO 4 8.01 a 2.01 0.30 b 0.10 0.33 b 0.04 0.10 b 0.03 K 671 a 67 491 b 13 470 b 14 570 ab 33 Ca 41.11 a 8.45 21.96 b 5.57 3.38 c 0.93 9.93 bc 0.41 Mg 3.56 a 1.74 < 0.01 b < 0.01 b < 0.01 b

PAGE 51

41 Table 3-5. Bulk soil chemistry from the surface 10 cm of soil. Values with a different letters are significantly different (P<0.05; n=6) Site Total C (g/kg) Total N (g/kg) Bulk 15 N () X 1 SE X 1 SE X 1 SE 1800 mm 25.21 a 2.79 2.86 a 0.81 2.99 a 0.61 2300 mm 28.60 ab 3.12 2.73 a 0.17 6.57 b 0.58 3100 mm 70.66 c 5.40 5.25 b 0.45 4.06 c 0.64 3500 mm 39.57 b 4.76 3.90 ab 0.36 5.67 d 0.72 Table 3-6. Soil bulk density and water content in the surface 10 cm of soil. Values are means ( 1 SE) of averages measured using 6 samples per site during the early wet season of 2002. Site Soil bulk density (g cm -3 ) Water content (g g -1 ) 1800 mm 0.92 0.02 0.53 0.02 2300 mm 0.73 0.03 0.58 .001 3100 mm 0.51 0.02 0.82 0.02 3500 mm 0.60 0.02 0.98 .006

PAGE 52

42 Mean Annual Precipitation 150020002500300035004000 Max-min litterfall rate 1015202530354045 r 2 = 0.98 Figure 3-1. The difference between seasonal maximum and minimum litterfall rates as a function of mean annual precipitation for four forest sites along a precipitation gradient in lowland Panama. Values are means (n=11).

PAGE 53

43 Mean annual precipitation (mm) 150020002500300035004000 13C () -31-30-29-28-27-26 Soil Litter r 2 = 0.93 Figure 3-2. Relationship between carbon isotope composition of soil (n=6) and litter (n=4) for four 1 ha forest plots along precipitation gradient in lowland Panama. Values are means 1 SE.

PAGE 54

44 Litter lignin:N 141618202224 N mineralization rate (mg kg-1) 0.000.250.500.751.001.251.50 r 2 = 0.77 Figure 3-3. Soil N mineralization rates for the top 10 cm (n=6) as a function of litter lignin to N ratio (n=4) for four forest sites in lowland Panama that vary in mean annual precipitation: 1800 mm (open circle); 2300 mm (open triangle); 3100 mm (closed circle); 3500 mm (closed triange). Values are means 1 SE.

PAGE 55

45 Litter Lignin (%) 1214161820222426 Soil N (g kg-1) 123456 Litter Lignin:N 12141618202224 Soil N (g kg-1) 123456 r 2 = 0.99r 2 = 0.81AB Figure 3-4. Soil N concentration as a function of (A) litter lignin and (B) litter lignin to N ratio for four forest study sites in lowland Panama. Values are means 1 SE. Symbols as in Figure 3-3.

PAGE 56

CHAPTER 4 HYDRAULIC CONDUCTIVITY AND WOOD DENSITY SCALE WITH LEAF PHOTOSYNTHETIC TRAITS IN PANAMANIAN FOREST CANOPY TREES Introduction Species often exhibit substantial variation in rates of carbon gain and resource use. In species-rich tropical forests, the extent to which observed patterns of resource utilization and carbon gain are species-specific is uncertain. Results of recent studies point to substantial convergence in plant functioning among species from diverse biomes (Reich et al. 1997). For example, maximum photosynthetic rates, stomatal conductance, and leaf surface area per unit mass (specific leaf area; SLA) are positively correlated and tend to decrease with increasing leaf life span across a wide array of study sites and angiosperm taxa (Ackerly and Reich 1999; Reich et al. 1999; Reich et al. 1997). However, within sites, variation in photosynthetic rate and leaf life span is as great or greater than variation in mean differences among biomes (Reich et al. 1999). Understanding hydraulic properties of the branch or whole plant scale may explain additional within-site variation in leaf characteristics among individuals and species (Meinzer and Goldstein 1996). Convergence in regulation of carbon economy at the leaf level may also be related to the life history features of a species. Photosynthesis, SLA, and leaf nitrogen concentration are generally related to rapid growth, high allocation to photosynthetic tissue, early attainment of reproductive age, and regeneration in high resource habitats (Cornelissen et al. 1997; Poorter and Remkes 1990; Reich et al. 1992; Wright and 46

PAGE 57

47 Westoby 1999). In addition, leaf photosynthetic traits are correlated with many of the same whole-organism traits that can be predicted by plant hydraulic conductance. Interspecific variation in leaf physiology may be related to xylem hydraulic properties, because stomatal conductance, a leaf area-based property, is often closely coordinated with the apparent hydraulic conductance of the soil-to-leaf pathway (Andrade et al. 1998; Kppers 1984; Meinzer and Grantz 1990; Sperry and Pockman 1993). Recent studies have suggested that allometric scaling of plant vascular systems is universal and thus reflects convergence among many species to overcome similar physical limitations of long-distance water transport (Enquist et al. 1998; Meinzer 2003; West et al. 1999). Furthermore, similar relationships in the scaling of plant transpiration and animal metabolism suggest that both share common scaling laws that reflect how resource requirements of organisms affect distribution in ecological communities (Enquist et al. 1998). Since photosynthesis is the sole mechanism of carbon assimilation in most vascular plants, and water is likely to limit photosynthesis at some time scale in most terrestrial environments, I expect coordination between photosynthetic capacity and plant hydraulic properties. Therefore, measurements of plant hydraulic properties may be related to other plant processes such as nutrient use and gross photosynthesis, thus integrating leaf level processes into a more complete understanding of whole-plant function. This study was designed to examine the allometry of branch hydraulic architecture, xylem biophysical properties and suites of leaf photosynthetic traits among 20 species of canopy trees growing in two Panamanian lowland forests. The primary objective was to determine the extent to which variation in leaf area-based hydraulic properties and xylem

PAGE 58

48 biophysical properties can explain variation in leaf gas exchange characteristics. Specific questions included: 1) Does allocation to leaf photosynthetic capacity correspond to capacity for hydraulic water supply? 2) Is hydraulic conductivity correlated with physiological leaf traits such as life span, nitrogen concentration and water use efficiency? 3) Does wood density constrain hydraulic conductivity to affect leaf gas exchange? 4) Can photosynthetic traits be predicted from xylem biophysical properties? Materials and Methods Study Site and Species The study was conducted from two canopy cranes operated by the Smithsonian Tropical Research Institute (STRI) in the Republic of Panama. Each crane is equipped with a gondola suspended by cables from a rotating boom that allows coverage of approximately 0.82 ha of forest. One crane is located in Parque Metropolitano, a secondary dry forest on the edge of Panama City that receives approximately 1800 mm of precipitation annually with a distinct dry season between December and April (Condit 1998). The other crane is located in an old-growth forest at Fort Sherman on the Caribbean side of the Panamanian Isthmus where mean annual precipitation is 3100 mm and the dry season is shorter and less intense than at Parque Metropolitano. During the dry season of 2002 (Feb-Mar), physiological and morphological characteristics in the upper crown of one to four individuals of 20 canopy tree species were measured. Both rare and common species were included in the study (Table 4-1). I measured at least five sun-exposed terminal branches in all species. In rare species with only one individual at the study site, sample branches were taken from different portions of the crown.

PAGE 59

49 Gas Exchange Measurements Maximum rates of net CO 2 assimilation (A) and stomatal conductance (g s ) were measured with a portable photosynthesis system (Model 6400, Li-Cor, Inc., Lincoln, NE, USA) between 0700 and 1100 h. Two to four newly formed mature leaves per branch were measured at 400 mol CO 2 and 1200 mol m -2 s -1 photosynthetic photon flux density (PPFD) provided by a red blue light source (model 6400-02B #SI-710, Li-Cor, Inc.). Only gas exchange values measured at a leaf to air vapor pressure deficit (VPD l ) below 1.5 kPa were used because significant stomatal closure above this value was observed. Hydraulic Conductivity Hydraulic conductivity (k h ) was measured on twigs excised from 1.0-1.5-mlong branches. Branches were longer than the measured maximum vessel length determined by the air pressure technique (Ewers and Fisher 1989). Upper crown branches exposed to full sun were cut directly after gas exchange measurements on days with similar environmental conditions and transported to the laboratory. In the laboratory, the first fully developed twigs supporting the leaves on which gas exchange was measured were excised from the rest of the branch under filtered water (0.2 m) to prevent xylem embolism. After removing 3 to 4 mm of bark from each end, the cut ends were shaved with Teflon coated razor blades and connected to the hydraulic conductivity apparatus (Sperry et al. 1988). Pith areas were plugged with plasticine when necessary. The hydraulic conductivity apparatus consisted of a beaker supplying filtered (0.2 m) water under low (1.4 kPa) gravitational pressure to the stem. A low hydraulic head insured no embolisms were removed and the apparatus was frequently flushed with 10% bleach

PAGE 60

50 solution to avoid microbial growth. Flow rates were determined volumetrically 5-15 minutes after connection when they became steady. Initial k h was estimated as the rate of water flux (J, mmol s -1 ) in a stem when a pressure gradient (dP/dx, MPa m -1 ) was applied across the stem k h = J /(dP/dx). (4-1) Maximum k h was measured after a 20-min high-pressure flush from a captive air tank when flow rates reached steady values. After hydraulic measurements, stem length, xylem diameter and pith diameter were measured for the calculation of k h and specific conductivity (k s ), k h per unit xylem area. Leaf area distal to the branch segment where k h was measured was recorded with a leaf area meter (model LI-3100, Li-Cor Inc.), and used to calculate leaf specific conductivity (LSC), k h per unit leaf area. Wood Density After conductivity measurements, outer bark, phloem and pith were removed from sample stems, wet mass was determined and stems were dried for 48 h at 65C and weighed to determine the saturated water content (SWC). Wood density was measured as the ratio of xylem dry mass to xylem volume. Percent loss of conductivity (PLC) was calculated from the ratio of initial to maximum k h Foliar Analysis After measurement of their area, gas exchange leaves were dried for 48 h at 65C and weighed for the determination of specific leaf area (SLA; cm 2 g -1 ) so that photosynthesis could be calculated both on a unit leaf area and leaf mass basis. Leaves were ground in a Wiley mill and one composite sample per branch was analyzed for total leaf nitrogen and carbon isotope discrimination ( 13 C) on an elemental analyzer

PAGE 61

51 connected to a continuous flow mass spectrometer at the University of Idaho. This allowed us to express photosynthesis per unit leaf nitrogen (A N ) and to evaluate intrinsic water use efficiency estimated by discrimination against 13 C (Farquhar and Richards 1984). Results There were significant positive relationships between mean species A area and stomatal conductance (g s ) when plotted against leaf specific hydraulic conductance (LSC; Figure 4-1). A area and g s showed substantial increases with increasing initial (LSC initial ) and maximum (LSC max ) measurements, but LSC max was a much stronger predictor (Figure 4-1) suggesting that plants allocate photosynthetic capacity proportionally to the maximum operational capacity of the xylem. There were also significant positive relationships between A area and specific conductivity (k s ). When A area was regressed against initial and maximum k s the relationships were again stronger with maximum than initial conductance (r 2 =0.22; P<0.05 and r 2 =0.39; P<0.005, respectively), consistent with the pattern observed in LSC. Leaf nitrogen per unit area (N area ) varied independently of LSC max (Figure 4-2). However, photosynthesis per unit N (A N ) increased significantly with increasing LSC max suggesting that hydraulic constraints limit the instantaneous efficiency with which N appears to be used (Fig 4-2). Median leaf life span was negatively related to LSC max suggesting that there is an evolutionary tradeoff between long-lived leaves and high LSC (r 2 =0.24; P<0.1) similar to the functional relationship between leaf photosynthetic capacity and leaf life span (Reich et al. 1992). Leaf carbon isotope discrimination ( 13 C), a measure of intrinsic water use efficiency was not significantly related to LSC initial

PAGE 62

52 (r 2 =0.00; P=0.86) or LSC max (r 2 =0.00; P=0.78) when measurement branches were averaged per species. However, instantaneous water use efficiency (A area /g s ) declined with increasing LSC max indicating a tradeoff between hydraulic conductivity and water loss per unit carbon gain (Figure 4-2). Relationships between wood density and photosynthetic and hydraulic characteristics generally showed negative correlations. Stem saturated water content (SWC) was negatively correlated with wood density demonstrating a possible tradeoff between sapwood water storage and mechanical strength (Figure 4-3). LSC max was negatively related to wood density suggesting that high LSC can be achieved by producing larger vessels or higher vessel density at the expense of having lighter wood with potentially reduced mechanical strength (Figure 4-3). Both A area (Figure 4-3) and A mass (r 2 =0.50; P<0.001) were negatively correlated with wood density; thus wood density appears to be negatively related to photosynthetic capacity through its effect on hydraulic conductance per unit leaf area. Discussion The results suggest that plant traits regulating photosynthetic and hydraulic capacity are highly interdependent. Leaf specific hydraulic conductivity (LSC) varied proportionally with photosynthetic CO 2 assimilation (A) and this relationship is consistent within a group of canopy trees in Panamanian lowland forest. Additionally, it appears that tradeoffs in relation to wood density and leaf life span allow diversity of allocation patterns among coexisting species in the plant community. For example, species with low photosynthetic and hydraulic capacity exhibited longer leaf life spans, potentially minimizing the nutrient cost of leaf replacement (Chapin 1980). Species with lower photosynthetic and hydraulic capacity also have denser wood, which may be a result of

PAGE 63

53 smaller diameter vessels that constrain maximum xylem conductance but also allow increased biomechanical support and an increased xylem pressure threshold (Hacke et al. 2001). Furthermore, species with lower photosynthetic and hydraulic capacity show evidence of higher water use efficiency, thus tradeoffs result in several solutions to balancing carbon gain with water loss. These results indicate that understanding xylem hydraulic capacity adds information to leaf level measurements when comparing species and links photosynthetic allocation patterns with processes at the branch and whole organism levels of organization. The relationship between photosynthesis and hydraulic conductance reflects a balance between carbon gain and water loss, thus the primary feature regulated by LSC is probably stomatal conductance (g s ), and therefore transpiration. Species usually have specific operating ranges or minimum values of leaf water potential governed by stomatal regulation (Meinzer and Grantz 1991). Thus, if LSC increases as a result of partial defoliation or leaf shading, g s usually increases, but leaf water potential remains about the same because the transpiration/LSC relationship is conserved (Meinzer and Grantz 1991; Pataki et al. 1998). Therefore, the coordination of photosynthetic and hydraulic capacity may be more a consequence of the often reported coupling of A and g s and of g s with LSC rather than a direct relationship between A and LSC. A appears to belong to a suite of coordinated characteristics related to plant hydraulic architecture and wood density. High rates of photosynthesis in relation to LSC will not impair plant functioning, but excessive transpiration can result in xylem cavitation and turgor loss. Therefore, the data showing proportional allocation to photosynthetic and hydraulic capacity may mean that

PAGE 64

54 species only invest in transport capacity that can be supported without experiencing physiological damage. Understanding the contributions of individual species to productivity and resource use in species-rich tropical forest is challenging. Previous attempts to understand and model inter-species physiological variation have involved dividing species into discrete functional groups. Recent studies of regulation of water use among diverse tropical forest canopy tree species have shown that contrasting patterns of regulation at the leaf level tend to converge when appropriate scaling and normalizing factors are applied (Andrade et al. 1998; Goldstein et al. 1998; Meinzer et al. 1997). The data suggest that continuous scaling relationships work to collapse inter-species variation into functional relationships that reflect plant capacity for carbon gain and tradeoffs that allow a broad spectrum of capacities. Therefore, understanding where a species falls along a functional evolutionary tradeoff continuum may be more informative than the discrete functional group approach.

PAGE 65

55 Table 4-1. Area-based maximum photosynthetic rate (A area ), maximum leaf specific hydraulic conductivity (LSC max ), and wood density for study species from two lowland tropical forest sites in Panama. Species Family A area (mol m -2 s -1 ) LSC max (mmol m -1 s -1 MPa -1 ) Wood density (g cm -3 ) Fort Sherman Aspidosperma cruenta Apocynaceae 9.7 15.81 0.70 Dussia mundia Fabaceae 12.3 37.44 0.53 Guateria dumentosa Annonaceae 12.2 38.42 0.42 Humiriastrum diguense Hernandiaceae 11.2 34.05 0.58 Manilkara bidentata Sapotaceae 10.3 42.04 0.61 Marila laxifolia Clusiaceae 9.9 29.89 0.48 Miconia borealis Melastomataceae 16.8 81.00 0.50 Nectandra purpurescens Lauraceae 11.1 58.91 0.55 Ocotea ira Lauraceae 12.6 45.78 0.58 Poulsenia armata Moraceae 11.8 39.49 0.43 Pourouma bicolor Moraceae 13.7 49.90 0.45 Simarouba amara Simaroubaceae 17.5 102.37 0.41 Tapirira guianense Anacardiaceae 12.9 61.99 0.43 Trattinickia aspera Bursuraceae 12.2 43.00 0.57 Virola sebifera Myristicaceae 13.5 60.78 0.50 Vochysia ferruginea Vochysiaceae 18.3 120.69 0.35 Parque Metropolitano Chrysophyllum cainito Sapotaceae 9.9 33.84 0.61 Cordia alliodora Boraginaceae 15.4 61.59 0.47 Ficus insipida Moraceae 19.2 123.14 0.34 Luehea seemannii Tiliaceae 17.0 115.19 0.33

PAGE 66

56 Aarea (mol m-2 s-1) 8101214161820 r 2 = 0.88r 2 = 0.63 LSCinitial (mmol m-1 s-1 MPa-1) 020406080 gs (mol m-2 s-1) 0.00.10.20.30.40.5 LSCmax (mmol m-1 s-1 MPa-1) 020406080100120 r 2 = 0.67r 2 = 0.62ABCD Figure 4-1. Leaf photosynthetic rate per unit area (A area ; A-B) and stomatal conductance (g s ; C-D) as a function of initial and maximum leaf specific hydraulic conductivity (LSC) for 20 canopy tree species of lowland tropical forest in Panama. Symbols: A. cruenta (black circle), C. cainito (black square), C. alliodora (black triangle), D. mundia (black upside down triangle), F. insipida (black diamond), G. dumentosa (black hexagon), H. diguense (white circle), L. seemannii (white square), M. bidentata (white triangle), M. laxifolia (white upside down triangle), M. borealis (white diamond), N. purpurescens (white hexagon), O. ira (white crosshair circle), P. armata (white crosshair square), P. bicolor (white crosshair triangle), S. amara (white crosshair upside down triangle), T. guianense (white crosshair diamond), T. aspera (white crosshair hexagon), V. sebifera (black crosshair circle), V. ferruginea (black crosshair diamond).

PAGE 67

57 LSCmax (mmol m-1 s-1 MPa-1) 020406080100120 Aarea/gs (mol mol-1) 20304050607080 Narea (mol m-2) 0.000.040.080.120.160.200.24 AN (mol CO2 [mol N]-1 s-1) 20406080100120140 r 2 = 0.59r 2 = 0.44 Figure 4-2. Leaf nitrogen per unit area (N area ), leaf photosynthetic rate per unit leaf nitrogen (A N ), and instantaneous water use efficiency (A area /g s ) as a function of maximum leaf specific hydraulic conductivity (LSC max ) for 20 canopy tree species in lowland Panama. Symbols as in Figure 4-1.

PAGE 68

58 Saturated water content (%) 4080120160200 LSCmax (mmol m-1 s-1 MPa-1) 020406080100120 Wood density (g cm-3) 0.30.40.50.60.70.8 Aarea (mol m-2 s-1) 81012141618 r 2 = 0.63r 2 = 0.68r 2 = 0.75 Figure 4-3. Stem saturated water content, maximum leaf specific hydraulic conductivity (LSC max ), and photosynthetic rate per unit leaf area (A area ), as a function of stem wood density for 20 species of lowland forest canopy trees in Panama. Symbols as in Figure 4-1.

PAGE 69

CHAPTER 5 LEAF DECOMPOSITION IN A WET TROPICAL FOREST: LINKING LEAF TRAITS WITH NUTRIENT CYCLING Introduction Climate, substrate, and the decomposer community are the three primary factors controlling decomposition processes. In tropical wet forest, it has been proposed that the structure and chemistry of substrate should have a relatively large effect, because tropical forest tends to have warm moist climate and high biological diversity (Meentemeyer 1978). Substrate chemistry of leaf litter is a product of the resources allocated to the living leaf, minus the nutrients retranslocated during senescence. Several general patterns relating species to litter quality have been noted. For example, species from low resource habitats often have leaves of low nutrient content that may decompose relatively slowly, whereas species in high resource habitats generally produce leaves that have higher nutrient content and decomposable readily (Chapin 1980). Furthermore, leaves share interdependent characteristics such as photosynthetic rate, N concentration, specific leaf area and leaf life span (Reich et al. 1997). The purpose of this paper is to evaluate the extent to which such functional leaf characteristics predict litter quality and decomposition rates among species in a wet lowland tropical forest. Of the variety of leaf chemical components that are correlated with decomposition rate, it is generally accepted that litter lignin:N ratio is the strongest predictor (Cornelissen 1996; Hobbie 1996; Melillo et al. 1982). Considerably less work has gone into investigating which leaf chemical, structural, or physiological traits of functioning 59

PAGE 70

60 leaves predict decomposition (Cornelissen 1996; Cornelissen and Thompson 1997; Grime et al. 1996; Wardle et al. 1998). This approach is of interest because it allows the potential for understanding how growth and reproduction are related to the decomposability of plant tissue. For example, Cornelissen and Thompson (1997) found that specific leaf area (SLA), leaf life span, and leaf N were significant predictors of % mass loss during decomposition among herbaceous monocots, suggesting that relationships between leaf physiological traits of living leaves may be extended into leaf litter decomposition. Furthermore, antiherbivore defenses have been found to continue to work against decomposing organisms (Grime et al. 1996; Wardle et al. 2002). Thus leaf physiology and antiherbivore defenses, two major determinants of leaf structure and function, might be related to the effects of any one species on decomposition and nutrient cycling. It would be informative to link leaf physiology and antiherbivore defense to decomposition because of the body of ecological literature that relates suites of leaf functional traits to growth strategies in specific environmental regimes (Chapin 1993; Coley et al. 1985; Grime 1977; Westoby 1998). Leaf physiological traits, especially photosynthesis and water use efficiency, have been correlated with fitness in some studies, although mostly through correlations with other traits (Arntz and Delph 2001; Dudley 1996; Geber 1990; Lechowicz 1984). Therefore, it appears that many traits are correlated along a minimal number of functional axes. The goal of this study is to determine if it is possible to extend our understanding of correlated plant traits to include leaf decomposition rate.

PAGE 71

61 Materials and Methods Study Site and Species The study was conducted at the Fort Sherman canopy crane site on the Caribbean coast of Central Panama. The site contains a 5-ha plot of lowland tropical forest within the 12,000 ha San Lorenzo protected area. Mean annual rainfall at the site is 3100 mm, with a mild dry season from January to March. I accessed the canopy with a 52-m-tall construction crane equipped with a gondola and operated by the Smithsonian Tropical Research Institute. Thirty six plant species that represent a broad selection of growth forms including palms, lianas, monocot herbs, canopy trees and pioneer trees were selected for study (Table 5-1). Litter Collection and Decomposition Leaf litter was collected during the dry season, between January and April, 2001, when many species shed or exchange leaves. Senescent leaves were collected by hand directly from at least three individuals of each study species. Entire leaves with complete discoloration were gently shaken and harvested only if they came off the plant with a light touch, indicating a well-formed abscission zone. Palm fronds and monocot herb leaves remain on the plant for several months after retranslocation, so the most recent senescing leaf or frond that had no remaining green pigmentation was selected. Leaves were air dried in an air-conditioned laboratory (45% RH and 24C) for >1 month. Ten grams of litter from each species was set aside for initial litter quality chemical analyses. Two to three grams of litter from each species were placed in 1X1 mm mesh nylon-covered fiberglass window screen. Leaves of several species can weigh more than 3 g and for these species one entire leaf was placed in each bag. For palms, a leaflet was separated from the rachis and treated as a leaf. Four sizes of litter bags: 10X10,

PAGE 72

62 10X30, 20X20 and 30X30 cm were used to accommodate different leaf sizes and minimize folding. Litterbags were heat sealed with a dry iron and placed in the field on March 30, 2001, about one month before the beginning of the rainy season that year. Dowel rods (1/4 diameter, 15 cm long) were placed in the field with litterbags as a standard to compare with other decomposition studies (LIDET 1995). At 1, 3, 6, 14, and 24 months, 5 bags per species were collected from the field. Bags were gently rinsed with distilled water to remove adhered soil particles and any roots that had grown into the bags were removed with tweezers. A subsample of each species was weighed fresh for moisture content, while the rest of the bags were placed in the freezer overnight. The contents of each litterbag were then dried for 48 hours at 65C. Dried samples were weighed and ground in a Wiley mill through a 40 size mesh. Initial Litter Quality A subsample of litter from each species was dried for 48 hours at 65C for the determination of initial litter quality. Carbon fraction analyses were performed using a series of increasingly aggressive extractants (Ryan et al. 1989). Dried litter samples were digested in a detergent solution which separated non-polar extractives (cell contents) from neutral detergent fiber (NDF), which includes cell wall constituents and fractions that are not immediately nutritionally available. A dilute acid detergent solution was then used to determine acid detergent fiber (ADF, lignocellulose) before lignin was separated from cellulose in 72% H 2 SO 4 A separate subsample was analyzed for C and N using an elemental analyzer (Carlo Erba, Milano, Italy). Gas Exchange and Leaf Chemistry In the wet season of 2000 and 2001 (June-November), maximum rates of net CO 2 assimilation (A) and stomatal conductance (g s ) were measured with an infrared gas

PAGE 73

63 analyzer (Model 6400, Li-Cor, Inc., Lincoln, NE, USA) between 0700 and 1100 h. Five newly formed mature leaves from three individuals of each species were measured at 370 mol mol -1 (slightly higher than ambient CO 2 concentration), and 1500 mol m -2 s -1 photosynthetic photon flux density (PFD) provided by a red blue light source (model 6400-02B #SI-710, Li-Cor, Inc.). Only gas exchange values measured at a leaf-to-air vapor pressure deficit (VPD l ) below 1.5 kPa were used because significant stomatal closure was observed above this value. Gas exchange was measured on leaves growing in the highest light environment in which the species tends to grow. For canopy trees, lianas, pioneer trees and some palms, terminal leaves were accessed using the canopy crane. For monocot herbs and understory palms, measurements were made from the forest floor. Following photosynthetic measurements, lamina thickness was measured between primary and secondary veins with a digital caliper (Mitutoyo, Japan). Leaves were dried for 48 hours at 65C, weighed for determination of specific leaf area (SLA) and ball milled to a fine powder. All leaves from the same individual were pooled for chemical analysis. One composite sample from each individual was analyzed for C and N using an elemental analyzer (Carlo Erba, Milano, Italy). Leaf toughness was measured in the field on freshly collected leaves with a penetrometer (Pesola, Switzerland), which measured the maximum force during punching through the lamina between primary and secondary veins at a steady, slow rate with a 1 mm diameter plunger. Such measurements do not replicate the action by which herbivores puncture leaves (fracture toughness) but leaf toughness was correlated with fracture toughness in a study of 42 tropical tree species (Choong et al. 1992), so I use it as an index of allocation to structure and potential

PAGE 74

64 resistance to decomposition. Leaf polyphenol concentration was analyzed with the Prussian Blue Procedure in a subset of the species studied, including canopy trees, pioneer trees, lianas, and one palm (S. Joseph Wright, unpublished data). Results Decomposition rate (k) varied from a minimum of 0.41 in the canopy tree Vochysia ferruginea to a maximum of 4.58 in the pioneer tree, Piper hispidum. Overall, there were no significant differences in k among plant growth forms (F=1.76; P= 0.1617; Table 5-2). Several indices of litter quality varied significantly between the monocots, (palms and monocot herbs), and dicots (canopy trees, pioneer trees, and lianas). For example, non-polar extractives were significantly lower in monocot herbs and palms than other growth forms (F=7.86; P=0.0002). Litter cellulose concentration was significantly higher in monocot herbs and palms (F=7.16; P=0.0003), and acid detergent fiber (ADF) was higher in palms than other growth forms (F=6.43; P=0.0007). Litter lignin concentration was negatively related to decomposition rate (Table 5-3, Figure 5-1). Although litter N concentration was not a significant predictor of decomposition, the composite variable, litter lignin:N, was the strongest predictor of decomposition rate (Table 5-3, Figure 5-1). Litter C concentration and C:N were also negatively related to decomposition (Table 5-3, Figure 5-2). No other indices of litter quality were related to decomposition except ADF, and this relationship was weak (Table 5-3). Several chemical and structural traits of living leaves were significantly related to decomposition. Leaf N and C:N were significantly related to decomposition (Table 5-4, Figure 5-3). SLA was positively related to decomposition (Table 5-4, Figure 5-4), and species with tougher leaves tended to decompose more slowly (Table 5-4). Leaf N and

PAGE 75

65 SLA, two of the strongest leaf predictors of decomposition were also significant predictors of photosynthetic rate (Figure 5-5), indicating a relationship between leaf physiology and decomposition through these structural and chemical traits. Although lignin and lignin:N were the strongest predictors of decomposition, there were several species with low lignin concentrations that decomposed slowly. A general linear model including the effects of lignin and polyphenol concentration explained more of the variance in decomposition than lignin alone (Table 5-5). This result suggests that although lignin is probably the best general leaf defensive characteristic for predicting decomposition, other antiherbivore defenses may play a similar role in deterring decomposers following senescence (Grime et al. 1996; Wardle et al. 2002), and when combined with lignin appear to explain a greater proportion of the variance. Discussion Plant Growth Forms and Ecosystem Processes In high diversity tropical ecosystems, it has been difficult to relate resource use characteristics of individual plant species to system-wide processes such as nutrient storage and cycling. This predicament has led to the functional group approach, in which species with similar resource use characteristics are grouped into a smaller, more manageable number of functional groups. This approach greatly simplifies modeling of plant processes and has achieved success in many ecosystems (Chapin et al. 1995; Ewel and Bigelow 1996; Krner 1993; Tilman et al. 1997). However, it appears that continuous functions reflecting convergence in leaf physiology can also be used to simplify the effects of many species. Therefore, my results suggest that for predicting the effect of a species on an ecosystem process such as litter decomposition, it is more

PAGE 76

66 important to know the value of a continuous variable such as litter lignin:N than any discrete classification such as growth form, canopy position, successional status, or phylogenetic association. Litter Quality and Decomposition The result that litter lignin:N is the strongest predictor of decomposition reported in this study is consistent with numerous other decomposition studies (Cornelissen 1996; Hobbie 1996; Melillo et al. 1982; Ostertag and Hobbie 1999). Similarly, the result that litter C concentration and C:N are significantly related to decomposition corroborates much of what exists in the ecological literature. Mechanistically, it is important to note that the ratio of lignin to N reflects a ratio between leaf defense and photosynthetic potential. High leaf lignin concentration is a characteristic of species in low resource habitats and species with long-lived leaves, and is considered a general antiherbivore defense (Chapin 1980; Coley 1983; Reich et al. 1997). To some degree, litter N is a reflection of the physiological capacity of the leaf. Therefore, litter lignin:N is a ratio of defense and structure to physiological capacity, although effects on decomposition are more strongly driven by lignin. Viewed in this respect, the strongest predictor of litter decomposition integrates plant allocation along the axes of defense and potentially carbon gain. Photosynthesis and Decomposition Photosynthesis is the primary function of leaves and therefore most of the features of leaves are under strong selection to maximize carbon gain over the life of the leaf (Kikuzawa 1991; Mulkey et al. 1995; Reich et al. 1997). Ecologists have identified a suite of highly interdependent leaf characteristics that occur repeatedly and appear important in the exploitation of specific habitats (Chapin 1980). Specific leaf area (SLA)

PAGE 77

67 and leaf N are two of these leaf characteristics that are positively related with photosynthesis across broad groups of plant species (Reich et al. 1992). SLA and leaf N are also the strongest leaf predictors of decomposition in this study and suggest that allocation to photosynthetic capacity is related to decomposition through these correlations (Figure 5-6). Interestingly, leaf N was correlated with decomposition rate, but litter N was not. Leaf N may be related to a host of other leaf characteristics that influence decomposition. Although photosynthetic rate and decomposition rate may show a positive correlation, litter does not photosynthesize. Therefore, it is necessary to consider chemical and structural leaf components that are related to photosynthesis to understand how decomposition is a product of selection to maximize carbon gain in leaves.

PAGE 78

68 Table 5-1. Summary of study species and leaf litter decomposition rate (k) organized by growth form. All leaf litter material was collected in lowland wet tropical forest at Fort Sherman, Panama. Species Family Growth form k (yr -1 ) Aspidosperma cruenta Apocynaceae Canopy tree 1.27 Brosimum utile Moraceae Canopy tree 0.47 Calophyllum longifolium Clusiaceae Canopy tree 0.46 Carapa guianensis Meliaceae Canopy tree 0.45 Dussia mundia Leguminosae Canopy tree 1.11 Manilkara bidentata Sapotaceae Canopy tree 0.66 Nectandra purpurascens Lauraceae Canopy tree 0.57 Pourouma bicolor Cecropiaceae Canopy tree 0.44 Simarouba amara Simaroubaceae Canopy tree 1.01 Tapirira guianensis Anacardiaceae Canopy tree 0.96 Vochysia ferruginea Vochysiaceae Canopy tree 0.41 Apeiba membranaceae Tiliaceae Pioneer tree 1.05 Cecropia insignis Cecropiaceae Pioneer tree 1.00 Jacaranda copaia Bignoniaceae Pioneer tree 0.80 Clidemia octona Melastomataceae Pioneer tree 0.69 Ochroma pyrmidale Bombacaceae Pioneer tree 0.58 Piper hispidum Piperaceae Pioneer tree 4.58 Trema micrantha Ulmaceae Pioneer tree 1.55 Arrabidaea verrucosa Bignoniaceae Liana 0.72 Cayaponia granatensis Cucurbitaceae Liana 0.72 Doliocarpus dentatus Dilleniaceae Liana 0.44 Heisteria scandens Olacaceae Liana 1.40 Maripa panamensis Convolvulaceae Liana 0.82 Phryganocydia corymbosa Bignoniaceae Liana 1.24 Pleonotoma variabilis Bignoniaceae Liana 0.48 Tontelea richardii Hippocrateaceae Liana 0.56 Costus pulverulentus Zingiberaceae Monocot herb 0.64 Diffenbachia pittieri Araceae Monocot herb 3.18 Heliconia pogonantha Heliconiaceae Monocot herb 0.81 Stromanthe jacquinii Maranthaceae Monocot herb 0.80 Zingiber officinale Zingiberaceae Monocot herb 1.76 Calyptrogyne costatifrons Palmae Palm 0.61 Carloduvica palmata Cyclanthaceae Palm 0.90 Genoma cuneata Palmae Palm 0.79 Oenocarpus mapora Palmae Palm 0.64 Socratea esxorrhiza Palmae Palm 0.46 Dowels 0.18

PAGE 79

69 Table 5-2. Leaf litter decomposition rates and initial litter quality of study species from Fort Sherman, Panama, averaged by growth form. Values are mean 1 SE. Values with the same letter are not significantly different (P>0.05). Canopy trees (n=11) Pioneer trees (n=7) Lianas (n=8) Monocot herbs (n=5) Palms (n=6) k (yr -1 ) 0.71 a 0.10 1.47 a 0.58 0.80 a 0.13 1.44 a .054 0.68 a 0.08 C (%) 47.27 a 1.38 43.23 a 2.66 46.45 a 2.07 40.71 a 1.40 41.98 a 1.11 N (%) 0.90 a 0.09 1.22 a 0.18 1.10 a 0.08 1.02 a 0.06 1.14 a 0.16 C:N 58.42 a 6.42 39.23 a 5.95 44.42 a 5.57 40.29 a 1.44 38.69 a 3.89 Lignin (%) 16.03 a 1.66 13.67 a 2.71 16.46 a 3.17 11.96 a 1.62 21.48 a 3.19 Lignin:N 19.90 a 3.34 11.81 a 2.76 16.80 a 4.93 11.74 a 1.37 19.9 a 4.30 Non-polar extractives (%) 54.23 a 3.49 55.48 a 5.07 50.35 a 3.01 35.81 b 6.33 29.59 b 1.65 Cellulose (%) 18.41 a 1.90 18.01 a 3.00 19.56 a 2.72 27.67 b 1.71 32.61 b 1.82 FiberADF (%) 45.77 a 3.49 44.52 a 5.07 49.65 a 3.01 64.18 a 6.33 70.39 b 1.64

PAGE 80

70 Table 5-3. Regressions of litter chemical parameters and leaf litter decomposition rate (k). All plant growth forms were combined for analyses. Only statistically significant regressions (P<0.1) are reported. ADF, acid detergent fiber. (n=36) Parameter Regression equation r 2 Significance Litter C ln(k)=6.91-1.89Litter C 0.17 0.0122 Litter C:N ln(k)=2.29-0.66Litter C:N 0.15 0.0181 Lignin ln(k)=1.66-0.70Lignin 0.30 0.0005 Lignin:N ln(k)=1.53-0.65Lignin:N 0.38 <0.0001 ADF ln(k)=1.78-0.55ADF 0.09 0.0772 Table 5-4. Regressions of leaf chemical and structural parameters and leaf litter decomposition rate (k). All plant growth forms were combined for analyses. Only statistically significant regressions (P<0.1) are reported. SLA, specific leaf area. (n=36) Parameter Regression equation r 2 Significance Leaf N ln(k)=-0.97+1.15Leaf N 0.20 0.0059 Leaf C:N ln(k)=2.48-0.85Leaf C:N 0.19 0.0077 Lamina toughness ln(k)=0.66-0.18lamina toughness 0.10 0.0575 SLA ln(k)=-2.43+0.48SLA 0.21 0.0050 Table 5-5. Results of general linear model for combined effects of leaf phenol (measured with the Prussian Blue Procedure) and litter lignin on litter decomposition rate. This analysis was conducted on a subset of the study species for which phenol data was available. The analysis included canopy trees, pioneer trees, lianas and one palm. k Parameter Estimate P Intercept 2.005 0.0052 Phenols -1.243 0.0589 Lignin -0.839 0.0026 Phenols*Lignin 0.439 0.1146 Overall model p-value 0.0044 Model R 2 0.49 n 23

PAGE 81

71 AB ln litter lignin (%) 1.61.82.02.22.42.62.83.03.23.43.6 ln k (yr-1) -1.5-1.0-0.50.00.51.01.52.0 canopy trees pioneer trees lianas monocot herbs palms r 2 = 0.30 ln litter lignin:N 1.52.02.53.03.54.0 ln k (yr-1) -1.5-1.0-0.50.00.51.01.52.0 r 2 = 0.38 Figure 5-1. Relationship between leaf litter decomposition rate (k) and (A) initial litter lignin concentration and (B) initial litter lignin to nitrogen ratio for 36 lowland tropical wet forest plant species from Fort Sherman, Panama.

PAGE 82

72 ln litter C (%) 3.43.53.63.73.83.94.04.1 ln k (yr-1) -1.5-1.0-0.50.00.51.01.52.0 r 2 = 0.17 ln litter C:N 345 ln k (yr-1) -1.5-1.0-0.50.00.51.01.52.0 r 2 = 0.15AB Figure 5-2. Relationship between leaf litter decomposition rate (k) and (A) initial litter carbon concentration and (B) initial litter carbon to nitrogen ratio for 36 wet forest species. Symbols as in figure 5-1.

PAGE 83

73 AB ln leaf N (%) 0.00.51.01.5 ln k (yr-1) -2-1012 r 2 = 0.20 ln leaf C:N 2.02.53.03.54.0 ln k (yr-1) -2-1012 r 2 = 0.19 Figure 5-3. Relationship between leaf litter decomposition rate (k) and (A) nitrogen concentration of living leaf and (B) carbon to nitrogen ratio of living leaf for 36 wet forest species. Symbols as in figure 5-1.

PAGE 84

74 ln SLA (cm2 g-1) 3.54.04.55.05.56.0 ln k (yr-1) -2-1012 r 2 = 0.21 Figure 5-4. Relationship between leaf litter decomposition rate (k) and specific leaf area (SLA) for 36 wet forest species. Symbols as in figure 5-1.

PAGE 85

75 SLA (cm2 g-1) 050100150200250300350 Amass (nmol g-1 s-1) 050100150200250 r 2 = 0.25 Leaf N (%) 1.01.52.02.53.03.54.0 Amass (nmol g-1 s-1) 050100150200250 r 2 = 0.21AB Figure 5-5. Relationship between leaf photosynthetic rate per unit mass (A mass ) and (A) specific leaf area (SLA) and (B) leaf nitrogen concentration for 36 wet forest species. Symbols as in figure 5-1.

PAGE 86

76 SLA A mass k Leaf N Figure 5-6. Schematic diagram depicting how photosynthesis is related to decomposition rate (k) through correlation with specific leaf area (SLA) and leaf nitrogen concentration.

PAGE 87

CHAPTER 6 TEST OF GAS EXCHANGE MEASUREMENTS ON EXCISED BRANCHES OF TEN TROPICAL TREE SPECIES: A TECHNICAL REPORT Introduction Plant physiological parameters that describe CO 2 uptake and water loss are used to model biosphere-atmosphere interactions at local and regional scales (Running and Coughlan 1988; Williams et al. 1996) and to predict the sensitivity of vegetation to climate change at regional to global scales (Neilson and Marks 1994). Measurements from the upper canopy are important for calculations of net primary productivity (NPP) and direct gas exchange measurements are needed to understand the extent to which NPP is controlled by plant community type and environmental heterogeneity (Haxeltine and Prentice 1996; Woodward et al. 1995). Towers, scaffolds, walkways and construction cranes have been employed to obtain physiological parameters from intact branches in the upper canopy of mature forest (Mulkey et al. 1996). However, these structures are not available everywhere and there is a need for reliable estimates of leaf-level processes in places where it is not practical or feasible to build such structures. Researchers have measured gas exchange of canopy leaves by cutting or shooting small branches and re-cutting stems under water to re-establish the xylem water column in the form of a potometer before measurement (Dang et al. 1997; Koyama 1981; Reich et al. 1995; Reich et al. 1998). Other researchers have measured gas exchange of montane forest trees (Gerrish 1992) or conifers (Ginn et al. 1991; Samuelson 1998) by simply detaching a group of leaves or fascicles and placing them in a cuvette for immediate measurement. 77

PAGE 88

78 Although numerous physiological observations on excised foliage have been published, there is evidence that excision may alter gas exchange measurements in several ways. Even if a stem is double cut to re-establish the xylem water column, cavitation, changes in xylem pressure, and reductions in hydraulic conductance can result (Boari and Malone 1993; Stahlberg and Cosgrove 1995). These alterations in the xylem stream can influence stomatal responses (Sperry et al. 1993; Williamson and Milburn 1995) and inhibit the transport of hormones and nutrients, which are important for regulating shoot water potential and stomatal control (Tardieu and Davies 1993; Zhang and Davies 1990). The time scale at which these changes occur is likely to be important in determining the degree of excision-induced effects on gas exchange. However, these responses vary widely across the plant kingdom leaving us with little information regarding the reliability of gas exchange measurements on excised foliage; few studies provide verification of how leaves on excised stems perform (Dang et al. 1997; Ginn et al. 1991). The purpose of this study is to compare gas exchange rates measured on excised and attached branches of tropical forest canopy trees. I investigated how measurements of gas exchange and biochemical parameters derived from photosynthetic light and CO 2 response curves varied between leaves on excised and attached branches. Our main objective was to determine if excision causes significant effects on measured gas exchange rates and parameters, i.e. I wanted to know if the effect of excision on gas exchange parameters is greater than the range of values observed on intact branches. Second, I hoped to evaluate whether species differ in their gas exchange responses to excision, and if there are any characteristics that might be used to predict the degree

PAGE 89

79 excision-induced effects. Although measuring gas exchange on excised branches is a convenient technique for acquiring data, it seems important to address the limitations of this approach so as to prevent the collection of spurious data and to interpret the results of studies that reported gas exchange results from excised stems. Materials and Methods Study Sites and Species The study was conducted from two construction cranes operated by the Smithsonian Tropical Research Institute (STRI) in the Republic of Panama. Each crane is equipped with a gondola suspended by cables from a rotating boom that allows coverage of approximately 0.82 ha of forest. The first crane is located in Parque Metropolitano, a secondary dry forest on the edge of Panama City that receives approximately 1800 mm of precipitation annually with a distinct dry season between December and April. The second crane is located at Fort Sherman, on the Caribbean side of the Panamanian Isthmus in old growth forest that receives approximately 3100 mm of precipitation annually with a shorter and less intense dry season. Ten tree species from 9 families, representing a variety of leaf morphology and gas exchange rates were investigated. At Fort Sherman, Apeiba membranacea Spruce ex. Benth. (Tiliaceae), Jacaranda copaia (Aubl.) D. Don (Bignoniaceae) and Vochysia ferruginea Mart. in Mart. & Zucc. (Vochysiaceae) have relatively high rates of photosynthesis and stomatal conductance (Table 6-1). Aspidosperma cruenta Woods (Apocynaceae), Brosimum utile (Moraceae), Manilkara bidentata (Sapotaceae) and Simarouba amara Aubl. var. typica Cronq. (Simaroubaceae), in contrast, have lower rates of photosynthesis and stomatal conductance, and tend to have thicker leaves. At Parque Metropolitano, Anacardium excelsum (Bertero & Balb.) Skeels (Anacardiaceae) has

PAGE 90

80 relatively low rates of photosynthesis and stomatal conductance, whereas Pseudobombax septenatum (Jacq.) Dug. (Bombacaceae) and Luehea seemannii Tr. & Planch. (Tiliaceae) have higher rates of photosynthesis and stomatal conductance. All nomenclature follows (Croat 1978). Gas Exchange Measurements Net CO 2 assimilation (A) and stomatal conductance (g s ) were measured with an infrared gas analyzer (Model 6400, Li-Cor, Inc., Lincoln, NE, USA) on leaves of branches with full sun exposure. One gas exchange measurement was taken on a newly formed mature intact and attached leaf at ambient temperature, 37.26 Pa CO 2 (equivalent to 370 mol mol -1 slightly higher than ambient), and 1500 mol m -2 s -1 photosynthetic photon flux density (PFD) provided by a red blue light source (model 6400-02B #SI-710, Li-Cor, Inc.). The segment of branch was then excised 100 cm from the measured leaf and immediately shortened to 50 cm by re-cutting under water to re-establish the xylem water column. Successive gas exchange measurements were taken within 3 min and then several times up to 60 min after excision. The response of A to PFD was measured on newly formed fully expanded mature leaves at a range of PFD from 0 to 1500 mol m -2 s -1 37.26 Pa CO 2 and ambient relative humidity. Measurements were made at ambient temperature unless leaf temperature exceeded 33 C, in which case, peltier plates adjacent to the cuvette were used to maintain the block temperature at 33 C. I first measured A at 1500 mol m -2 s -1 PFD, then light was decreased in a stepwise fashion for a total of 10 measurement points down to 0 mol m -2 s -1 Once the light response curve was completed, the branch was immediately re-cut under water as described above and the curve was repeated in the

PAGE 91

81 gondola with the stem in a potometer on a leaf of similar age adjacent to the original leaf. An attempt was made to maintain the cut stem in sunlight to prevent stomatal closure as a result of low light availability in the gondola. Trials were performed on two to four branches per species. Parameters for light response curves were fit to a nonrectangular hyperbola (Sims and Pearcy 1991), which allows calculation of apparent quantum yield on the basis of incident photons (), the light-saturated rate of gross CO 2 assimilation at saturating irradiance (A max ), the radius of curvature (and dark respiration (R d ). The response of A to intercellular CO 2 partial pressure (p i ) was measured on three branches per species. I conducted measurements on intact branches by first inducing the leaf with 1700 mol m -2 s -1 PFD and 37.26 Pa CO 2 Measurements were then taken at 0 Pa CO 2 and the concentration was increased over a total of 14 measurement points to a final concentration of 151 Pa. The branch was then cut and re-cut as described above and a second CO 2 response curve was conducted on a leaf of similar age adjacent to the original leaf to avoid feedback inhibition. Parameters for CO 2 response curves were modeled by two equations that describe CO 2 -limited and rubisco-limited rates of photosynthesis (Lambers et al. 1998; von Caemmerer and Farquhar 1981). CO 2 -limited rate of photosynthesis [A(c)] can be calculated as (6-1) AcVppKRciimday()max Where Vc max is the rate of CO 2 assimilation at saturating p i is the CO 2 compensation point, K m is the Michaelis-Menten constant for the carboxylation reaction

PAGE 92

82 and R day is the rate of respiration during photosynthesis. The rubisco-limited rate of calculated as: calculated as: photosynthesis [A(j)] can be can be AjJppRiiday()max42 (6-2) (6-2) Where J max is the maximum rate of electron transport and all other parameters follow equation 1. Repeated measure analysis of variance (ANOVA) was used to assess the effect of excision on light and CO 2 response curves with branches as a random factor nested within treatments. Where J max is the maximum rate of electron transport and all other parameters follow equation 1. Repeated measure analysis of variance (ANOVA) was used to assess the effect of excision on light and CO 2 response curves with branches as a random factor nested within treatments. Results and Discussion Results and Discussion Time Course Measurements Time Course Measurements After excision, all species showed a reduction in both A and g s within 3 min (Figures 6-1 and 6-2). However, after 10 min, most species showed fewer fluctuations and appeared to stabilize for the remainder of the 60 min measurement period. This pattern indicates that hurrying to take a measurement on an excised branch may minimize, but not necessarily avoid changes in gas exchange rates caused by excision. At Ft. Sherman, the three species with the highest A for attached leaves, A. membranacea, J. copaia, and V. ferruginea (Table 6-1), also maintained the highest A at 60 min after excision (Figure 6-1). The other four species at Ft. Sherman, A. cruenta, B. utile, M. bidentata, and S. amara, suffered a greater reduction in A 60 min after cutting (Figure 6-1) and all produce latex or resin in the stem. A. cruenta, B. utile and M. bidentata After excision, all species showed a reduction in both A and g s within 3 min (Figures 6-1 and 6-2). However, after 10 min, most species showed fewer fluctuations and appeared to stabilize for the remainder of the 60 min measurement period. This pattern indicates that hurrying to take a measurement on an excised branch may minimize, but not necessarily avoid changes in gas exchange rates caused by excision. At Ft. Sherman, the three species with the highest A for attached leaves, A. membranacea, J. copaia, and V. ferruginea (Table 6-1), also maintained the highest A at 60 min after excision (Figure 6-1). The other four species at Ft. Sherman, A. cruenta, B. utile, M. bidentata, and S. amara, suffered a greater reduction in A 60 min after cutting (Figure 6-1) and all produce latex or resin in the stem. A. cruenta, B. utile and M. bidentata are from plant families well-known for this characteristic: the Apocynaceae, Moraceae, Sapotaceae, respectively. Simarouba amara has a clear yellow resin, also characteristic of its family, the Simaroubaceae (Croat 1978). Wounding stimulates latex and resin production (Kramer and Kozlowski 1979), which acts as a wound response and can

PAGE 93

83 physically clog the xylem and prevent water loss and pathogen attack. However, once an excised stem is in water, if latex or resin leaks in the vicinity of the cut section of active xylem it can clog the xylem thus inhibiting water supply to the leaf and reducing A and g s Therefore, these data suggest that species with latex or resin do not perform well following excision. At Parque Metropolitano, L. seemannii maintained a mean of 86.5% of intact photosynthetic rates 60 min after excision (Figure 6-1), whereas P. septenatum maintained only 13.4 and 5.5% of the average intact values for A and g s respectively, after excision (Figs. 6-1 and 6-2). A. excelsum, maintained only 26.6 and 48.6% of intact A and g s respectively after excision. With the exception of P. septenatum, species at Parque maintained higher percentages of intact g s than A 60 min after excision (Figs. 6-1 and 6-2). In contrast, all species in the wetter forest at Ft. Sherman maintained a higher percentage of intact A, relative to g s 60 min after excision (Figs. 6-1 and 6-2). This trend may be indicative of greater stomatal sensitivity in wet forest trees as a result of acclimation to humid conditions. Response to Light Photosynthetic light response curves following excision varied among species (Figure 6-3). Species such as A. membranacea and L. seemannii showed no differences between intact and excised branches, whereas B. utile and S. amara showed substantial differences in light-saturated photosynthesis (Figure 6-3). Although the curve from the excised branch was always lower, this difference was less than the variation between leaves from different branches, so excision did not significantly reduce A as a function of PFD in any of the six species investigated (Table 6-2). Therefore, the light response curve of any excised leaf is likely to reach lower values than measurements on the same leaf

PAGE 94

84 when it is attached, but these lower values are also likely to fall within the range of values for attached leaves. Excision decreased the calculated value of A max in A. cruenta, B. utile, and S. amara (Table 6-3) but not in the other three species investigated. Since A max is probably the most widely used parameter derived from light response curves and half of the species tested showed significant reductions upon excision, estimates of A max derived from light response curves on excised foliage must be interpreted with caution. Minor differences between , and R d from intact and excised light response curves were also observed in some cases (data not shown), but these differences were generally within the 95% confidence interval of the parameter in curve fitting. Overall, most species appeared to perform better during light response curves than over time course measurements. Maintaining a portion of the leaf in the constant environment of the cuvette may help maintain consistency. How a leaf's ability to respond to short-term changes in light and VPD is affected following excision is unknown, but measurements suggest that constant conditions may result in more accurate measurements. (Reich et al. 1995) kept excised branches in sunlight, which is likely to have maintained the leaf in an induced state. In that study, attached foliage from pioneer species was compared with excised foliage from mature canopy species and it was found that the excised foliage of canopy species produced lower gas exchange rates. Since excision can negatively affect gas exchange rates in a systematic way, it is difficult to discern whether the lower gas exchange rates of canopy trees observed by Reich et al. (1995) are the result of excision-induced effects or represent characteristics that have evolved to specific environmental regimes. In either case, the species-specific responses

PAGE 95

85 to excision presented in this paper suggest that it is important to consider the sensitivity to excision when comparing species or functional groups. Response to CO 2 The effect of branch excision on leaf photosynthetic CO 2 responses was again species specific (Figure 6-4). Excision changed the general shape of the curve in A. cruenta B. utile, and S. amara, whereas in A. membranacea, L. seemannii, and V. ferruginea, lower maxima were the main effect. L. seemannii and V. ferruginea were the only species exhibiting a significant effect of excision on photosynthetic responses to p i (Table 6-2). In the other four species, A as a function of p i was reduced following excision, but the magnitude of the effect of excision was small compared to the range of values encountered between branches and therefore, no statistical effect was detected. The reduction in A during CO 2 response curves following excision may be indicative of non-uniform stomatal closure. Heterogeneous gas exchange over small areas of the leaf has been shown to occur in response to water stress (Downton et al. 1988). Heterogeneous stomatal closure has also been implicated in causing errors in estimating p i and may explain observed reductions in A even when p i does not appear to be limiting (Mansfield et al. 1990). Cheeseman (1991) demonstrated through simulation models that the effects of heterogeneous stomatal behaviour can only account for minor effects in calculating p i and therefore, on CO 2 response curves. Yet, the effects of patchy stomatal closure on CO 2 response curves presented by (Cheeseman 1991) are similar to the patterns of reduction following excision in this study, suggesting that stem water status is altered by excision and the reduction in g s is heterogeneous over the leaf surface. Therefore, non-uniform stomatal behaviour may explain why for several species the main effect of excision on the CO 2 response curve was a slight reduction of the maximum. This

PAGE 96

86 reduction probably reflects changes in stomatal behaviour rather than a reduction of the biochemical functioning of the photosystem. Excision reduced the calculated values of both Vc max and J max in L. seemannii and S. amara whereas in A. membranacea and A. cruenta only the calculation of Vc max was significantly affected (Table 6-3). In spite of multi-fold variation in Vc max and J max among species, and in some cases between treatments, there was strong correlation between these two parameters (r 2 =0.89). Excision did not produce a significantly different relationship (t=-0.174; P>0.25) between J max and Vc max so a single linear regression was used. A slightly positive intercept of 13.7 mol m -2 s -1 was observed, along with a slope of 1.63. These results are consistent with those reported in a review of 109 species (Wullschleger 1993). In that study, Wullschleger (1993) also reported linear regression between J max and Vc max with a positive intercept and a slope of 1.64. The consistency of the relationship between J max and Vc max suggests that effects on water transport and patchy stomatal behavior are responsible for the differences in calculating biochemical parameters, rather than changes in biochemical activity. Declining values of A at high p i in CO 2 response curves on both attached and excised branches indicate an apparent limitation by triose phosphate utilization (Sharkey 1985) in L. seemannii and V. ferruginea (Figure 6-4). Limitation of A by TPU was once thought to occur rarely, but there is now increasing evidence that this limitation may be more common under conditions of high light and high p i (Harley et al. 1992; Wullschleger 1993). Tropical forest canopies often experience high light and our data indicate that TPU limitation can be important for at least two species in these forests.

PAGE 97

87 This finding has implications for modeling the responses of tropical forest to increases in atmospheric CO 2 Conclusions I investigated the utility of a technique for measuring gas exchange on excised foliage in tropical forest canopy trees. Responses to excision were species specific and there is some evidence that responses are related to phylogeny, potentially allowing prediction of how a species might perform. All species with latex or resin performed relatively poorly after excision and I suggest that these types of wound responses can clog cut xylem ends, thus reducing water transport. Patterns of reduction of A and g s in time course measurements, light and CO 2 response curves appear to be mediated by stomatal closure as opposed to reductions in biochemical photosynthetic capacity. I recommend individual trials on all species to be investigated. It appears that anatomical, morphological, and possibly, phylogenetic differences can influence the response of gas exchange measurements to excision. Therefore, patterns of response to excision at taxonomic levels higher than species, and different growth forms, such as trees, vines and shrubs, should also be considered in broad species comparisons of photosynthetic measurements on excised branches.

PAGE 98

88 Table 6-1. Light-saturated rate of net CO 2 assimilation per unit mass (A mass ) and area (A area ), leaf diffusive conductance (g s ), and specific leaf area (SLA) 1 S.D. for 10 tropical tree species from the Republic of Panama. Species A mass (nmol g -1 s -1 ) A area (mol m -2 s -1 ) g s (mmol m -2 s -1 ) SLA (cm 2 g -1 ) Parque Metropolitano Anacardium exelsum 65.1 19.1 7.0 2.0 129.7 37.7 93.1 6.8 Luehea seemannii 114.8 12.6 14.8 1.6 592.2 140.0 79.5 11.9 Pseudobombax septenatum 146.8 20.3 12.4 1.7 347.0 57.1 118.6 7.7 Fort Sherman Apeiba membranacea 200.4 16.1 16.1 1.3 659.7 103.6 124.8 9.6 Aspidosperma cruenta 104.5 5.3 14.2 0.7 387.3 68.9 74.4 7.3 Brosimum utile 94.5 14.2 10.4 1.6 298.2 83.8 91.2 4.2 Jacaranda copaia 157.8 27.1 14.6 2.5 538.0 212.7 96.6 6.4 Manilkara bidentata 58.1 16.6 11.8 3.4 199.3 83.9 49.4 1.7 Simarouba amara 106.3 11.4 15.7 1.7 443.8 113.2 68.1 5.4 Vochysia ferruginea 134.5 7.5 15.2 0.8 670.2 83.9 89.5 9.7 Table 6-2. Repeated measures analysis of variance for the effects of excision on photosynthetic light and CO 2 response curves for 6 canopy tree species from the Republic of Panama. Light CO 2 Species SS F P SS F P Apeiba membranacea 13.20 4.38 0.171 0.22 2.98 0.227 Aspidosperma cruenta 9.27 12.90 0.173 50.93 40.84 0.099 Brosimum utile 10.36 2.46 0.361 76.13 1.35 0.452 Luehea seemannii 3.75 4.96 0.156 16.24 523 0.002 Simarouba amara 13.17 2.21 0.377 404.86 3.47 0.314 Vochysia ferruginea 4.58 18.30 0.146 13.82 4.97 0.038

PAGE 99

89 Table 6-3. Means of model parameters from photosynthetic light and CO 2 response curves on leaves of intact (IN) and excised (EX) branches determined by least-squared estimates. Values of light-saturated rate of gross photosynthesis (A max ) from light response curves, and maximum rate of carboxylation (Vc max ) and electron transport (J max ) from CO 2 response curves followed by the same letter are not significantly different (P<0.05) based on Duncans multiple range test. Comparisons are only valid within species. Species Method A max (mol m -2 s -1 ) Vc max (mol m -2 s -1 ) J max (mol m -2 s -1 ) Apeiba membranacea IN EX 20.0 a 18.9a 51.9 a 42.5b 102.9 a 86.3a Aspidosperma cruenta IN EX 16.0 a 13.8b 56.2 a 45.1b 92.3 a 81.4a Brosimum utile IN EX 12.8 a 9.8b 35.0 a 29.2a 64.5 a 54.5a Luehea seemannii IN EX 14.9 a 13.3a 38.3 a 27.5b 82.7 a 65.2b Simarouba amara IN EX 17.9 a 15.6b 61.4 a 32.9b 108.8 a 59.9b Vochysia ferruginea IN EX 16.6 a 14.8a 50.9 a 53.6a 108.1 a 104.0a

PAGE 100

90 020406080100 Fort ShermanParque MetropolitanoTime after cut (min) 0102030405060 % Intact photosynthesis 020406080100 Figure 6-1. Percent of maximum photosynthesis of 3-5 leaves from excised branches at two canopy crane sites: Fort Sherman and Parque Metropolitano, Republic of Panama. Species at Fort Sherman: (filled circle) Apeiba membranacea, (open circle) Aspidosperma cruenta,(filled upside down triange) Brosimum utile, (open upside down triange) Jacaranda copaia, (filled square) Manilkara bidentata, (open square) Simarouba amara var. typica and (filled diamond) Vochysia ferruginea. Species at Parque Metropolitano: (open diamond) Anacardium excelsum, (filled triange) Luehea seemannii, and (open triangle) Pseudobombax septenatum.

PAGE 101

91 020406080100 Time after cut (min) 0102030405060 % Intact stomatal conductance 020406080100 Parque MetropolitanoFort Sherman Figure 6-2. Percent of maximum stomatal conductance of 3-5 leaves from excised branches at two canopy crane sites: Fort Sherman and Parque Metropolitano, Republic of Panama. Symbols as in Figure l.

PAGE 102

92 050010001500 -404812 Vochysia ferrugineaPAR (mol m-2 s-1) 050010001500 051015 Aspidosperma cruenta 050010001500 051015 Luehea seemannii Apeiba membrancea 050010001500 051015 Intact Excised Simarouba amara 050010001500 -505101520 Brosimum utile 050010001500 Net CO2 assimilation (mol m-2 s-1) -4-20246810 Figure 6-3. Representative curves of net CO 2 assimilation as a function of PFD, under 370 mol mol -1 CO 2 and ambient temperature (30-33 C) and relative humidity (70-85%) for leaves of intact and excised branches.

PAGE 103

93 Luehea seemannii 020406080100120140160 -10-50510152025 Simarouba amara 020406080100120140 -10-5051015202530 020406080100120140160 -10-50510152025 Intact Excised Apeiba membranacea Vochysia ferrugineaInternal CO2 partial pressure (Pa) 020406080100120140 -50510152025 Aspidosperma cruenta 020406080100120140 -10-50510152025 020406080100120140 Net CO2 assimilation (mol m-2 s-1) -505101520 Brosimum utile Figure 6-4. Representative curves of net CO 2 assimilation as a function of internal CO 2 partial pressure (p i ) under 1700 mol m -2 s -1 PFD, and ambient temperature (30-33 C) and relative humidity (70-85%) for leaves of intact and excised branches.

PAGE 104

CHAPTER 7 CONCLUSIONS This dissertation emphasizes the relationship between leaf physiological traits and effects of plant species and communities on processes at the ecosystem scale. It appears that plant allocation to leaf photosynthetic capacity has the potential to reveal itself at larger spatial and temporal scales. The leaf traits evaluated in this dissertation have been shown to be interrelated and appear to have evolved in response to specific environmental regimes (Chapin 1980; Reich et al. 1992). The result that there is a relationship between leaf physiology and leaf litter decomposition suggests that the body of ecological literature on plant traits and life history characteristics may be extended to understand the effects of species on ecosystem processes. 94

PAGE 105

LIST OF REFERENCES Ackerly DD, Reich PB (1999) Convergence and correlations among leaf size and function in seed plants: a comparative test using independent contrasts. American Journal of Botany 86:1272-1281 ACP (2002) Panama Canal Authority, Meteorology and Hydrology Branch, http://www.stri.org/tesp/ACP.htm. November 2002, Republic of Panama Andrade JL, Meinzer FC, Goldstein G, Holbrook NM, Caverier J, Jackson P, Silvera K (1998) Regulation of water flux through trunks, branches and leaves in trees of a lowland tropical forest. Oecologia 115:463-471 Arntz AM, Delph LF (2001) Pattern and process: evidence for the evolution of photosynthetic traits in natural populations. Oecologia 127:455-467 Austin AT, Vitousek PM (1998) Nutrient dynamics on a precipitation gradient in Hawai'i. Oecologia 113:519-529 Binkley D, Matson P (1983) Ion exchange resin bag method for assessing forest soil nitrogen availability. Soil Science Society of America Journal 47:1050-1052 Bjrkman O, Demmig B (1987) Photon yield of O 2 evolution and chlorophyll fluorescence characteristics at 77K among vascular plants of diverse origins. Planta 170:489-504 Boari F, Malone M (1993) Wound-induced hydraulic signals: Survey of occurrence in a range of species. Journal of Experimental Botany 44:741-746 Borchert R (1994) Water status and development of tropical trees during seasonal drought. Trees 8:115-125 Chabot BF, Hicks DJ (1982) The ecology of leaf life spans. Annual Review of Ecology and Systematics 13:229-259 Chapin FS, III (1980) The mineral nutrition of wild plants. Annual Review of Ecology and Systematics 11:233-260 Chapin FS, III (1993) Functional role of growth forms in ecosystem and global processes. In: Ehleringer JR, Field CB (eds) Scaling Physiological Processes. Academic Press, New York, pp 287-308 95

PAGE 106

96 Chapin FS, III, Shaver GR, Giblin AE, Nadelhoffer KJ, Laundre JA (1995) Responses of Arctic tundra to experimental and observed changes in climate. Ecology 76:694-711 Cheeseman JM (1991) PATCHY: simulating and visualizing the effects of stomatal patchiness on photosynthetic CO 2 exchange studies. Plant, Cell and Environment 14:593-599 Choong MF, Lucas PW, Ong JSY, Pereira B, Tan HTW, Turner IM (1992) Leaf fracture toughness and sclerophylly: their correlations and ecological implications. New Phytologist 121:597-610 Clark DA, Brown S, Kicklighter DW, Chambers JQ, Thomlinson JR, Ni J (2001a) Measuring net primary production in forests: Concepts and field methods. Ecological Applications 11:356-370 Clark DA, Brown S, Kicklighter DW, Chambers JQ, Thomlinson JR, Ni J, Holland EA (2001b) Net primary production in tropical forests: an evaluation and synthesis of existing field data. Ecological Applications 11:371-384 Coley PD (1983) Herbivory and defensive characteristics of tree species in a lowland tropical forest. Ecological Monographs 53:209-233 Coley PD, Bryant JP, Chapin FS, III (1985) Resource availability and plant anti-herbivore defense. Science 230:895-899 Condit R (1998) Ecological implications of changes in drought patterns: Shifts in forest composition in Panama. Climate Change 39:413-427 Condit R, Pitman N, Leigh EG, Chave J, Terborgh J, Foster RB, Nuez P, Aguilar S, Valencia R, Villa G, Muller-Landau HC, Losos E, Hubbell SP (2002) Beta-diversity in tropical forest trees. Science 295:666-669 Cornelissen JHC (1996) An experimental comparison of leaf decomposition rates in a wide range of temperate plant species and types. Journal of Ecology 84:573-582 Cornelissen JHC, Thompson K (1997) Functional leaf attributes predict litter decomposition rate in herbaceous plants. New Phytologist 135:109-114 Cornelissen JHC, Werger MJA, Castro-Dez P, vanRheenen JWA, Rowland AP (1997) Foliar nutrients in relation to growth, allocation and leaf traits in seedlings of a wide range of woody plant species and types. Oecologia 111:460-469 Crews TE, Kitayama K, Fownes JH, Riley RH, Herbert DA, Mueller-Dombois D, Vitousek PM (1995) Changes in soil phosphorus fractions and ecosystem dynamics across a long soil chronosequence in Hawaii. Ecology 76:1407-1424

PAGE 107

97 Croat TB (1978) Flora of Barro Colorado Island. Stanford University Press, Stanford, California Cunningham SA, Summerhayes B, Westoby M (1999) Evolutionary divergences in leaf structure and chemistry, comparing rainfall and soil nutrient gradients. Ecological Monographs 69:569-588 Dang Q-L, Margolis HA, Coyea MR, Sy M, Collatz GJ (1997) Regulation of branch-level gas exchange of boreal trees: roles of shoot water potential and vapor pressure difference. Tree Physiology 17:521-535 Deka RN, Wairiu M, Mtakwa PW, Mullins CE, Veenendaal EM, Townsend J (1995) Use and accuracy of the filter-paper technique for measurement of soil matric potential. European Journal of Soil Science 46:233-238 Delieu T, Walker D (1981) Polarographic measurement of photosynthetic oxygen evolution by leaf-disks. New Phytologist 89:165-178 Dietrich WE, Windsor DM, Dunne T (1982) Geology, climate, and hydrology of Barro Colorado Island. In: Leigh EG, Rand AS, Windsor DM (eds) The Ecology of a Tropical Forest. Smithsonian Institution Press, Washington D.C., pp 21-46 Downton WJS, Loveys BR, Grant WJR (1988) Stomatal closure fully accounts for the inhibition of photosynthesis by abscisic acid. New Phytologist 108:263-266 Dudley SA (1996) Differing selection on plant physiological traits in response to environmental water availability: a test of adaptive hypotheses. Evolution 50:92-102 Eamus D, Prior L (2001) Ecophysiology of trees of seasonally dry tropics: comparisons among phenologies. Advances in Ecological Research 32:113-197 Enquist BJ, Brown JH, West GB (1998) Allometric scaling of plant energetics and population density. Nature 395:163-166 Evans RD (2001) Physiological mechanisms influencing plant nitrogen isotope composition. Trends in Plant Science 6:121-126 Ewel JJ, Bigelow SW (1996) Plant Life-Forms and Tropical Ecosystem Functioning. In: Orians, Dirzo, Cushman (eds) Biodiversity and Ecosystem Functioning in Tropical Forests. Springer-Verlag, Berlin, pp 101-126 Ewers FW, Fisher JB (1989) Techniques for measuring vessel lengths and diameters in stems of woody plants. American Journal of Botany 76:645-656 Farquhar GD, Richards RA (1984) Isotopic composition of plant carbon correlates with water-use efficiency of wheat genotypes. Australian Journal of Plant Physiology 33:539-552

PAGE 108

98 Field CB (1991) Ecological scaling of carbon gain to stress and resource availability. In: Mooney HA, Winner WE, Pell EJ (eds) Response of Plants to Multiple Stresses. Academic Press, New York, pp 35-65 Geber MA (1990) The cost of meristem limitation in Polygonum arenastrum: negative genetic correlations between fecundity and growth. Evolution 44:799-819 Gentry AH (1988) Changes in plant community diversity and floristic composition on environmental and geographical gradients. Annals of the Missouri Botanical Garden 75:1-34 Gerrish G (1992) Changing photosynthetic capacity during leaf ontogeny in juvenile and mature Metrosideros polymorpha trees. Pacific Science 46:315-324 Ginn SE, Seiler JR, Cazell BH, Kreh RE (1991) Physiological and growth responses of eight-year-old pine stands to thinning. Forest Science 37:1030-1040 Givnish TJ (2002) Adaptive significance of evergreen vs. deciduous leaves: solving the triple paradox. Silva Fennica 36:703-743 Goldstein G, Andrade JL, Meinzer FC, Holbrook NM, Cavelier J, Jackson P, Celis A (1998) Stem water storage and diurnal patterns of water use in tropical forest canopy trees. Plant, Cell and Environment 21:397-406 Goldstein G, Sarmiento G, Meinzer FC (1986) Patrones diarios y estacionales en las relaciones hdricas de rboles siempreverdes de la sabana tropical. Oecologia Plantarum 7:107-119 Graham EA, Mulkey SS, Kitajima K, Phillips NG, Wright SJ (2003) Cloud cover limits net CO 2 uptake and growth of a rainforest tree during tropical rainy seasons. Proceedings of the National Academy of Sciences 100:572-576 Grime JP (1977) Evidence for the existence of three primary strategies in plants and its relevance to ecological and evolutionary theory. American Naturalist 111:1169-1194 Grime JP, Cornelissen JHC, Thompson K, Hodgson JG (1996) Evidence of a causal connection between anti-herbivore defence and the decomposition rate of leaves. Oikos 77:489-494 Hacke UG, Sperry JS, Pockman WT, Davis SD, McCulloh KA (2001) Trends in wood density and structure are linked to prevention of xylem implosion by negative pressure. Oecologia 126:457-461 Harley PC, Thomas RB, Reynolds JF, Strain BR (1992) Modelling photosynthesis of cotton grown in elevated CO 2 Plant, Cell and Environment 15:271-282

PAGE 109

99 Haxeltine A, Prentice IC (1996) BIOME 3: An equilibrium terrestrial biosphere model based on ecophysiological constraints, resource availability and competition among plant functional types. Global Biogeochemical Cycles 10:693-709 Hobbie SE (1992) Effects of Plant Species on Nutrient Cycling. Trends in Ecology and Evolution 7:336-339 Hobbie SE (1996) Temperature and plant species control over litter decomposition in Alaskan tundra. Ecological Monographs 66:503-522 Holbrook NM, Whitbeck JL, Mooney HA (1995) Drought responses of neotropical dry forest trees. In: Bullock HA, Mooney HA, Medina E (eds) Seasonally Dry Tropical Forests. Cambridge University Press, Cambridge, pp 243-276 Kikuzawa K (1991) A cost-benefit-analysis of leaf habit and leaf longevity of trees and their geographic pattern. American Naturalist 138:1250-1263 Krner C (1993) Scaling from species to vegetation: The usefullness of functional groups. In: Schulze E-D, Mooney HA (eds) Biodiversity and Ecosystem Function. Springer-Verlag, Berlin, pp 117-140 Koyama H (1981) Photosynthetic rates in lowland rain forest trees of peninsular Malaysia. Japanese Journal of Ecology 31:361-369 Kramer PJ, Kozlowski TT (1979) Physiology of Woody Plants. Academic Press, New York Kppers M (1984) Carbon relations and competition between woody species in a central European hedgerow II. Stomatal responses, water use, and hydraulic conductivity in the root/leaf pathway. Oecologia 64:344-354 Kursar TA, Wright SJ, Radulovich R (1995) The effects of the rainy season and irrigation on soil water and oxygen in a seasonal forest in Panama. Journal of Tropical Ecology 11:479-516 Lambers H, Chapin FS, III, Pons TL (1998) Plant Physiological Ecology. Springer-Verlag, New York Lechowicz MJ (1984) The effects of individual variation in physiological and morphological traits on the reproductive capacity of the common cocklebur, Xanthium strumarium L. Evolution 38:833-844 LIDET (1995) Long-term Intersite Decomposition Experiment Team (LIDET) Meeting the Challenges of Long-Term, Broad Scale Ecological Experiments. LTER, Seattle, WA, p 23

PAGE 110

100 Mansfield TA, Hetherington AM, Atkinson CJ (1990) Some current aspects of stomatal physiology. Annual Review of Plant Physiology and Plant Molecular Biology 41:55-75 Medina E, Francisco M (1994) Photosynthesis and water relations of savanna tree species differing in leaf phenology. Tree Physiology 14:1367-1381 Meentemeyer V (1978) Microclimate and lignin control of litter decomposition rates. Ecology 59:465-472 Meinzer FC (2003) Functional convergence in plant responses to the environment. Oecologia 134:1-11 Meinzer FC, Andrade JL, Goldstein G, Holbrook NM, Cavelier J, Jackson P (1997) Control of transpiration from the upper canopy of a tropical forest: the role of stomatal, boundary layer and hydraulic architecture components. Plant, Cell and Environment 20:1242-1253 Meinzer FC, Goldstein G (1996) Scaling up from leaves to whole plants and canopies for photosynthetic gas exchange. In: Mulkey SS, Chazdon RL, Smith AP (eds) Tropical Forest Plant Ecophysiology. Chapman & Hall, New York, pp 114-138 Meinzer FC, Grantz DA (1990) Stomatal and hydraulic conductance in growing sugarcane: stomatal adjustment to water transport capacity. Plant, Cell and Environment 13:383-388 Meinzer FC, Grantz DA (1991) Coordination of stomatal, hydraulic, and canopy boundary layer properties: Do stomata balance conductances by measuring transpiration? Physiologia Plantarum 83:324-329 Melillo JM, Aber JD, Muratore JF (1982) Nitrogen and lignin control of hardwood leaf litter decomposition dynamics. Ecology 63:621-626 Miller AJ, Schuur EAG, Chadwick OA (2001) Redox control of phosphorus pools in Hawaiian montane forest soils. Geoderma 102:219-237 Monk CD (1966) An ecological significance of evergreenness. Ecology 47:504-505 Mooney HA, Ferrar PJ, Slatyer RO (1978) Photosynthetic capacity and carbon allocation patterns in diverse growth forms of Eucalyptus. Oecologia 36:103-111 Mulkey SS, Kitajima K, Wright SJ (1995) Photosynthetic capacity and leaf longevity in the canopy of a dry tropical forest. Selbyana 16:169-173 Mulkey SS, Kitajima K, Wright SJ (1996) Plant physiological ecology of tropical forest canopies. Trends in Ecology and Evolution 11:408-412

PAGE 111

101 Nadelhoffer KJ, Fry B (1988) Controls on natural nitrogen-15 and carbon-13 in forest soil organic matter. Soil Science Society of America Journal 52:1633-1640 Neilson RP, Marks D (1994) A global perspective of regional vegetation and hydrologic sensitivities from climate change. Journal of Vegetation Science 5:715-730 Niinemets (2001) Global-scale climate controls of leaf dry mass per area, density, and thickness in trees and shrubs. Ecology 82:453-469 Ostertag R (2001) Effects of nitrogen and phosphorus availability on fine-root dynamics in Hawaiian montane forests. Ecology 82:485-499 Ostertag R, Hobbie SE (1999) Early stages of root and leaf decomposition in Hawaiian forests: effects of nutrient availability. Oecologia 121:564-573 Pataki DE, Oren R, Phillips N (1998) Responses of sap flux and stomatal conductance of Pinus taeda L. trees to stepwise reductions in leaf area. Journal of Experimental Botany 49:871-876 Paton S, Wright SJ (2003) Terrestrial Environmental Science Program http:/www.stri.org/tesp/. January 2003. Smithsonian Tropical Research Institute, Balboa, Republic of Panama Pitman NCA, Terborgh JW, Silman MR, Nuez P, Neill DA, Cern CE, Palacios WA, Aulestia M (2001) Dominance and distribution of tree species in upper Amazonian terre firme forests. Ecology 82:2101-2117 Poorter H, Remkes C (1990) Leaf area ratio and net assimilation rate of 24 wild species differing in relative growth rate. Oecologia 83:553-559 Prado CHBA, DeMoraes JAPV (1997) Photosynthetic capacity and specific leaf mass in twenty woody species of Cerrado vegetation under field conditions. Photosynthetica 33:103-112 Pyke CR, Condit R, Aguilar S, Lao S (2001) Floristic composition across a climatic gradient in a neotropical lowland forest. Journal of Vegetation Science 12:553-566 Radulovich R, Sollins P (1991) Nitrogen and phosphorus leaching in zero-tension drainage from a humid tropical soil. Biotropica 23:84-87 Reich PB (1995) Phenology of tropical forests: patterns, causes, and consequences. Canadian Journal of Botany 73:164-174 Reich PB, Borchert R (1984) Water stress and tree phenology in a tropical dry forest in the lowlands of Costa Rica. Journal of Ecology 72:61-74

PAGE 112

102 Reich PB, Ellsworth DS, Uhl C (1995) Leaf carbon and nutrient assimilation and conservation in species of differing successional status in an oligiotrophic Amazonian forest. Functional Ecology 9:65-76 Reich PB, Ellsworth DS, Walters MB, Vose JM, Gresham C, Volin JC, Bowman WD (1999) Generality of leaf trait relationships: a test across six biomes. Ecology 80:1955-1969 Reich PB, Walters MB, Ellsworth DS (1997) From tropics to tundra: Global convergence in plant functioning. Proceedings of the National Academy of Science 94:13730-13734 Reich PB, Walters MB, Ellsworth DS, Vose JM, Volin JL, Greshan CA, Bowman WD (1998) Relationships of leaf dark respiration to leaf nitrogen, specific leaf area and leaf life-span: a test across biomes and functional groups. Oecologia 114:471-482 Reich PB, Walters MB, Ellsworth ES (1992) Leaf life-span in relation to leaf, plant, and stand characteristics among diverse ecosystems. Ecological Monographs 62:365-392 Riley RH, Vitousek PM (1995) Nutrient dynamics and nitrogen trace gas flux during ecosystem development in montane rain-forest. Ecology 76:292-304 Running SW, Coughlan JC (1988) A general model for forest ecosystem processes for regional application. I. Hydrologic balance, canopy gas exchange, and primary production processes. Ecological Modeling 42:125-154 Ryan MG, Melillo JM, Ricca A (1989) A comparison of methods for determining proximate carbon fractions of forest litter. Canadian Journal of Forest Research 20:166-171 Samuelson LJ (1998) Influence of intensive culture on leaf net photosynthesis and growth of sweetgum and loblolly pine seedlings. Forest Science 44:308-316 SAS (1985) SAS user's guide. SAS Institute, Cary, North Carolina, USA Schimper AFW (1898) Plant Geography upon a Physiological Basis. Weinheim, J. Cramer, New York Schuur EAG, Chadwick OA, Matson PA (2001) Carbon cycling and soil carbon storage in mesic to wet Hawaiian montane forests. Ecology 82:3182-3196 Schuur EAG, Matson PA (2001) Net primary productivity and nutrient cycling across a mesic to wet precipitation gradient in Hawaiian montane forest. Oecologia 128:431-442 Scott NA, Binkley D (1997) Foliage litter quality and annual net N mineralization: comparison across North American forest sites. Oecologia 111:151-159

PAGE 113

103 Sharkey TD (1985) Photosynthesis in intact leaves of C 3 plants: physics, physiology and rate limitation. The Botanical Review 51:53-105 Shearer G, Kohl DH (1986) N 2 -fixation in field settings: estimations based on natural 15 N abundance. Australian Journal of Plant Physiology 13:699-756 Sims DA, Pearcy RW (1991) Photosynthesis and respiration in Alocasia macrorrhiza following transfers to high and low light. Oecologia 86:447-453 Sobrado MA (1986) Aspects of tissue water relations and seasonal changes of leaf water potential components of evergreen and deciduous species coexisting in tropical dry forests. Oecologia 68:413-416 Sperry JS, Alder NN, Eastlack SE (1993) The effect of reduced hydraulic conductance on stomatal conductance and xylem cavitation. Journal of Experimental Botany 44:1075-1082 Sperry JS, Donnelly JR, Tyree MT (1988) A method for measuring hydraulic conductivity and embolism in xylem. Plant, Cell and Environment 11:35-40 Sperry JS, Pockman WT (1993) Limitation of transpiration by hydraulic conductance and xylem cavitation in Betula occidentalis. Plant, Cell and Environment 16:279-287 Stahlberg R, Cosgrove DJ (1995) Comparison of electric and growth responses to excision in cucumber and pea seedlings. II. Long-distance effects are caused by the release of xylem pressure. Plant, Cell and Environment 18:33-41 Tardieu F, Davies WJ (1993) Integration of hydraulic and chemical signalling in the control of stomatal conductance and water status of droughted plants: opinion. Plant, Cell & Environment 16:341-349 Tilman D (1988) Plant Strategies and the Dynamics and Structure of Plant Communities. Princeton University Press, Princeton, New Jersey Tilman D, Knops J, Wedin D, Reich P, Ritchie M, Siemann E (1997) The influence of functional diversity and composition on ecosystem processes. Science 277:1300-1302 Torn M, Trumbore SE, Chadwick OA, Vitousek PM, Hendricks DM (1997) Mineral control of soil organic carbon storage and turnover. Nature 389:170-173 Vzquez JA, Givnish T (1998) Altitudinal gradients in tropical forest composition, structure, and diversity in the Sierra de Manantln. Journal of Ecology 86:999-1020 Veenendaal EM, Swaine MD, Blay D, Yelifari NB, Mullins CE (1996) Seasonal and long-term soil water regime in West African tropical forest. Journal of Vegetation Science 7:473-482

PAGE 114

104 Vitousek PM (1995) The Hawaiian Islands as a model system for ecosystem studies. Pacific Science 49:2-16 Vitousek PM, Sanford RL (1986) Nutrient cycling in moist tropical forest. Annual Review of Ecology and Systematics 17:137-167 Vitousek PM, Turner DR, Parton WJ, Sanford RL (1994) Litter decomposition on the Mauna Loa environmental matrix, Hawaii: patterns, mechanisms, and models. Ecology 75:418-429 von Caemmerer S, Farquhar GD (1981) Some relationships between the biochemistry of photosynthesis and the gas exchange of leaves. Planta 153:376-387 von Ende CN (1993) Repeated-measures analysis: growth and other time-dependent measures. In: Guretitch J, Scheiner, SM (eds) Design and Analysis of Ecological Experiments. Chapman & Hall, New York, pp 113-137 Wardle DA, Barker GM, Bonner KI, Nicholson KS (1998) Can comparative approaches based on plant ecophysiological traits predict the nature of biotic interactions and individual plant species effects in ecosystems. Journal of Ecology 86:405-420 Wardle DA, Bonner KI, Barker GM (2002) Linkages between plant litter decomposition, litter quality, and vegetation responses to herbivores. Functional Ecology 16:585-595 Werger MJA, Morris JW (1991) Climatic control of vegetation structure and leaf characteristics along an aridity gradient. Annali di Botanica IL:203-215 West GB, Brown JH, Enquist BJ (1999) A general model for the structure and allometry of plant vascular systems. Nature 400:664-667 Westoby M (1998) A leaf-height-seed (LHS) plant ecology strategy scheme. Plant and Soil 199:213-227 Williams M, Rastetter EB, Fernandes DN, Goulden ML, Wofsy SC, Shaver GR, Melillo JM, Munger JW, Fan S-M, Nadelhoffer KJ (1996) Modelling the soil-plant atmosphere continuum in a Quercus-Acer at Harvard Forest: the regulation of stomatal conductance by light, nitrogen and soil/plant hydraulic properties. Plant, Cell and Environment 19:911-927 Williamson VG, Milburn JA (1995) Cavitation events in cut stems kept in water: implications for cut flower senescence. Scientia Horticulturae 64:219-232 Woodring W, Steward R, Steward J (1980) Geological map of the Panama Canal. U.S. Geological Survey, Reston, VA Woodward FI, Smith TM, Emanuel WR (1995) A global primary productivity and phytogeography model. Global Biogeochemical Cycles 9:471-490

PAGE 115

105 Wright IJ, Westoby M (1999) Differences in seedling growth behaviour among species: trait correlations across species, and trait shifts along nutrient compared to rainfall gradients. Journal of Ecology 87:85-97 Wright SJ (1992) Seasonal drought, soil fertility and the species density of tropical forest plant communities. Trends in Ecology and Evolution 7:260-263 Wullschleger SD (1993) Biochemical limitations to carbon assimilation in C 3 plantsA retrospective analysis of the A/C i curves from 109 species. Journal of Experimental Botany 44:907-920 Yavitt JB (2000) Nutrient dynamics of soil derived from different parent material on Barro Colorado Island, Panama. Biotropica 32:198-207 Yavitt JB, Wieder RK, Wright SJ (1993) Soil nutrient dynamics in response to irrigation of a Panamanian tropical moist forest. Biogeochemistry 19:1-25 Zhang J, Davies WJ (1990) Changes in the concentration of ABA in xylem sap as a function of changing soil water status can account for changes in leaf conductance and growth. Plant, Cell and Environment 13:227-285

PAGE 116

BIOGRAPHICAL SKETCH Louis Stephen Santiago was born on February 13, 1970 in Aiea, Oahu, Hawaii. After several years, he moved with his family to San Francisco Bay, where he was raised. During his youth, Louis spent time playing in the dunes of Half Moon Bay, and walking in the East Bay hills and Sierra Nevada foothills with his father. These activities stimulated an interest in the natural world. For undergraduate studies, Louis attended the University of California at Berkeley. Although he began in Chemistry, he soon switched to Integrative Biology. He had the opportunity to work on several independent projects in the East Bay hills, in the Mojave Desert, and on the island of Moorea, Tahiti, in the South Pacific. These experiences inspired a strong interest in tropical biology. After graduating from the University of California in 1993, Louis began a masters program at the University of Hawaii, under the direction of Dr. Guillermo Goldstein. Louis pursued a masters project in the cloud forests of Waikamoi, Maui. Before graduating from the University of Hawaii, however, he spent five months at Dr. Goldsteins lab in Panama; and began developing the doctoral project presented in this dissertation. Louis then returned to Hawaii, graduated in 1998, and immediately began a doctoral program at the University of Florida with Dr. Stephen Mulkey. After completing his doctoral studies, Louis will return to the University of California at Berkeley for a postdoctoral position in stable isotope ecology. 106


Permanent Link: http://ufdc.ufl.edu/UFE0001358/00001

Material Information

Title: Leaf Traits of Canopy Trees on a Precipitation Gradient in Panama: Integrating Plant Physiological Ecology and Ecosystem Science
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0001358:00001

Permanent Link: http://ufdc.ufl.edu/UFE0001358/00001

Material Information

Title: Leaf Traits of Canopy Trees on a Precipitation Gradient in Panama: Integrating Plant Physiological Ecology and Ecosystem Science
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0001358:00001


This item has the following downloads:


Full Text












LEAF TRAITS OF CANOPY TREES ON A PRECIPITATION GRADIENT IN
PANAMA: INTEGRATING PLANT PHYSIOLOGICAL ECOLOGY AND
ECOSYSTEM SCIENCE
















By

LOUIS STEPHEN SANTIAGO


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA


2003















ACKNOWLEDGMENTS

Many people and organizations contributed to the work presented in this

dissertation, and I am grateful for all of their help. I thank my committee members Steve

Mulkey, Kaoru Kitajima, Jack Putz, Tim Martin and Joe Wright, and everyone in the

Plant Ecology group for five years of stimulating interactions and discussions leading to

the ideas presented here. I am grateful to Guillermo Goldstein, Rick Meinzer, Ted

Schuur, Michelle Mack, Tom Kursar, Lissy Coley, Klaus Winter, Steve Hubbell, and

Allen Herre for comments and discussions on this project at various stages of its

development. Tim Jones, Dave Woodruff, Katy Balatero, Mirna Sarmaniego, Aurelio

Virgo, Augustin Somoza, Elizabeth Osorio, Steve Davis, Eric Graham, Aurelio Virgo,

Kate Moran and Sarah Bouchard helped with lab and field procedures. I thank Edwin

Andrade, Jose Herrera, Oscard and Pitti for their patience and skill in operating the

canopy crane. Rick Condit and Suzanne Lao of CTFS provided generous use of their tree

database. Funding was provided by the U.S. Environmental Protection Agency, the

Smithsonian Tropical Research Institute (STRI), the Florida-Georgia Alliance, the

Mellon Foundation, and the University of Florida. Katia Silvera, the Silvera family and

my parents provided support throughout the project.
















TABLE OF CONTENTS
Page

A C K N O W L E D G M E N T S .................................................................................................. ii

LIST OF TABLES ....................................................... ............ ....... ....... vi

L IST O F F IG U R E S .... ...... ................................................ .. .. ..... .............. vii

ABSTRACT .............. .......................................... ix

CHAPTER

1 IN T R O D U C T IO N ............................................................................. .............. ...

2 LEAF PHOTOSYNTHETIC TRAITS OF CANOPY TREES ALONG A
PRECIPITATION GRADIENT IN LOWLAND TROPICAL FOREST ................3

In tro d u ctio n ..................................................................................................... .... .. 3
M materials and M methods ................................................................. ....................... 5
Site Characterization and Species ........................................ ...... ............... 5
P photosynthesis ...................................... ............................ ............... 7
L eaf Structure and C hem istry..................................................................... .... .8
Tw o-Site Com prisons .......................................................................9
D ata A n aly sis ................................................................... ............... 10
R e su lts ................... ...................................... ................. ................ 1 0
Soil W ater Potential ........................................................ .. .......... 10
Photosynthesis .................................. ........................... .... ........10
Leaf Structure and Chem istry........................................................... ..... ........ .11
Tw o-Site Com prisons ......................................................... ............... 12
D iscu ssio n ...................................... ................................................. 12

3 NUTRIENT CYCLING ON A PRECIPITATION GRADIENT IN LOWLAND
T R O P IC A L F O R E ST ..................................................................... ... ...................25

In tro du ctio n ...................................... ................................................ 2 5
M materials and M methods ....................................................................... ..................27
Study Site............................................. 27
Foliar Chem istry ............... .. .................. ........... ........ .... .. ........ .... 28
Litterfall C collection and Processing ........................................ .....................28
Leaf Litter Chem istry .......................................................... ............... 29
Nutrient Availability .................. ......................... .... ....................... 29









Extractable nitrogen and phosphorus ................................. ...............29
Nitrogen mineralization and nitrification...............................................30
Ion exchange resins ............................ .............. .............. ......... 30
Soil nutrient pools ............................................. ............ ............. 31
R e su lts ................................................................................................................... 3 1
F o liar C h em istry .............................................................................. 3 1
Litter Production and Chem istry ........................................ ...... ............... 31
N u trient A v ailab ility ................................................................. ....................32
Extractable nitrogen and phosphorus ................................. ...............32
Ion ex ch an g e resin s ............................................................ ................ .. 3 3
Soil nutrient pools ....................................... ...... ............ ............. 33
D isc u ssio n ..............................................................................................3 3
Foliar Chemistry .......... .... ......................................................33
Litter Production and Chem istry ........................................ ...... ............... 34
N utrient A availability .................. .......................... .. .. .... .. ........ .... 35
C o n c lu sio n s ................................................................................................... 3 6

4 HYDRAULIC CONDUCTIVITY AND WOOD DENSITY SCALE WITH LEAF
PHOTOSYNTHETIC TRAITS IN PANAMANIAN FORCES CANOPY TREES ....46

Introdu action ...................................... ................................................. 4 6
M materials and M methods ....................................................................... ..................4 8
Study Site and Species............ .... ........................................ .... .. ........... 48
G as Exchange M easurem ents................................................... ........ ....... 49
H ydraulic C onductivity ............................................... ............................ 49
W ood D density .................................................................. ..........50
F o lia r A n a ly sis............................................................................................... 5 0
R e su lts ...........................................................................................5 1
D isc u ssio n ............................................................................................................. 5 2

5 LEAF DECOMPOSITION IN A WET TROPICAL FOREST: LINKING LEAF
TRAITS WITH NUTRIENT CYCLING.................. ........................................59

Intro du action ...................................... ................................................ 59
M materials and M methods ....................................................................... ..................6 1
Stu dy Site and Species............ ........ .................................... ........ ... ........... 6 1
Litter Collection and D ecom position ...................................... ............... 61
Initial Litter Quality .............................................. .. ...... ................. 62
Gas Exchange and Leaf Chemistry .......................................... ............... 62
R e su lts ...................................... .................................................... 6 4
D discussion ................................... .............. .................. ........... 65
Plant Growth Forms and Ecosystem Processes...............................................65
Litter Quality and Decomposition ..... .................... ...............66
Photosynthesis and Decomposition...................... ... ........................ 66










6 TEST OF GAS EXCHANGE MEASUREMENTS ON EXCISED BRANCHES OF
TEN TROPICAL TREE SPECIES: A TECHNICAL REPORT ...............................77

In tro d u ctio n ............................................................................................................ 7 7
M materials and M methods ........................................................................ ..................79
Study Sites and Species .......................................................... ............... 79
G as Exchange M easurem ents..................................... ............................. ....... 80
R esu lts an d D iscu ssion ............................................................. ...................... 82
Tim e C ourse M easurem ents .......................................................................... 82
R esp on se to L ight .............................................................. ......................83
R e sp o n se to C O 2 ................................................................. ... .. .. ... .. .... 8 5
C o n clu sio n s ..................................................... ................ 8 7

7 CON CLU SION S ........................................................ ..............94

L IST O F R E FE R E N C E S ............................................................................. .............. 95

BIOGRAPH ICAL SKETCH .............................................................. ............... 106




































v
















LIST OF TABLES


Table page

2-1. Site characteristics for 1 ha census plots ........................................ ............... 16

2-2. Study species from four sites along precipitation gradient.............................17

2-3. Correlation coefficients between leaf defensive characteristics.............................18

3-1. Characteristics of sites along precipitation gradient........................................38

3-2. Percentage of N and C in canopy sun leaves.................................. ............... 38

3-3. Litterfall rates separated by component............. ........................... ..................39

3-4. Soil chemistry to a depth of 10 cm ............. .. ..... ........................ ............... 40

3-5. Bulk soil chemistry from the surface 10 cm of soil..................... .................41

3-6. Soil bulk density and w ater content....................................... ......................... 41

4-1. Area-based maximum photosynthetic rate (Aarea).......................... .........................55

5-1. Summary of study species and leaf litter decomposition rate (k) ..............................68

5-2. Leaf litter decomposition rates and initial litter quality ...........................................69

5-3. Regressions of litter chemical parameters and leaf litter decomposition .................70

5-4. Regressions of leaf chemical and structural parameters.................. ..................70

5-5. Results of general linear m odel ....................... ...... ......... ........................ 70

6-1. Light-saturated rate of net CO2 assimilation .................. ..... ..................88

6-2. Repeated measures analysis of variance...... ...................... ...............88

6-3. M means of m odel param eters......... .................. .................................. ............... 89
















LIST OF FIGURES


Figure pge

2-1. Seasonal variation in soil water potential ( oil) ....................................... ........... 19

2-2. Relationship between mean annual precipitation and photosynthesis...................20

2-3. Relationship between mean annual precipitation and leaf N concentration .............21

2-4. Relationship between mean annual precipitation and lamina thickness...................22

2-5. Maximum photosynthetic 02 evolution rate (Pmass) .......................................23

2-6. Seasonal variation in leaf water potential ( eaf) ............................... ................. 24

3-1. The difference between seasonal maximum and minimum litterfall rates ...............42

3-2. Relationship between carbon isotope composition of soil and litter........................43

3-3. Soil N mineralization rates for the top 10 cm .................................. ............... 44

3-4. Soil N concentration as a function of (A) litter lignin and (B) litter lignin:N............45

4-1. Leaf photosynthetic rate per unit area ......................... .................... ...............56

4-2. Leaf nitrogen per unit area (Narea)................ ............... .............. ................57

4-3. Stem saturated w ater content......................................................................... ... ... 58

5-1. Relationship between leaf litter decomposition rate (k) and initial litter lignin .........71

5-2. Relationship between leaf litter decomposition rate (k) and initial litter carbon........72

5-3. Relationship between leaf litter decomposition rate (k) and nitrogen........................73

5-4. Relationship between leaf litter decomposition rate (k) and specific leaf area ..........74

5-5. Relationship between Amass specific leaf area and leafN .........................................75

5-6. Schematic diagram depicting how photosynthesis is related to decomposition.........76

6-1. Percent of maximum photosynthesis ........................................................ 90









6-2. Percent of maximum stomatal conductance..........................................................91

6-3. Representative curves of net CO2 assimilation as a function of PFD.........................92

6-4. Representative curves of net CO2 assimilation as a function of internal CO2............93















Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

LEAF TRAITS OF CANOPY TREES ON A PRECIPITATION
GRADIENT IN PANAMA: INTEGRATING PLANT PHYSIOLOGICAL
ECOLOGY AND ECOSYSTEM SCIENCE

By

Louis Stephen Santiago

August 2003

Chair: Stephen S. Mulkey
Major Department: Botany

There is increasing awareness in ecology of the importance of species effects on

processes at the ecosystem scale. This dissertation uses a comparative approach involving

many species along a precipitation gradient (1800-3500 mm/yr) in lowland Panama to

understand how species traits vary among different plant communities, and how these

traits feed back into ecosystem processes such as decomposition and soil nutrient

availability. As precipitation increases from South to North across the Isthmus of

Panama, there is a gradual change in canopy leaf traits from short-lived leaves with high

photosynthetic rates in seasonally dry forest, to relatively long-lived leaves with lower

photosynthetic rates and increased allocation to structural defense in wet forest. Increases

in leaf litter lignin:N also accompany increases in precipitation, indicating a decrease in

potential decomposability of leaf litter in wetter sites. Leaf litter lignin:N was negatively

correlated with soil N mineralization rates, and positively correlated with total soil N

pools indicating that slowly decomposing litter reduces mineralization, but conserves N









in the soil organic matter matrix. Leaf litter lignin:N was the strongest litter quality

predictor of decomposition at the one site where decomposition was studied.

Decomposition was positively related to specific leaf area, leafN concentration and

photosynthetic rate per unit mass suggesting that these traits may be useful predictors of

the effects of species on ecosystem processes. Photosynthetic rate per unit area and

stomatal conductance were positively related to leaf specific hydraulic conductivity and

negatively related to branch wood density indicating that leaf traits controlling gas

exchange correlate with processes at the branch and whole plant levels of organization.

Overall this dissertation provides evidence that many plant traits are correlated along a

minimal number of axes, and that these traits can be used to predict the movement of

matter and energy between plants and their environments.














CHAPTER 1
INTRODUCTION

This dissertation is an attempt to understand how changes in plant community

composition along a precipitation gradient feed back into nutrient cycling and ecosystem

processes. This research area integrates plant physiology, community ecology, and

ecosystem science. The ideas in this dissertation are also based on studies of how species

respond to environmental factors, and the effects that species have on those factors

(Cornelissen 1996; Hobbie 1992; Wardle et al. 1998). Several studies have demonstrated

that many plant species traits are correlated, and that suites of traits appear to have

evolved in response to specific environmental regimes vary along a minimal number of

axes (Chapin 1980; Grime 1977; Tilman 1988). More recently, it has been shown that

leaf physiological traits governing the carbon economy of the leaf are interrelated and

reflect fundamental evolutionary tradeoffs and biochemical constraints (Reich et al.

1992). Together, these studies suggest that if we can link the effects of plant species on

ecosystem processes with ecophysiological or life history characteristics, then we may

augment our understanding of ecosystem functioning, by drawing upon evolutionary and

ecological principles.

In this dissertation, I take two main approaches to understanding how species

traits feed back into ecosystem processes. The first is along a precipitation gradient in

lowland Panama. As precipitation increases in the lowland tropics, there is a gradual

change in canopy species from deciduous species in seasonally dry forest, to evergreen

species in wet, seasonal forest. Leaf turnover represents a major pathway of energy and









matter between the plant and soil components of the ecosystem. Therefore, along this

precipitation gradient, I present data on how patterns of canopy phenology are related to

photosynthesis, and leaf life span. In turn, I relate leaf traits to indices of litter quality and

discuss ways in which variation in species composition along this precipitation gradient

can influence the cycling of nutrients and the size of soil nutrient pools.

The second approach I take to understanding how species traits feed back into

ecosystem processes is at one wet forest site along the precipitation gradient. At this site I

present data on how leaf physiological traits that control carbon assimilation and water

loss are related to hydraulic and biophysical characteristics at the branch and whole plant

scale, and discuss how relationships between leaf and branch physiology reflect

evolutionary tradeoffs. I also compare how plant growth forms vary in the litter quality

and decomposition rates of their leaves, and whether leaf litter decomposition can be

predicted from leaf physiological characteristics. Together, this collection of studies

provides evidence that changes in plant community composition with precipitation in

lowland Panama can influence nutrient cycling, and that leaf physiological characteristics

provide information about the potential effects of species on ecosystem processes.














CHAPTER 2
LEAF PHOTOSYNTHETIC TRAITS OF CANOPY TREES ALONG A
PRECIPITATION GRADIENT IN LOWLAND TROPICAL FOREST

Introduction

Tropical forests often exhibit gradients of vegetation structure and species

composition in relation to precipitation (Gentry 1988; Schimper 1898; Wright 1992).

Forests on the dry side of this gradient commonly have a higher proportion of deciduous

canopy species that lose their leaves during seasonal dry periods when low soil water

availability may limit physiological activity. Deciduous leaves of tropical forest reduce

whole-plant transpiration and respiration during drought; and often have higher rates of

photosynthesis per unit leaf mass (Pmass) than evergreen species (Chabot and Hicks 1982;

Eamus and Prior 2001; Prado and DeMoraes 1997; Reich et al. 1992). In contrast,

evergreen leaves have lower Pmass, but exhibit a potentially longer photosynthetic season;

and appear to reduce demand on soil nutrients required to replace leaves (Chabot and

Hicks 1982; Cunningham et al. 1999; Vazquez and Givnish 1998). These respective costs

and benefits partially explain why deciduous trees dominate seasonally dry tropical

forest; why evergreen species dominate wet, seasonal forest; and suggest that Pmass of

canopy species should decrease with increasing precipitation in the tropics. There are few

data available for photosynthetic traits on broad precipitation gradients in lowland

tropical forest. On a global scale, however, Pmass and specific leaf area are reported to

increase with increasing moisture availability (Niinemets 2001; Reich et al. 1999). The

purpose of our study is to determine how patterns of canopy phenology are related to









photosynthesis and leaf life span along a regional precipitation gradient in lowland

Panama.

Interest in leaf trait variation along climate gradients dates back to the time of

Theophrastus and formed some of the earliest ecological works. Recent studies have

focused on precipitation as a driver of resource availability with direct and indirect

effects on plant processes. For example, plants of relatively dry environments in Australia

exhibited more sclerified vasculature than did species in high rainfall sites, possibly

reflecting adaptation to resist wilting and minimize cell damage when water availability

is low (Cunningham et al. 1999). Relatively high leaf N per unit area in dry habitats in

Australia may represent a mechanism by which plants capitalize on higher light

availability in dry habitats (Cunningham et al. 1999; Mooney et al. 1978). LeafN of

montane forest species in Hawaii was also reported to decrease with increasing

precipitation and was related to decreasing soil N availability with increasing

precipitation (Schuur and Matson 2001), suggesting that evergreenness in wet tropical

forest may be a response to relatively low nutrient availability (Monk 1966). Precipitation

may therefore directly affect vegetation structure and community composition through

constraints imposed by water deficit, or indirectly through effects on availability of light

and/or nutrients (Schuur and Matson 2001).

Most studies of leaf trait variation with climate have been conducted along

gradients with a maximum precipitation below 2500 mm yr-1 (Cunningham et al. 1999;

Mooney et al. 1978; Werger and Morris 1991). Other studies on precipitation gradients

with maximum precipitation above 5000 mm yr-1 have focused on phenotypic changes

within individual plant species (Austin and Vitousek 1998; Schuur and Matson 2001). In









contrast, our study addresses variation in leaf traits caused by large changes in species

community composition over short (10 km) distances (Condit et al. 2002). I measured

leaf physiological and structural traits to understand how patterns of resource allocation

to leaves are related to climate on this regional gradient, which lies at the high end of the

global precipitation range (1800 to 3500 mm yr-1). Specifically I wanted to link the leaf

functional traits of dominant species to shifts in community phenology and leaf longevity

characteristics.

Materials and Methods

Site Characterization and Species

Our study was conducted in lowland tropical forest along a precipitation gradient in

the Panama Canal Watershed. Mean annual precipitation (MAP) across this part of the

Panamanian Isthmus ranges from 1800 mm yr- on the Pacific Coast to 4000 mm yr- on

the Caribbean Coast (Condit 1998). I selected four 1-ha lowland forest study plots

established by the Center for Tropical Forest Science (CTFS) over a range of

precipitation with minimal changes in altitude and therefore temperature (Table 2-1). All

sites have a mean monthly precipitation of >100 mm during the wet season (between

May and December) but dry season length (mean number of 30-day periods with <100

mm precipitation) varies between 129 days at the 1800 mm site and 67 days near the

3500 mm site (ACP 2002) Between December and May, the probability of a site

receiving <100 mm of monthly precipitation varies from 80% at the 1800 mm site to 22%

at the 3500 mm site (Paton and Wright 2003). Variation in rainfall during the study

period (2000-2002) was within 15% of MAP at all sites. Variation in the amount and

distribution of annual precipitation has the potential to influence light availability and

relative humidity. Average daily photosynthetically active radiation (PAR) decreases









with increasing precipitation from 32.3 mol m-2 at the 1800 mm site (Juan Posada,

unpublished data); to 31.9 mol m-2 at Barro Colorado Island in the middle of the isthmus

and to 31.4 mol m-2 at the 3100 mm site (Paton and Wright 2003). Mean relative

humidity appears to be highest on the wet Caribbean coast (94.6% at the 3100 mm site);

the 1800 mm site and Barro Colorado Island maintain values of 87.9 and 84.0%,

respectively (Paton and Wright 2003).

The soils in the Panama Canal forests are well-drained clays high in Ca, Mg and

N; and low in K and P relative to other tropical soils (Dietrich et al. 1982; Kursar et al.

1995; Yavitt et al. 1993). All of the study sites are on volcanic substrate (except Ft.

Sherman, which lies on sedimentary substrate). However, similarities in soil

characteristics between sedimentary and volcanic substrates on Barro Colorado Island

(which lies in the middle of the isthmus) suggest that in this area nutrient availability is

determined more by weathering and nutrient cycling by vegetation than by parent

material (Yavitt 2000).

Soil water potential ( oil) was measured with the filter paper technique (Deka et al.

1995) at six randomly selected locations in each 1-ha study plot, nine times over a 14-

month period (February 2001 to March 2002). One 42.5 mm diameter filter paper

(Whatman no. 42, batch #711492, Whatman, Kent, UK) was equilibrated for 6 days with

a fresh soil sample taken from 15 to 20 cm depth. Then the gravimetric moisture content

of the filter paper was used to predict water potential using the regression equation of

Deka et al. (1995). Assuming that gravitational and solute potential are negligible, the

resulting values represent soil matric potential. I also determined the gravimetric soil

water content on a 5 g subsample by drying at 1050C for 24 h.









At each site, I mesaured the eight canopy tree species with the largest relative

proportion of basal area. Species composition and phenological habit of the most

common canopy trees change rapidly across this gradient, as species richness increases

steeply with mean annual precipitation (Table 2-2) (Pyke et al. 2001). One recent study

comparing beta-diversity (how species composition changes with distance) of tree

communities between lowland forests in Ecuador and Peru found that distant forests

(>1000 km) with similar climate shared a much larger proportion of the most common

species than would be expected by chance (Pitman et al. 2001). In contrast, tree

community composition varies substantially even over 10 to 20 km distances in Panama;

and such variation appears to be regulated largely by climate variation (Condit et al.

2002). All study plots are located in mature forest (>500 yr), except the 1800 mm site,

which is a forest of mixed age (70-100 yr). Measurements on the largest canopy trees at

the 1800 mm site are comparable to other sites even though this forest is younger,

because canopy composition was representative of mature moist and dry forest (Croat

1978).

Photosynthesis

At the 1800 and 3100 mm sites, canopy leaves were collected using canopy cranes

maintained by the Smithsonian Tropical Research Institute. The youngest fully expanded

mature leaves were cut at the petiole; and immediately sealed in a darkened humidified

container. At the 2300 and 3500 mm sites, leaves were collected from the upper canopy

using a shotgun and then treated the same. Canopy trees were defined as individuals with

approximately 80% of the crown exposed to full sun. All leaves were collected before

1030 h the day after rainfall and were transported to the laboratory within 2.5 h where

photosynthesis was measured immediately.









Photosynthetic capacity was measured as the maximum rate of 02 evolution with a

Clark electrode (Model LD2, Hansatech, Norfolk, UK) (Delieu and Walker 1981) on 3 to

8 leaves of each of 2 to 4 individuals for a total of 15 leaves per species from a site. A 10-

cm2 leaf disk was placed in a sealed chamber containing 10% CO2 and maintained at

290C. Leaves were first induced with 250 and 560 [tmol photosynthetic photon flux

density (PPFD) before the maximum rate of 02 evolution at 2000 [tmol was measured.

Light was provided by a quartz-halogen lamp (Bjorkman and Demmig 1987) with

attenuation achieved by inserting neutral density filters and verified with a quantum

sensor (Model LI 190SB, Li-Cor Inc., Lincoln, NE). At each light level, a stable signal

was usually obtained in 5 to 7 min. The chamber was flushed with 10% CO2 for 2 min

between changes in light intensity. The high concentration of CO2 in the 02 electrode

chamber bypasses all stomatal and cuticular resistance, so that the measured maximum

rate reflects the Rubisco-limited rate of photosynthesis; and thus is a good index of

enzyme allocation to photosynthetic capacity. Maximum rates of photosynthetic 02

evolution were correlated with measurements of CO2 assimilation conducted with an

infrared gas analyzer (Model 6400, Li-Cor, Inc.) on a subset of study species (r2=0.82;

P<0.0001; n=12).

Leaf Structure and Chemistry

After photosynthetic measurements, I measured lamina thickness between primary

and secondary veins with a digital caliper (Mitutoyo Inc., Japan). Leaf discs were dried

overnight at 650C; weighed for determination of specific leaf area (SLA); and ball milled.

All leaf discs from the same tree were pooled for chemical analysis. One composite

sample from each tree was analyzed for N using an elemental analyzer (Model NCS









2500, Carlo Erba, Milano, Italy). A separate subsample was analyzed for 615N at the

University of Georgia Institute of Ecology. Leaf carbon fraction analyses were performed

on a single composite sample per species at a site using forest-product techniques (Ryan

et al. 1989). Dried leaf samples were digested in a detergent solution by which soluble

and nutritionally available cell contents were separated from neutral detergent fiber

(NDF), which includes all cell wall constituents and is not immediately nutritionally

available. A dilute acid detergent solution was then used to determine acid detergent fiber

(ADF, lignocellulose) before lignin was separated from cellulose in 72% H2SO4. Leaf

toughness was measured in the field on freshly collected leaves with a penetrometer

(Pesola, Switzerland), which measured the maximum force during punching through the

lamina between primary and secondary veins at a steady, slow rate with a 1 mm diameter

plunger. Such measurements are not equivalent to the material property of fracture

toughness, which is a more theoretically relevant measure of toughness against herbivore

action. However, leaf toughness was correlated with fracture toughness for species at the

1800 mm site (Kaoru Kitajima, unpublished data).

Two-Site Comparisons

Canopy access at the two crane sites allowed for ease in data collection of leaf

water potential and leaf life span of study species. Leaf water potentials ( feaf) were

measured at predawn (0600 h) and midday (1200 h) for the study species at the 1800 and

3100 mm sites using a pressure chamber (PMS Instruments, Corvallis, Oregon, USA). At

both sites, canopy cranes were used to collect three terminal shoots from two individuals

per species from the upper canopy in July 2001, October 2001, and February 2002

representing the early and late wet season and dry season, respectively. Leaf life span was









measured on marked sun leaves that were censused every 6 weeks for 5 years as part of a

larger study on canopy leaf dynamics (S. Joseph Wright, unpublished data).

Data Analysis

Analyses of the statistical effect of climate on leaf traits were conducted with

one-way general linear models, to test whether leaf traits varied significantly with

precipitation. Analyses of the relationships between leaf traits were performed with

one-way analyses of covariance to test for heterogeneity of means around regression

slopes. Each analysis of covariance was conducted independently with Pmass as the

dependent variable; MAP as the grouping factor; and leaf N, SLA or leaf life span as the

covariate. Mean 4eaf between sites was compared by repeated measure analysis of

variance and profile analysis with site as a between-subject factor; and measurement date

and time of day as multiple within-subject factors (von Ende 1993). All analyses were

conducted using SAS version 6.12 (SAS 1985).

Results

Soil Water Potential

Soil water potential reinforced precipitation data, demonstrating that the main

difference between sites is the length and intensity of the dry season and not differences

during the wet season (Figure 2-1). Furthermore, maximum gravimetric soil moisture, an

index of soil water-holding capacity, increased with precipitation from a maximum of

49.4 to 89.1% of dry soil mass from the driest to wettest site.

Photosynthesis

Mean annual precipitation (MAP) explained 13% of the variation in area-based

maximum photosynthetic rates (Parea) of the eight most abundant canopy tree species at

each of the sites (Figure 2-2A). The Parea showed a marginally significant decrease with









increasing MAP. When photosynthesis was expressed on a mass basis (Pmass), MAP

explained 42% of the variation and Pmass decreased significantly with increasing

precipitation (Figure 2-2B). There was no clear relationship between leafN per unit area

and MAP (r2=0.00, P=0.95). However, MAP explained 21% of the variation in leaf N per

unit mass and 34% of the variation in 615N, both of which decreased significantly with

increasing precipitation (Figure 2-3A,B). Both photosynthesis and leafN showed

stronger correlations with precipitation when expressed on a mass basis than on an area

basis because there was an increase in leaf thickness with increasing MAP (Figure 2-4A).

In a related index of leaf structure, SLA decreased with increasing MAP (Figure 2-4B).

Analysis of covariance revealed no heterogeneity of means around regression

slopes caused by MAP in the relationship between Pmass and leaf life span (F=0.23;

P=0.58) or leaf N (F=0.51; P=0.68), and only a marginally significant heterogeneity of

slope caused by MAP in the relationship between Pmass and SLA (F=2.95; P=0.05).

Therefore, species from all sites were regressed together in the same predictive

relationships, regardless of site, although species from different sites tended to occupy

different ranges of the relationship (Figure 2-5). The Pmass was positively correlated with

leaf N across all species and sites; and this predictive relationship improved from an r2 of

0.56 to 0.78 by removing the statistical outlier (Figure 2-5A). The Pmass and SLA were

positively correlated across all species and sites; and were expressed as a log-linear

function (Figure 2-5B).

Leaf Structure and Chemistry

Toughness of canopy leaves significantly increased with MAP and lamina

thickness (Table 2-3). Fiber (NDF) and cellulose concentration were positively related to









lamina thickness; thus the proportion of leaf cell-wall material increases as leaves

increase in thickness. With increasing precipitation, I also noted trends toward reduced

nutritional content and increased antiherbivore defense (fiber and lignin per unit leafN)

(Table 2-3) (Cunningham et al. 1999).

Two-Site Comparisons

Median leaf life spans were longer at the 3100 mm site than they were at the

1800 mm site (t=-6.25; P=0.003; df=4); and were negatively correlated with Pmass (Figure

2-5C). Midday leaf water potential ( faf) was lower at the 1800 mm site than at the 3100

mm site during the early (F1,14=6.11; P<0.05) and late wet season (F1,14=3.17; P<0.1;

Figure 2-6); but not during the dry season. There were no significant differences in

predawn qeaf among sites during any measurement period.

Discussion

As precipitation increases from South to North across the Isthmus of Panama, there

is a gradual change in canopy leaf traits from short-lived leaves with high Pmass in

seasonally dry forest, to relatively long-lived leaves with lower Pmass and increased

allocation to structural defense in wet forest. Relatively high Pmass and short leaf life

spans of canopy leaves in seasonally dry forest may allow canopy species to take

advantage of high light availability when water is available; and to minimize water loss

and respiration costs during rainless periods. Longer leaf life spans exhibited by canopy

trees in wetter forest appear to place constraints on Pmass by necessitating increased

allocation to structural defenses. Although substantial variation in leaf traits exist along

this precipitation gradient, it is important to note that precipitation drives the expression

of leaf traits through soil moisture and tree water status (Reich 1995; Reich and Borchert









1984); and potentially by influencing light and nutrient availability, although these

indirect effects are less documented (Schuur and Matson 2001).

Our data show coupling between Voil and seasonality in precipitation and show

how variation in dry season intensity among study sites appears to be the greatest source

of variation in soil water availability among sites (Figure 2-1). The range of Voil values

observed are consistent with other studies of lowland Neotropical and West African

forest reporting upper soil values near 0 MPa during the wet season to -3.5 MPa or lower

during dry periods (Goldstein et al. 1986; Holbrook et al. 1995; Veenendaal et al. 1996).

Our values of predawn and midday feaf are also comparable to studies conducted in

seasonally dry tropical forest in Venezuela, Australia, and Costa Rica (Borchert 1994;

Eamus and Prior 2001; Goldstein et al. 1986; Medina and Francisco 1994; Sobrado

1986). Differences in midday leaf between the 1800 and 3100 mm sites during the wet

season but not during the dry season (despite strong differences in Voil in the dry season)

indicate that differences in atmospheric water content and stomatal function can affect

Vlaf independently of Voil. Additionally, two of the deciduous species at the 1800 mm

site had no leaves during February 2002 and were therefore not measured; whereas the

remaining species maintained leaf values between -1.0 and -1.5 MPa and thus

minimized variation between the 1800 and 3100 mm sites, resulting in no statistical

difference among the measured species. Sobrado (1986) and Medina and Francisco

(1994) also found a stronger decrease in leaf in deciduous than in evergreen species

during the dry season, suggesting that maintaining functional leaves during the dry

season requires greater resistance to desiccation, possibly through allocation to deeper

roots.









Several studies have addressed the effects of precipitation on nutrient availability

as a factor governing the expression of leaf traits at the community scale (Austin and

Vitousek 1998; Givnish 2002; Schuur and Matson 2001). Evergreen vegetation tends to

dominate nutrient-poor habitats (Monk 1966). Dominance of evergreen species in wet

forest is consistent with the notion of reduced nutrient availability as precipitation

increases. Both Schuur and Matson (2001) and Austin and Vitousek (1998) found the

lowest foliar N concentrations at highest precipitation sites along precipitation gradients

in Hawaiian montane forest. Decreasing leafN per unit mass in lowland Panamanian

forest is consistent with this line of reasoning. Decreasing foliar 615N with increasing

precipitation has been interpreted as signifying increasing N-limited conditions (Austin

and Vitousek 1998; Schuur and Matson 2001; Shearer and Kohl 1986). However, our

bulk soil 615N did not match this pattern (Chapter 3). Thus decreasing foliar 615N with

increasing precipitation in lowland Panamanian canopy trees may reflect the fact that

species with shorter leaf life spans retranslocate leaf N more frequently than evergreen

species and that 615N becomes enriched during re-assimilation of nitrate and leaf N re-

metabolism (Evans 2001).

Decreasing leafN with increasing precipitation may also reflect diminishing

returns for N allocation to canopy leaves if the decrease in light availability with

increasing precipitation is sufficiently strong to drive such a pattern. A difference of

0.9 mol m-2 for average daily PAR exists across this precipitation gradient. It is unknown

whether this amount could contribute to patterns of N allocation to canopy leaves.

Reduced N allocation to leaves with lower light availability is consistent with the

functional convergence hypothesis (which predicts that plants should only allocate









resources to photosynthetic capacity that can be used despite constraints imposed by

limiting resources such as light and water) (Field 1991). One species at the 1800 mm site,

Luehea seemannii, increased rates of CO2 assimilation, branch growth, and fruit

production in response to experimental light enrichment; supporting the notion that

canopy trees of lowland forest have the potential to acclimate to higher light availability

(Graham et al. 2003). Therefore, decreasing light availability with increasing

precipitation may contribute to relatively low photosynthetic rates in wet tropical forest.

Our pattern of decreasing Pmass and SLA with increasing moisture availability on

a regional precipitation gradient is consistent with the notion of a bimodal distribution of

leaf longevity (Chabot and Hicks 1982; Kikuzawa 1991). On a global scale,

deciduousness is highest in mid-precipitation and mid-latitude ecosystems. Our

relationships between Pmass and leaf life span and between leaf N and SLA show slopes

similar to those of global comparisons (Reich et al. 1992). This suggests that fundamental

tradeoffs between leaf traits are constant; but at the high end of global precipitation

range, where water availability may exceed biological demand for much of the annual

cycle, Pmass and SLA decrease with increasing precipitation, in contrast to the pattern of

increasing Pmass and SLA with increasing precipitation common in other biomes (Reich et

al. 1999).






16


Table 2-1. Site characteristics for 1 ha census plots in the Panama Canal Watershed.
Basal area, species richness, and tree density represent species with stems >10
cm in diameter.


Site
Parque Metropolitano
Pipeline Road
Fort Sherman
Santa Rita


Mean annual
precipitation
(mm yr-1)
1800
2300
3100
3500


Elevation
(m)

60
210
140
282


Basal
area
(m2 ha-)
25.39
26.59
32.50
25.74


Species
richness
(no. ha-)
36
95
87
162


Tree
density
(no. ha-)
318
560
569
497









Table 2-2. Study species from four sites along precipitation gradient in Panamanian
lowland tropical forest, including phenological classification into deciduous
(losing leaves for more than a few weeks), brevi-deciduous (losing leaves
once per year and immediately flushing a new set) and evergreen.
Classification based on field observations and the Flora ofBarro Colorado
Island (Croat 1978).


Species
1800 mm
Anacardium excelsum
Astronium graveolens
Calycophyllum candidissimum
Chrysophyllum cainito
Enterolobium cyclocarpum
Luehea seemannii
Pseudobombax septenatum
Spondias mombin
2300 mm
Poulsenia armata
Pourouma bicolor
Sterculia apetala
Tabebuia guayacan
Tapirira guianensis
Terminalia amazonica
Trattinickia aspera
Virola sebifera
3100 mm
Aspidosperma cruenta
Brosimum utile
Calophyllum longifolium
Dussia mundia
Manilkara bidentata
Marila laxiflora
Poulsenia armata
Tapirira guianensis
3500 mm
Aspidosperma cruenta
Carapa guianensis
Cassipourea eliptica
Erisma blancoa
Sacaglottis trygynum
Sterculia costaricana
Virola koschnyi
Zvzia ramiflora


Family


Phenology


Anacardiaceae
Anacardiaceae
Rubiaceae
Sapotaceae
Fabaceae
Tiliaceae
Bombacaeae
Anacardiaceae

Moraceae
Moraceae
Sterculiaceae
Bignoniaceae
Anacardiaceae
Combretaceae
Burseraceae
Myristicaceae

Apocynaceae
Moraceae
Clusiaceae
Fabaceae
Sapotaceae
Clusiaceae
Moraceae
Anacardiaceae

Apocynaceae
Meliaceae
Rhizophoraceae
Vochysiaceae
Hernandiaceae
Sterculiaceae
Myrsticaceae
Fabaceae


Brevi-deciduous
Deciduous
Evergreen
Evergreen
Deciduous
Brevi-deciduous
Deciduous
Deciduous

Brevi-deciduous
Evergreen
Deciduous
Deciduous
Evergreen
Brevi-deciduous
Deciduous
Evergreen

Evergreen
Evergreen
Evergreen
Evergreen
Evergreen
Evergreen
Brevi-deciduous
Evergreen

Evergreen
Evergreen
Evergreen
Evergreen
Evergreen
Brevi-deciduous
Evergreen
Evergreen


I Y









Table 2-3. Correlation coefficients between leaf defensive characteristics and mean
annual precipitation and lamina thickness for 32 species of lowland tropical
forest canopy trees at four levels of mean annual precipitation.
Mean annual precipitation Lamina thickness
(mm yr-1) (mm)
r P r P
Toughness (kg) 0.535 0.0016 0.843 <0.0001
Fiber-NDF (% dry mass) 0.127 0.490 0.356 0.0454
Fiber-ADF (% dry mass) 0.124 0.500 0.276 0.1270
Lignin (% dry mass) 0.077 0.674 0.091 0.6223
Cellulose (% dry mass) 0.168 0.358 0.434 0.0130
NDF:N 0.470 0.0067 0.674 <0.0001
ADF:N 0.438 0.0121 0.600 0.0003
Lignin:N 0.376 0.0339 0.442 0.0113
Note: Bold type indicates significant correlation.



























J F M A M J J A S O N D J F


100

90


Ieeo


J F M A M J J A SONDJ F

2001 Date 2002


Figure 2-1. Seasonal variation in (A) soil water potential ( oil) and (B) gravimetric water
content determined between 15-20 cm depth in four 1-ha lowland tropical
forest study plots in Panama. Points represent mean (+1SE) (n=6).






























2000 2500


2000 2500 3000


A


3000 3500 4000


3500 4'


B







000


Mean annual precipitation (mm)


Figure 2-2. Relationship between mean annual precipitation and maximum
photosynthetic oxygen evolution rate (A) per unit area (Parea) and (B) per unit
mass (Pmass) for the eight most common canopy species at four sites along a
precipitation gradient in lowland Panamanian forest.


0 r2 = 0.13
p < 0.05
0

8 0 0
o 0
0o
0 0

o o
o o 0

0


CM
E
O(N

75
E

a)


150
1500


350

300

250

200

150

100

50


O'N



U)
U)
o
E

0C
r
a


r2 = 0.42
0 p < 0.0001






o
00
0 8




0

80


500
1500













40

35

S30

E 25

Z 20
4-
( 15
-j


10


5
1500 2000


-2
1500


2500 3000 3500


2000 2500


3000 3500


Mean annual precipitation (mm)


Figure 2-3. Relationship between mean annual precipitation and (A) leaf N concentration
per unit mass and (B) N isotopic composition (615N) for the most common
canopy species (excluding legumes) at four sites along a precipitation gradient
in Panama.


O 0
O
-O
- 0



- 0
O
O 0

r2 = 0.21
p < 0.01 0


r2 = 0.34
0 p < 0.001

0
O




08
0 0
0
o
o o


4000


4000













0.6

E 0.5
E
S0.4
c)
0 0.3
M-0
c 0.2
C

c 0.1
-j

0.0
150C


240


200

160

120

80

40


2000


2500 3000 3500 4000


1500 2000 2500 3000 3500


4000


Mean annual precipitation (mm)


Figure 2-4. Relationship between mean annual precipitation and (A) lamina thickness and
(B) specific leaf area (SLA) for the eight most common canopy species at four
sites along a precipitation gradient in Panama.


r2 = 0.30
p = 0.001 0


o 8



8o
oo
o
8


CN
E

.J
C J


r2 = 0.23
o p = 0.005








O O
0 0
0
0
O

















































350
300
250
200
150
100

50
0
1


r2 =0.78
p < 0.0001 0

v V o
V 0v
V
0 00
01*Z





5 10 15 20 25 30 35 4C

Leaf N (mg g-1)

r2 = 0.66
p < 0.0001


v v
0 0 0











Log (SLA)


r2= 0.52
0 p =0.005
0
0


00 *
S*


S


200 300 400 500

Median leaf life span (days)


Figure 2-5. Maximum photosynthetic 02 evolution rate (Pmass) as a function of (A) leafN
from study species at the 1800 mm (open circles), 2300 mm (open triangles),
3100 mm (closed circles), and 3500 mm (closed triangles) sites. The outlier
from the 3500 mm site (Zygia ramiflora) was not included in the regression
due to large studentized residuals; Pmass=95.56.(leaf N)-17.81. (B) Pmass as a
function of log-transformed specific leaf area (SLA); Pmass=301 log(SLA)-
429.7. (C) Pmass as a function of median leaf life span for study species at the
1800 mm (open circles) and 3100 mm (closed circles) canopy crane sites

Pmass=0.53.(leaf life span)-314.93.


00














0.0




- -0.5




S-1 .
S-1.0


- se.0


Dry season


3100 mm predawn 1
-*- 1800 mm predawn
-0- 3100 mm midday
-0- 1800 mm midday
-1.5 -1
J J A S O N D J F M

2001 Month 2002


Figure 2-6. Seasonal variation in leaf water potential ( eaf) for canopy trees at two sites
with contrasting precipitation in lowland Panama. Each point represents the
mean (+1SE) of two individuals from 6-8 species at each site.














CHAPTER 3
NUTRIENT CYCLING ON A PRECIPITATION GRADIENT IN LOWLAND
TROPICAL FOREST

Introduction

Water availability has the potential to control components of nutrient cycles, such

as nutrient pool sizes and fluxes between pools in terrestrial ecosystems. Water may

affect nutrient cycling directly through soil processes such as leaching, weathering and

decomposition of organic matter. Water may also influence nutrient cycling indirectly,

through effects on plant community composition, since many of the plant characteristics

that influence nutrient cycles, such as litter quality and productivity, vary with

precipitation. Lowland tropical forest is important in global nutrient cycles and therefore,

understanding how nutrient pool sizes and cycling respond to water availability is crucial.

Studies of nutrient cycling in humid montane tropical forest (>2000 mm precipitation

yr1) suggest that C pool sizes increase whereas N availability decreases with increasing

precipitation (Austin and Vitousek 1998; Schuur et al. 2001; Schuur and Matson 2001).

These montane forest studies were conducted in Hawaii, where precipitation may vary

while other state factors such as parent material, species composition and temperature

remain constant (Vitousek 1995). The purpose of this study is to use a precipitation

gradient to sample patterns of nutrient cycling in lowland tropical forests that vary in soil

parent material and plant community composition.

As precipitation inputs exceed biological demand, such as in the wet tropics,

several key soil processes that influence soil nutrient cycling and storage are likely to









occur. Removal of highly mobile nutrients in the soil solution may occur via leaching, a

process that may contribute to reduced nutrient availability under conditions of high

water inputs (Radulovich and Sollins 1991; Schuur and Matson 2001). Where poorly

drained soils are prevalent, oxygen availability may limit microbial activity that

mineralizes N. Several lines of evidence also suggest that water availability may affect

soil nutrient pools by weathering soil to produce secondary minerals with a higher

surface area and thus higher capacity to adsorb organic matter (Torn et al. 1997).

Therefore, water availability has the potential to shape ecosystem nutrient dynamics

through effects on the size of soil nutrient pools and the rate at which nutrients become

available for plant uptake.

Differences in water availability also result in distinct species assemblages that

have the potential to influence nutrient cycling through effects on litter productivity and

quality. Species in low resource environments tend to minimize tissue turnover and

produce long-lived leaves with high concentrations of carbon-based defenses and slow

rates of decomposition (Chapin 1980). In contrast, species from high resource habitats

produce relatively short-lived leaves with higher photosynthetic rates and higher N

concentrations, resulting in faster decomposition rates (Chapin 1980). The influence of

litter quality in determining nutrient availability is thought to increase with actual

evapotranspiration (AET) (Meentemeyer 1978). Tropical forest has high AET, so litter

quality is expected to have a strong effect on nutrient availability. I present data on

nutrient cycling from a precipitation gradient in lowland tropical forest in Panama.

Substantial changes in species composition of canopy trees exist along this precipitation

gradient (Chapter 2). The objective was to use the precipitation gradient to sample









diversity of leaf phenological patterns across plant communities, which are also affected

by soil properties.

Materials and Methods

Study Site

This study was conducted in lowland tropical forest along a rainfall gradient in

Central Panama (Chapter 2). Mean annual precipitation (MAP) across these sites varied

from 1800 mm yr1 at the driest site to approximately 3500 mm yr1 at the wettest site

(ACP 2002). Four 1-ha lowland forest study plots established by the Center for Tropical

Forest Science (CTFS) over a range of precipitation with minimal changes in altitude and

temperature were used as study sites (Table 3-1). All sites have a mean monthly

precipitation >100 mm during the wet season between May and December, but dry

season length (mean number of 30-day periods with <100 mm precipitation) varies

between 129 days at the 1800 mm site and 67 days near the 3500 mm site (Condit 1998).

Variation in rainfall during the study period (2000-2002) was within 15% of MAP at all

sites.

The soils in the Panama Canal forests are generally well-drained and rich in clay,

Ca, Mg and N, but poor in K and P relative to other tropical soils (Dietrich et al. 1982;

Kursar et al. 1995; Yavitt et al. 1993). All of the study sites lie on volcanic substrate

except the 3100 mm site, which lies on sedimentary substrate (Table 3-1). The 1800 mm

site is derived from the early to late Oligocene, principally agglomerate, generally

andesitic in fine-grained tuff and includes stream-deposited conglomerate (Woodring et

al. 1980). Soils of the 2300 and 3500 mm sites are described as derived from altered

basaltic and andesitic lavas and tuff, including dioritic and dacitic intrusive rocks. The









3100 mm site is derived from the late Miocene or early Pliocene with massive, generally

fine-grained sandstone (Woodring et al. 1980).

Species composition and phenological habit of the most common canopy trees

change rapidly across this gradient as species richness increases steeply with mean annual

precipitation (Pyke et al. 2001). The dominant canopy tree species in drier forest tend to

exhibit shorter leaf life spans and a dry season deciduous leaf phenology (S.J. Wright,

unpublished data).

Foliar Chemistry

As part of a larger study on photosynthetic leaf traits of canopy tree species along

this precipitation gradient (Chapter 2), the eight canopy tree species at each site with the

largest relative proportion of basal area were studied. Young, fully-expanded mature sun

leaves were collected from 2-4 individuals for a total of 15 leaves per species from a site.

Leaves were collected using canopy cranes maintained by the Smithsonian Tropical

Research Institute at the 1800 and 3100 mm sites. At the 2300 and 3500 mm sites, leaves

were collected from the upper canopy using a shotgun. Leaf material was dried for 48 h.

at 650C. Leaf samples from the same tree were pooled for chemical analysis. One

composite sample from each tree was analyzed for C and N using an elemental analyzer

(Model NCS 2500, Carlo Erba, Milano, Italy).

Litterfall Collection and Processing

Litter was collected in 0.25 m2 traps randomly located at 20X20-m grid points in

each 1 ha plot. Large items, such as palm fronds often fell across traps and only material

that fell in the area above the trap was collected. Litter was collected 11 times at intervals

ranging from 1 to 12 weeks for the period between February 2001 and February 2002.

Litter used in chemical analyses was collected at intervals of 7-10 days in February-June









2001, October 2001, and February 2002. Litter collected at intervals >2 weeks were

adjusted for mass loss within traps using decomposition data from the 3100 mm site

(Chapter 5), and were not used in chemical analyses. Litter was sorted into four classes:

(1) fine woody debris <1 cm in diameter; (2) leaves; (3) reproductive structures including

fruits, flowers and seeds; and (4) other components of litterfall including insects, frass,

canopy soil, and items too decomposed to identify. Litter was dried for 48 hours at 650C

and the separate classes were weighed. Entire samples from each site and pickup date

were ground in a Wiley mill (mesh size 40) and homogenized. A 120-ml subsample was

then retained for further chemical analyses.

Leaf Litter Chemistry

Leaf carbon fraction analyses were performed using a series of increasingly

aggressive extractants (Ryan et al. 1989). Dried, ground litter samples were digested in a

detergent solution to separate labile cell contents from neutral detergent fiber (NDF),

which includes all cell wall components. A dilute acid detergent solution was then used

to determine acid detergent fiber (ADF, lignocellulose) before cellulose was separated

from lignin and insoluble ash in 72% H2SO4. Litter C and N concentrations were

determined with an elemental analyzer (Model ECS 4010, Costech, Valencia, CA).

Isotopic ratios of C (613C) and N (615N) were determined with a continuous flow isotope

ratio mass spectrometer (Model Delta plus XL, Thermo Finnigan, Germany).

Nutrient Availability

Extractable nitrogen and phosphorus

Soil cores were taken at each of the four sites in July 2002, approximately 2 months

after the beginning of the 8-month wet season when nutrient availability is expected to be









highest. Six 10 cm deep soil samples were taken with a slide hammer corer at random

locations in each 1 ha plot. Samples were returned to the lab and hand sorted to remove

roots and rocks, and three subsamples were taken. Weakly sorbed P from the first 10 g

subsample was extracted in 50 ml 0.5 M NaHCO3 (Crews et al. 1995; Miller et al. 2001).

Extracts were shaken for one minute and after 24 h a 10-ml sample of the supernatant

was removed and frozen until transported to the University of Florida for analysis. I

extracted NH4 and NO3 from the second 10-g subsample for 24 h in 50 ml 2 M KC1.

Nitrogen mineralization and nitrification

The third 10-g subsample of each core was weighed into a 100-ml sample cup,

covered, and allowed to incubate aerobically for 10 days in a dark cabinet at 240C. After

10 days, incubated samples were extracted in 2M KC1 as described above and the

difference in NO3 and NH4 was recorded as net mineralization (Riley and Vitousek

1995).

Ion exchange resins

Soil nutrient availability was determined with ion exchange resins (Binkley and

Matson 1983). Three grams of anion exchange resin (Biorad, AG 1-X8, 20-50 mesh, C1

form) and the same amount of cation exchange resin (Biorad, AG 50W-X8 20-50 mesh,

H form) were weighed into separate 5X6 cm undyed monopolyester bags (approx. 190

|tm mesh size). At each of the four sites, resin bags were placed vertically 4 to 7 cm deep.

One anion and one cation resin bag were secured to a plastic stake with monofilament

line at 6 random locations in each 1 ha plot. Resin bags were collected from the field after

21 days and rinsed with DI water to remove soil particles. Ions were extracted with 20 ml

of 0.5 M HC1 and then neutralized with 20 ml of 0.5 M NaOH. PO4 and NO3 from anion









extracts and NH4, K, Ca and Mg from cation extracts were measured colorimetrically

using an autoanalyzer at the University of Florida, Food and Agricultural Sciences Soils

Testing Lab to determine nutrient content per bag.

Soil nutrient pools

The remaining sample of each soil core was dried at 500C for 48 h, and sifted to

pass through a 2-mm sieve. Total soil C and N as well as 613C and 615N were analyzed on

an elemental analyzer (Model ECS 4010, Costech) connected to a continuous flow

isotope ratio mass spectrometer (Model Delta plus XL, Thermo Finnigan, Germany).

Bulk density of surface 10 cm was estimated using 5 cm diameter volumetric cores dried

at 1050C. Soil moisture was determined gravimetrically on bulk density samples and

expressed as grams of water per unit mass of dry soil.

Results

Foliar Chemistry

Foliar N of the eight most common canopy tree species was highest at the 1800 mm

site and lowest at the 3100 mm site, showing a general decrease in foliar N as

precipitation increases (Table 3-2). In contrast, foliar C values were similar at all sites.

Foliar C:N of the eight most common canopy tree species decreased significantly with

increasing precipitation (r2=0.32; P<0.001), and was largely driven by changes in N.

Litter Production and Chemistry

Total litter production varied from 12.47 Mg ha-1 yr- at the 1800 mm site to 9.80

Mg ha-1 yr1 at the 3500 mm site. Fine woody litterfall represented 11-17% of total litter

and was about 20% higher at the two wetter sites (Table 3-3). Leaf litter represented 60-

75% of total litter and was 45% higher at the 1800 mm site than the other three sites

(Table 3-3). Reproductive structures comprised 7-15% of total litter fall and showed no









clear pattern with precipitation (Table 3-3). Other litter components were 5-9% of total

litter production (Table 3-3). Litterfall rates were seasonal with more litter falling during

the dry season, and seasonal differences between maximum and minimum litterfall rates

decreased with increasing precipitation (Figure 3-1).

There was significant variation in mean litter N concentration of leaf litter with

precipitation (F=5.07; P<0.05) and values tended to increase with increasing

precipitation (Table 3-3). There was also significant variation in litter 615N among sites

with different soil parent material (Table 3-3; F=28.05; P<0.0001), but values did not

vary in any predictable manner with precipitation. Litter 613C decreased linearly with

increasing precipitation indicating greater integrated water use efficiency of leaves in

drier forest (Figure 3-2). Litter C increased (F=33.01; P<0.0001) with increasing

precipitation and may be related to variation in litter lignin concentration among sites

(Table 3-3; F= 1.76; P<0.001). Litter lignin:N significantly increased with increasing

precipitation (F=3.74; P<0.05) whereas cellulose concentration showed no significant

variation with precipitation (Table 3-3).

Nutrient Availability

Extractable nitrogen and phosphorus

Extractable P was significantly higher at the driest site than at the three wetter sites

(Table 3-4; F=3.50; P<0.05). Extractable NO3 was significantly higher at the wettest site

than at the three drier sites (Table 3-4; F=4.80; P<0.05). Extractable NH4, net N

mineralization and nitrification showed no clear pattern with precipitation (Table 3-4),

but N mineralization decreased linearly with increasing litter lignin:N (Figure 3-3)

indicating an effect of litter quality on N availability.









Ion exchange resins

Resin exchangeable NO3 was about 420% higher at the two sites on pre-tertiary

basalt (2300 and 3500 mm sites) compared to the other two sites (Table 3-4; F=5.81,

P<0.01). There were no clear patterns in exchangeable NH4 or K, but exchangeable Ca

decreased 76% from the driest to wettest sites (Table 3-4; F=11.75; P<0.0005) and

measurable quantities of exchangeable Mg were only detected at the driest site (Table

3-4).

Soil nutrient pools

Total soil N and C pools of the top 10 cm were higher at the two wetter sites than at

the two drier sites (Table 3-5). Soil N showed an exponential increase in relation to litter

lignin concentration (Figure 3-4A) and a strong linear increase in relation to litter

lignin:N (Figure 3-4B) suggesting that lignin-bound proteins in the soil organic matter

matrix comprise a large part of the total soil N pool. Soil 613C and 615N were more

enriched than litter inputs, possibly due to faster turnover of lighter isotopes (Figure 3-2;

Table 3-5) (Nadelhoffer and Fry 1988). Soil bulk density decreased with increasing

precipitation (Table 3-6).

Discussion

Foliar Chemistry

Our results suggest that patterns of foliar C per unit N reflect changes in species

composition across this precipitation gradient. Decreasing foliar N of canopy species may

reflect increased allocation of carbon-based leaf defenses to canopy leaves in association

with longer leaf life spans in wetter forest. Decreasing foliar N with increasing

precipitation does not appear to be caused by reduced N availability with increasing

precipitation as has been observed on rainfall gradients in Hawaii given that soil N









availability did not decrease with increasing precipitation. This pattern, whether driven

by photosynthetic or anti-herbivore allocation patterns, may provide an important

feedback to nutrient availability by affecting litter quality since litter C:N is negatively

correlated with decomposition rate (Chapter 5).

Litter Production and Chemistry

Patterns of litterfall suggest that the driest site may be slightly more productive than

wetter forests. The primary differences are greater leaf litter production and stronger

seasonality from the 1800 mm site, which has the highest proportion of dry season

deciduous tree species. Therefore increased leaf litter production at the driest site may be

the result of increased leaf turnover. However, in order to understand the extent to which

productivity is regulated by water availability, other components of productivity such as

root growth, trunk growth and respiration need to be incorporated. Nonetheless, litterfall

is often the greatest fraction of productivity (Clark et al. 2001a). Several recent studies

have also revealed that productivity in humid tropical forest may actually decline at high

annual precipitation (>2500 mm) (Clark et al. 2001b; Schuur and Matson 2001).

Therefore, assuming that all sites are at steady state, the result reported here is consistent

with reduced ecosystem productivity at extremely high precipitation in the tropics due to

light or nutrient limitation (Schuur and Matson 2001).

The reduction in leaf litter quality with increasing precipitation suggests that

nutrient mineralization slows with increasing precipitation. Because fine root litter

quality may respond similarly to leaves (Ostertag 2001), belowground litter may

contribute to reduced nutrient mineralization in sites with lower litter quality (Figure 3-

3). Decomposition rates at the 3100 mm site decrease substantially with increases in litter

lignin:N (r2=0.38; P<0.05; Chapter 5). Changes in litter lignin and lignin:N suggest that









compound specific changes in litter quality may be important across this gradient. The

increase in lignin suggests that a higher proportion of litter will enter directly into the

slow decomposing pool or organic matter (Vitousek et al. 1994). This is a potential

explanation for the increase in soil C and N storage with increasing precipitation and

decreasing litter quality.

Nutrient Availability

The availability of P, Ca and Mg varied in relation to precipitation, whereas N

appeared more responsive to parent material. Decreasing available P, Ca and Mg with

increasing precipitation is consistent with the observation that weathering can leach these

elements from the soil profile and reduce availability to plants. Relatively high P

availability at the 1800 mm site may contribute to higher litterfall productivity, since P is

often considered to be the most limiting nutrient in lowland tropical forest, and both P

availability and litterfall productivity are highest at the 1800 mm site and relatively low at

all other sites. However, litterfall is extremely variable and measurements over several

years are needed to determine if this trend is robust.

In contrast to P, N appeared to be under stronger regulation of parent material

with both sites on tertiary basalt exhibiting high exchangeable NO3. Both Austin and

Vitousek (1998) and Schuur and Matson (2001) found decreasing N availability with

increasing precipitation in Hawaiian montane forests in sites with consistent parent

material. Our results suggest that parent material may alter the relationship between

precipitation and N availability in tropical forest.

The results of increasing total soil N and C with increasing precipitation support

the notion that soil organic matter increases with increasing precipitation in humid forest

(Schuur et al. 2001). Soil N pools in this part of Panama are high relative to both









temperate and montane tropical forest (Austin and Vitousek 1998; Schuur and Matson

2001; Vitousek and Sanford 1986). The increases in soil N and C pools with increasing

litter lignin and litter lignin:N suggest that lower quality litter may function to increase

soil organic matter accumulation. Since soil waterlogging appears to be of minimal

importance in these study sites, the observed increase in soil organic matter with

increased precipitation may reflect an indirect effect of precipitation on species

composition with more evergreen species producing lower quality litter as precipitation

increases.

Total soil N increases as a function of litter lignin:N, whereas soil N

mineralization rates decrease with increasing litter lignin:N (Figure 3-3) (Scott and

Binkley 1997). Therefore, as lignin:N increases, it appears that a higher proportion of soil

N is tightly held in the organic matter matrix and the mineralization rate of that N is

slower. Lower litter quality may function as an N conservation mechanism to prevent N

losses through leaching by decreasing the rate of organic N reactivity.

Conclusions

Overall, the results suggest that variation in plant community composition along

this precipitation gradient can have substantial effects on soil nutrient pools and on how

nutrients are cycled by vegetation. Several general patterns, such as decreasing litter

quality and decreasing soil availability of PO4, Ca and Mg with increasing precipitation

appear to corroborate patterns described for island ecosystems. Despite the diversity of

soil substrates and plant community compositional changes across even short distances in

Panama, there appear to be some patterns of C and N cycling and accumulation that are

consistent with previous findings in Hawaiian ecosystems, where greater control over soil

and species has generated theoretical predictions. Clearly, reduced litter quality and









increased soil nutrient pools with increasing precipitation, which I have shown to be

related, are two components of nutrient cycling that vary similarly in Panamanian

lowland forest and Hawaiian montane forest. However, other patterns of N cycling are

more strongly related to soil parent material than precipitation, making some predictions

developed in relatively homogeneous island systems more difficult to apply in more

heterogeneous landscapes. Further studies in lowland continental tropical forest are likely

to contribute to our understanding of nutrient cycling in humid ecosystems and dispel

myths or corroborate patterns observed in model island systems.











Table 3-1. Characteristics of sites along precipitation gradient across the Isthmus of
Panama.
Site CTFS Mean annual Parent material Order Suborder
plot precipitation
code' (mm)


Parque Metropolitano PM 1800 Panama formation Ultisol Ustult
Pipeline Road 8 2300 Pre-Tertiary basalt Ultisol Humult
Fort Sherman S3 3100 Chagres Sandstone Histosol Saprist
Santa Rita 31 3500 Pre-Tertiary basalt Ultisol Humult
"Pike et al. (2001)
bWoodring et al. (1980)

Table 3-2. Percentage of N and C in canopy sun leaves from the eight most common
canopy tree species across a precipitation gradient in lowland forest in
Panama (Chapter 2). Values are means 1 SE. (n=8)
Site Foliar N (%) Foliar C (%)
1800 mm 2.26a + 0.24 49.02a + 0.83
2300 mm 2.01ab 0.20 47.72a 1.57
3100 mm 1.43b 0.12 49.24a+ 1.10
3500 mm 1.64ab 0.28 49.42a 1.08









Table 3-3. Litterfall rates separated by component for the year between February 2001-
February 2002 and litter quality of leaf litterfall from 4 sites along a
precipitation gradient in lowland Panamanian forest. Values for leaf litter
quality are means 1 SE (n=4).


Litterfall (Mg ha1 yr-)
Fine woody debris
Leaf
Reproductive structures
Other
Total


Site
1800 mm

1.37
9.47
0.94
0.69
12.47


2300 mm

1.37
6.33
1.40
0.93
10.03


3100mm

1.65
6.45
1.79
0.62
10.51


3500 mm

1.58
6.74
0.64
0.83
9.79


Litter chemistry
N (%)
615N (%o)
C (%)
Lignin (%)
Lignin:N
Cellulose (%)


0.96a + 0.04
-0.46a 0.24
39.7a + 0.6
15.4a 1.9
16.0a 1.5
20.0a + 0.7


1.34b + 0.06
1.24b + 0.09
44.3b + 0.5
20.9b + 1.6
16.0a + 0.9
18.4a + 2.5


1.13ab + 0.06
0.25c + 0.10
47.1c + 0.3
23.8b + 1.6
21.1b 2.0
22.6a 1.8


1.28b + 0.09
1.23b + 0.06
45.5bc 0.5
23.7b 1.2
18.5ab 1.7
21.6a 1.2









Table 3-4. Soil chemistry to a depth of 10 cm. Values with different letters are
significantly different at a P-value of 0.1.


Site
1800 mm


2300 mm


3100 mm


3500 mm


Extractable nutrients
(mg kg-1)
P
NO3
NH4
Net N nitrification
(mg kg- d-1)
Net N mineralization
(mg kg- d-1)

Exchangeable nutrients
(mg kg resin1 21 d-1)
NO3
NH4
P04
K
Ca
Mg


5.26a 0.78
0.07a 0.08
2.73a+ 1.19
0.71a 0.20

0.74a 0.18





2.34a 1.46
0.47a + 0.08
8.01a 2.01
671a 67
41.11a+ 8.45
3.56a 1.74


3.21b + 0.53
0.73a + 0.61
2.95a + 0.28
1.13a+ 0.21

1.17a 0.17





14.75b + 1.80
0.55a 0.10
0.30b 0.10
491b + 13
21.96b + 5.57
< 0.01b


2.85b 0.77
0.23a 0.25
3.91a 0.40
0.57a 0.20

0.35b 0.15





3.82a + 2.10
0.5Oa + 0.02
0.33b + 0.04
470b + 14
3.38c + 0.93
< 0.01b


3.33b + 0.36
3.73b + 1.59
4.73a + 0.85
0.94a 0.43

0.74a + 0.38





17.4b + 5.20
0.54a + 0.01
0.10b 0.03
570ab 33
9.93b + 0.41
< 0.01b









Table 3-5. Bulk soil chemistry from the surface 10 cm of soil. Values with a different
letters are significantly different (P<0.05; n=6)
Site Total C Total N Bulk 615N
(g/kg) (g/kg) (%o)
X 1 SE X+1 SE X+1 SE
1800 mm 25.21a + 2.79 2.86a 0.81 2.99a 0.61
2300 mm 28.60b + 3.12 2.73a 0.17 6.57b + 0.58
3100 mm 70.66c 5.40 5.25b + 0.45 4.06c + 0.64
3500 mm 39.57b 4.76 3.90Ab + 0.36 5.67d + 0.72


Table 3-6. Soil bulk density and water content in the surface 10 cm of soil. Values are
means ( 1 SE) of averages measured using 6 samples per site during the early
wet season of 2002.
Site Soil bulk density (g cm-3) Water content (g g-1)
1800 mm 0.92 0.02 0.53 0.02
2300 mm 0.73 0.03 0.58 + .001
3100 mm 0.51 0.02 0.82 0.02
3500 mm 0.60 + 0.02 0.98 .006




























2000 2500


3000 3500 4000


Mean Annual Precipitation


Figure 3-1. The difference between seasonal maximum and minimum litterfall rates as a
function of mean annual precipitation for four forest sites along a precipitation
gradient in lowland Panama. Values are means (n=l 1).


45

40

35

30

25

20

15


10
1500


r2 = 0.98




r 2=0.98






43





-26



-27 4



0 -28

O
o -29

r2 = 0.93

-30
Soil
o Litter
-3 1 1 1 1 '
1500 2000 2500 3000 3500 4000

Mean annual precipitation (mm)


Figure 3-2. Relationship between carbon isotope composition of soil (n=6) and litter
(n=4) for four 1 ha forest plots along precipitation gradient in lowland
Panama. Values are means + 1 SE.












1.50 .
r2 = 0.77
0 1.25

1.00

-N 0.75

v "" 0.50

z 0.25

0.00
14 16 18 20 22 24
Litter lignin:N


Figure 3-3. Soil N mineralization rates for the top 10 cm (n=6) as a function of litter
lignin to N ratio (n=4) for four forest sites in lowland Panama that vary in
mean annual precipitation: 1800 mm (open circle); 2300 mm (open triangle);
3100 mm (closed circle); 3500 mm (closed triange). Values are means + 1 SE.












6


5

0)
- 4

Z
S3
O
U)
2


1


12 14 16 18 20 22 24


A







26


Litter Lignin (%)


6


5


e 4

z
= 3
0
0)
2


1


14 16 18 20


Litter Lignin:N


Figure 3-4. Soil N concentration as a function of (A) litter lignin and (B) litter lignin to N
ratio for four forest study sites in lowland Panama. Values are means + 1 SE.
Symbols as in Figure 3-3.


r2 = 0.81


I I I I I I














CHAPTER 4
HYDRAULIC CONDUCTIVITY AND WOOD DENSITY SCALE WITH LEAF
PHOTOSYNTHETIC TRAITS IN PANAMANIAN FOREST CANOPY TREES

Introduction

Species often exhibit substantial variation in rates of carbon gain and resource use.

In species-rich tropical forests, the extent to which observed patterns of resource

utilization and carbon gain are species-specific is uncertain. Results of recent studies

point to substantial convergence in plant functioning among species from diverse biomes

(Reich et al. 1997). For example, maximum photosynthetic rates, stomatal conductance,

and leaf surface area per unit mass (specific leaf area; SLA) are positively correlated and

tend to decrease with increasing leaf life span across a wide array of study sites and

angiosperm taxa (Ackerly and Reich 1999; Reich et al. 1999; Reich et al. 1997).

However, within sites, variation in photosynthetic rate and leaf life span is as great or

greater than variation in mean differences among biomes (Reich et al. 1999).

Understanding hydraulic properties of the branch or whole plant scale may explain

additional within-site variation in leaf characteristics among individuals and species

(Meinzer and Goldstein 1996).

Convergence in regulation of carbon economy at the leaf level may also be related

to the life history features of a species. Photosynthesis, SLA, and leaf nitrogen

concentration are generally related to rapid growth, high allocation to photosynthetic

tissue, early attainment of reproductive age, and regeneration in high resource habitats

(Cornelissen et al. 1997; Poorter and Remkes 1990; Reich et al. 1992; Wright and









Westoby 1999). In addition, leaf photosynthetic traits are correlated with many of the

same whole-organism traits that can be predicted by plant hydraulic conductance.

Interspecific variation in leaf physiology may be related to xylem hydraulic properties,

because stomatal conductance, a leaf area-based property, is often closely coordinated

with the apparent hydraulic conductance of the soil-to-leaf pathway (Andrade et al. 1998;

Kuippers 1984; Meinzer and Grantz 1990; Sperry and Pockman 1993). Recent studies

have suggested that allometric scaling of plant vascular systems is universal and thus

reflects convergence among many species to overcome similar physical limitations of

long-distance water transport (Enquist et al. 1998; Meinzer 2003; West et al. 1999).

Furthermore, similar relationships in the scaling of plant transpiration and animal

metabolism suggest that both share common scaling laws that reflect how resource

requirements of organisms affect distribution in ecological communities (Enquist et al.

1998). Since photosynthesis is the sole mechanism of carbon assimilation in most

vascular plants, and water is likely to limit photosynthesis at some time scale in most

terrestrial environments, I expect coordination between photosynthetic capacity and plant

hydraulic properties. Therefore, measurements of plant hydraulic properties may be

related to other plant processes such as nutrient use and gross photosynthesis, thus

integrating leaf level processes into a more complete understanding of whole-plant

function.

This study was designed to examine the allometry of branch hydraulic architecture,

xylem biophysical properties and suites of leaf photosynthetic traits among 20 species of

canopy trees growing in two Panamanian lowland forests. The primary objective was to

determine the extent to which variation in leaf area-based hydraulic properties and xylem









biophysical properties can explain variation in leaf gas exchange characteristics. Specific

questions included: 1) Does allocation to leaf photosynthetic capacity correspond to

capacity for hydraulic water supply? 2) Is hydraulic conductivity correlated with

physiological leaf traits such as life span, nitrogen concentration and water use

efficiency? 3) Does wood density constrain hydraulic conductivity to affect leaf gas

exchange? 4) Can photosynthetic traits be predicted from xylem biophysical properties?

Materials and Methods

Study Site and Species

The study was conducted from two canopy cranes operated by the Smithsonian

Tropical Research Institute (STRI) in the Republic of Panama. Each crane is equipped

with a gondola suspended by cables from a rotating boom that allows coverage of

approximately 0.82 ha of forest. One crane is located in Parque Metropolitano, a

secondary dry forest on the edge of Panama City that receives approximately 1800 mm of

precipitation annually with a distinct dry season between December and April (Condit

1998). The other crane is located in an old-growth forest at Fort Sherman on the

Caribbean side of the Panamanian Isthmus where mean annual precipitation is 3100 mm

and the dry season is shorter and less intense than at Parque Metropolitano. During the

dry season of 2002 (Feb-Mar), physiological and morphological characteristics in the

upper crown of one to four individuals of 20 canopy tree species were measured. Both

rare and common species were included in the study (Table 4-1). I measured at least five

sun-exposed terminal branches in all species. In rare species with only one individual at

the study site, sample branches were taken from different portions of the crown.









Gas Exchange Measurements

Maximum rates of net CO2 assimilation (A) and stomatal conductance (gs) were

measured with a portable photosynthesis system (Model 6400, Li-Cor, Inc., Lincoln, NE,

USA) between 0700 and 1100 h. Two to four newly formed mature leaves per branch

were measured at 400 [[mol C02, and 1200 [tmol m-2 s-1 photosynthetic photon flux

density (PPFD) provided by a red blue light source (model 6400-02B #SI-710, Li-Cor,

Inc.). Only gas exchange values measured at a leaf to air vapor pressure deficit (VPDi)

below 1.5 kPa were used because significant stomatal closure above this value was

observed.

Hydraulic Conductivity

Hydraulic conductivity (kh) was measured on twigs excised from 1.0-1.5-m- long

branches. Branches were longer than the measured maximum vessel length determined

by the air pressure technique (Ewers and Fisher 1989). Upper crown branches exposed to

full sun were cut directly after gas exchange measurements on days with similar

environmental conditions and transported to the laboratory. In the laboratory, the first

fully developed twigs supporting the leaves on which gas exchange was measured were

excised from the rest of the branch under filtered water (0.2 [tm) to prevent xylem

embolism. After removing 3 to 4 mm of bark from each end, the cut ends were shaved

with Teflon coated razor blades and connected to the hydraulic conductivity apparatus

(Sperry et al. 1988). Pith areas were plugged with plasticine when necessary. The

hydraulic conductivity apparatus consisted of a beaker supplying filtered (0.2 [tm) water

under low (1.4 kPa) gravitational pressure to the stem. A low hydraulic head insured no

embolisms were removed and the apparatus was frequently flushed with 10% bleach









solution to avoid microbial growth. Flow rates were determined volumetrically 5-15

minutes after connection when they became steady. Initial kh was estimated as the rate of

water flux (J, mmol s-') in a stem when a pressure gradient (dP/dx, MPa m-1) was applied

across the stem

kh = J/(dP/dx). (4-1)

Maximum kh was measured after a 20-min high-pressure flush from a captive air

tank when flow rates reached steady values. After hydraulic measurements, stem length,

xylem diameter and pith diameter were measured for the calculation of kh and specific

conductivity (ks), kh per unit xylem area. Leaf area distal to the branch segment where kh

was measured was recorded with a leaf area meter (model LI-3100, Li-Cor Inc.), and

used to calculate leaf specific conductivity (LSC), kh per unit leaf area.

Wood Density

After conductivity measurements, outer bark, phloem and pith were removed from

sample stems, wet mass was determined and stems were dried for 48 h at 650C and

weighed to determine the saturated water content (SWC). Wood density was measured as

the ratio of xylem dry mass to xylem volume. Percent loss of conductivity (PLC) was

calculated from the ratio of initial to maximum kh.

Foliar Analysis

After measurement of their area, gas exchange leaves were dried for 48 h at 650C

and weighed for the determination of specific leaf area (SLA; cm2 g-) so that

photosynthesis could be calculated both on a unit leaf area and leaf mass basis. Leaves

were ground in a Wiley mill and one composite sample per branch was analyzed for total

leaf nitrogen and carbon isotope discrimination (613C) on an elemental analyzer









connected to a continuous flow mass spectrometer at the University of Idaho. This

allowed us to express photosynthesis per unit leaf nitrogen (AN) and to evaluate intrinsic

water use efficiency estimated by discrimination against 13C (Farquhar and Richards

1984).

Results

There were significant positive relationships between mean species Aarea and

stomatal conductance (gs) when plotted against leaf specific hydraulic conductance (LSC;

Figure 4-1). Area and gs showed substantial increases with increasing initial (LSCinitial)

and maximum (LSCmax) measurements, but LSCmax was a much stronger predictor

(Figure 4-1) suggesting that plants allocate photosynthetic capacity proportionally to the

maximum operational capacity of the xylem. There were also significant positive

relationships between Aarea and specific conductivity (ks). When Aarea was regressed

against initial and maximum ks, the relationships were again stronger with maximum than

initial conductance (r2=0.22; P<0.05 and r2=0.39; P<0.005, respectively), consistent with

the pattern observed in LSC.

Leaf nitrogen per unit area (Narea) varied independently of LSCmax (Figure 4-2).

However, photosynthesis per unit N (AN) increased significantly with increasing LSCmax

suggesting that hydraulic constraints limit the instantaneous efficiency with which N

appears to be used (Fig 4-2). Median leaf life span was negatively related to LSCmax

suggesting that there is an evolutionary tradeoff between long-lived leaves and high LSC

(r2=0.24; P<0.1) similar to the functional relationship between leaf photosynthetic

capacity and leaf life span (Reich et al. 1992). Leaf carbon isotope discrimination (613C),

a measure of intrinsic water use efficiency was not significantly related to LSCinitial









(r2=0.00; P=0.86) or LSCmax (r2=0.00; P=0.78) when measurement branches were

averaged per species. However, instantaneous water use efficiency (Aarea/gs) declined

with increasing LSCmax, indicating a tradeoff between hydraulic conductivity and water

loss per unit carbon gain (Figure 4-2).

Relationships between wood density and photosynthetic and hydraulic

characteristics generally showed negative correlations. Stem saturated water content

(SWC) was negatively correlated with wood density demonstrating a possible tradeoff

between sapwood water storage and mechanical strength (Figure 4-3). LSCmax was

negatively related to wood density suggesting that high LSC can be achieved by

producing larger vessels or higher vessel density at the expense of having lighter wood

with potentially reduced mechanical strength (Figure 4-3). Both Aarea (Figure 4-3) and

Amass (r2=0.50; P<0.001) were negatively correlated with wood density; thus wood

density appears to be negatively related to photosynthetic capacity through its effect on

hydraulic conductance per unit leaf area.

Discussion

The results suggest that plant traits regulating photosynthetic and hydraulic

capacity are highly interdependent. Leaf specific hydraulic conductivity (LSC) varied

proportionally with photosynthetic CO2 assimilation (A) and this relationship is consistent

within a group of canopy trees in Panamanian lowland forest. Additionally, it appears

that tradeoffs in relation to wood density and leaf life span allow diversity of allocation

patterns among coexisting species in the plant community. For example, species with low

photosynthetic and hydraulic capacity exhibited longer leaf life spans, potentially

minimizing the nutrient cost of leaf replacement (Chapin 1980). Species with lower

photosynthetic and hydraulic capacity also have denser wood, which may be a result of









smaller diameter vessels that constrain maximum xylem conductance but also allow

increased biomechanical support and an increased xylem pressure threshold (Hacke et al.

2001). Furthermore, species with lower photosynthetic and hydraulic capacity show

evidence of higher water use efficiency, thus tradeoffs result in several solutions to

balancing carbon gain with water loss. These results indicate that understanding xylem

hydraulic capacity adds information to leaf level measurements when comparing species

and links photosynthetic allocation patterns with processes at the branch and whole

organism levels of organization.

The relationship between photosynthesis and hydraulic conductance reflects a

balance between carbon gain and water loss, thus the primary feature regulated by LSC is

probably stomatal conductance (gs), and therefore transpiration. Species usually have

specific operating ranges or minimum values of leaf water potential governed by stomatal

regulation (Meinzer and Grantz 1991). Thus, if LSC increases as a result of partial

defoliation or leaf shading, gs usually increases, but leaf water potential remains about the

same because the transpiration/LSC relationship is conserved (Meinzer and Grantz 1991;

Pataki et al. 1998). Therefore, the coordination of photosynthetic and hydraulic capacity

may be more a consequence of the often reported coupling of A and gs, and of gs with

LSC rather than a direct relationship between A and LSC. A appears to belong to a suite

of coordinated characteristics related to plant hydraulic architecture and wood density.

High rates of photosynthesis in relation to LSC will not impair plant functioning, but

excessive transpiration can result in xylem cavitation and turgor loss. Therefore, the data

showing proportional allocation to photosynthetic and hydraulic capacity may mean that









species only invest in transport capacity that can be supported without experiencing

physiological damage.

Understanding the contributions of individual species to productivity and resource

use in species-rich tropical forest is challenging. Previous attempts to understand and

model inter-species physiological variation have involved dividing species into discrete

functional groups. Recent studies of regulation of water use among diverse tropical forest

canopy tree species have shown that contrasting patterns of regulation at the leaf level

tend to converge when appropriate scaling and normalizing factors are applied (Andrade

et al. 1998; Goldstein et al. 1998; Meinzer et al. 1997). The data suggest that continuous

scaling relationships work to collapse inter-species variation into functional relationships

that reflect plant capacity for carbon gain and tradeoffs that allow a broad spectrum of

capacities. Therefore, understanding where a species falls along a functional evolutionary

tradeoff continuum may be more informative than the discrete functional group approach.













Table 4-1. Area-based maximum photosynthetic rate (Aarea), maximum leaf specific
hydraulic conductivity (LSCmax), and wood density for study species from two
lowland tropical forest sites in Panama.
Species Family area LSCmax Wood density
(manol ms2 s-1) (mmol m1 s'1 MPa ') (g cm3)


Fort Sherman
Aspidosperma cruenta
Dussia mundia
Guateria dumentosa
Humiriastrum diguense
Manilkara bidentata
Marila laxifolia
Miconia borealis
Nectandra purpurescens
Ocotea ira
Poulsenia armata
Pourouma bicolor
Simarouba amara
Tapirira guianense
Trattinickia aspera
Virola sebifera
Vochysiaferruginea
Parque Metropolitano
( hi ,, 1l I ito,, cainito
Cordia alliodora
Ficus insipida
Luehea seemannii


Apocynaceae
Fabaceae
Annonaceae
Hernandiaceae
Sapotaceae
Clusiaceae
Melastomataceae
Lauraceae
Lauraceae
Moraceae
Moraceae
Simaroubaceae
Anacardiaceae
Bursuraceae
Myristicaceae
Vochysiaceae

Sapotaceae
Boraginaceae
Moraceae
Tiliaceae


9.7
12.3
12.2
11.2
10.3
9.9
16.8
11.1
12.6
11.8
13.7
17.5
12.9
12.2
13.5
18.3

9.9
15.4
19.2
17.0


15.81
37.44
38.42
34.05
42.04
29.89
81.00
58.91
45.78
39.49
49.90
102.37
61.99
43.00
60.78
120.69

33.84
61.59
123.14
115.19


0.70
0.53
0.42
0.58
0.61
0.48
0.50
0.55
0.58
0.43
0.45
0.41
0.43
0.57
0.50
0.35

0.61
0.47
0.34
0.33







56



20
r2 0.63 r2=0.88
18

^ 16
E A
0 14
E 0
12 a
1o A B
0)5
0 A A



0.5 r2 0.67 r2 =0.62

0- .4
U)

E 0 V or
0 o
O o O o o
E 0.2
w aC ^ D
) */ A 0
0.1

0.0
0 20 40 60 80 0 20 40 60 80 100 120

LSCinitial (mmol m-1 s-1 MPa-1) LSCmax (mmol m1 s1 MPa1)



Figure 4-1. Leaf photosynthetic rate per unit area (Area; A-B) and stomatal conductance
(gs; C-D) as a function of initial and maximum leaf specific hydraulic
conductivity (LSC) for 20 canopy tree species of lowland tropical forest in
Panama. Symbols: A. cruenta (black circle), C. cainito (black square), C.
alliodora (black triangle), D. mundia (black upside down triangle), F. insipida
(black diamond), G. dumentosa (black hexagon), H. diguense (white circle),
L. seemannii (white square), M. bidentata (white triangle), M. laxifolia (white
upside down triangle), M. borealis (white diamond), N. purpurescens (white
hexagon), 0. ira (white crosshair circle), P. armata (white crosshair square),
P. bicolor (white crosshair triangle), S. amara (white crosshair upside down
triangle), T guianense (white crosshair diamond), T aspera (white crosshair
hexagon), V. sebifera (black crosshair circle), V. ferruginea (black crosshair
diamond).







57






0.24
*
0.20

E 0.16 -

E 0.12 -
CU
a 0.08
CU
Z
0.04

r 0.00
o.oo --------------------

r, 140
Z 120 0
75
E. 100-

S80-
0 0
75 60
E
S 40 r2 0.59
z
Z
< 20
80

75 70
E v
5 60 -
E
:= 50 e

40 0- <--

Sr2 = 0.44
20
0 20 40 60 80 100 120

LSCmax (mmol m-1 s1 MPa-1)



Figure 4-2. Leaf nitrogen per unit area (Narea), leaf photosynthetic rate per unit leaf
nitrogen (AN), and instantaneous water use efficiency (Aarea/gs) as a function of
maximum leaf specific hydraulic conductivity (LSCmax) for 20 canopy tree
species in lowland Panama. Symbols as in Figure 4-1.














200


160


120


80


0U


Cn)

E
"6
E
E




Cn)
CN
E
-5
"T



E

co


8 L
0.3


r2= 0.75


r2 = 0.68


> r I 0.63
0 r2 =0.63


0.4 0.5 0.6 0.7
0.4 0.5 0.6 0.7


Wood density (g cm3)



Figure 4-3. Stem saturated water content, maximum leaf specific hydraulic conductivity
(LSCmax), and photosynthetic rate per unit leaf area (Aarea), as a function of
stem wood density for 20 species of lowland forest canopy trees in Panama.
Symbols as in Figure 4-1.


O














CHAPTER 5
LEAF DECOMPOSITION IN A WET TROPICAL FOREST: LINKING LEAF TRAITS
WITH NUTRIENT CYCLING

Introduction

Climate, substrate, and the decomposer community are the three primary factors

controlling decomposition processes. In tropical wet forest, it has been proposed that the

structure and chemistry of substrate should have a relatively large effect, because tropical

forest tends to have warm moist climate and high biological diversity (Meentemeyer

1978). Substrate chemistry of leaf litter is a product of the resources allocated to the

living leaf, minus the nutrients retranslocated during senescence. Several general patterns

relating species to litter quality have been noted. For example, species from low resource

habitats often have leaves of low nutrient content that may decompose relatively slowly,

whereas species in high resource habitats generally produce leaves that have higher

nutrient content and decomposable readily (Chapin 1980). Furthermore, leaves share

interdependent characteristics such as photosynthetic rate, N concentration, specific leaf

area and leaf life span (Reich et al. 1997). The purpose of this paper is to evaluate the

extent to which such functional leaf characteristics predict litter quality and

decomposition rates among species in a wet lowland tropical forest.

Of the variety of leaf chemical components that are correlated with decomposition

rate, it is generally accepted that litter lignin:N ratio is the strongest predictor

(Cornelissen 1996; Hobbie 1996; Melillo et al. 1982). Considerably less work has gone

into investigating which leaf chemical, structural, or physiological traits of functioning









leaves predict decomposition (Cornelissen 1996; Comelissen and Thompson 1997;

Grime et al. 1996; Wardle et al. 1998). This approach is of interest because it allows the

potential for understanding how growth and reproduction are related to the

decomposability of plant tissue. For example, Cornelissen and Thompson (1997) found

that specific leaf area (SLA), leaf life span, and leaf N were significant predictors of %

mass loss during decomposition among herbaceous monocots, suggesting that

relationships between leaf physiological traits of living leaves may be extended into leaf

litter decomposition. Furthermore, antiherbivore defenses have been found to continue to

work against decomposing organisms (Grime et al. 1996; Wardle et al. 2002). Thus leaf

physiology and antiherbivore defenses, two major determinants of leaf structure and

function, might be related to the effects of any one species on decomposition and nutrient

cycling.

It would be informative to link leaf physiology and antiherbivore defense to

decomposition because of the body of ecological literature that relates suites of leaf

functional traits to growth strategies in specific environmental regimes (Chapin 1993;

Coley et al. 1985; Grime 1977; Westoby 1998). Leaf physiological traits, especially

photosynthesis and water use efficiency, have been correlated with fitness in some

studies, although mostly through correlations with other traits (Arntz and Delph 2001;

Dudley 1996; Geber 1990; Lechowicz 1984). Therefore, it appears that many traits are

correlated along a minimal number of functional axes. The goal of this study is to

determine if it is possible to extend our understanding of correlated plant traits to include

leaf decomposition rate.









Materials and Methods

Study Site and Species

The study was conducted at the Fort Sherman canopy crane site on the Caribbean

coast of Central Panama. The site contains a 5-ha plot of lowland tropical forest within

the 12,000 ha San Lorenzo protected area. Mean annual rainfall at the site is 3100 mm,

with a mild dry season from January to March. I accessed the canopy with a 52-m-tall

construction crane equipped with a gondola and operated by the Smithsonian Tropical

Research Institute. Thirty six plant species that represent a broad selection of growth

forms including palms, lianas, monocot herbs, canopy trees and pioneer trees were

selected for study (Table 5-1).

Litter Collection and Decomposition

Leaf litter was collected during the dry season, between January and April, 2001,

when many species shed or exchange leaves. Senescent leaves were collected by hand

directly from at least three individuals of each study species. Entire leaves with complete

discoloration were gently shaken and harvested only if they came off the plant with a

light touch, indicating a well-formed abscission zone. Palm fronds and monocot herb

leaves remain on the plant for several months after retranslocation, so the most recent

senescing leaf or frond that had no remaining green pigmentation was selected.

Leaves were air dried in an air-conditioned laboratory (45% RH and 24C) for >1

month. Ten grams of litter from each species was set aside for initial litter quality

chemical analyses. Two to three grams of litter from each species were placed in 1X1 mm

mesh nylon-covered fiberglass window screen. Leaves of several species can weigh more

than 3 g and for these species one entire leaf was placed in each bag. For palms, a leaflet

was separated from the rachis and treated as a leaf. Four sizes of litter bags: 10X10,









10X30, 20X20 and 30X30 cm were used to accommodate different leaf sizes and

minimize folding. Litterbags were heat sealed with a dry iron and placed in the field on

March 30, 2001, about one month before the beginning of the rainy season that year.

Dowel rods (1/4" diameter, 15 cm long) were placed in the field with litterbags as a

standard to compare with other decomposition studies (LIDET 1995).

At 1, 3, 6, 14, and 24 months, 5 bags per species were collected from the field.

Bags were gently rinsed with distilled water to remove adhered soil particles and any

roots that had grown into the bags were removed with tweezers. A subsample of each

species was weighed fresh for moisture content, while the rest of the bags were placed in

the freezer overnight. The contents of each litterbag were then dried for 48 hours at 650C.

Dried samples were weighed and ground in a Wiley mill through a 40 size mesh.

Initial Litter Quality

A subsample of litter from each species was dried for 48 hours at 650C for the

determination of initial litter quality. Carbon fraction analyses were performed using a

series of increasingly aggressive extractants (Ryan et al. 1989). Dried litter samples were

digested in a detergent solution which separated non-polar extractives (cell contents)

from neutral detergent fiber (NDF), which includes cell wall constituents and fractions

that are not immediately nutritionally available. A dilute acid detergent solution was then

used to determine acid detergent fiber (ADF, lignocellulose) before lignin was separated

from cellulose in 72% H2S04. A separate subsample was analyzed for C and N using an

elemental analyzer (Carlo Erba, Milano, Italy).

Gas Exchange and Leaf Chemistry

In the wet season of 2000 and 2001 (June-November), maximum rates of net CO2

assimilation (A) and stomatal conductance (gs) were measured with an infrared gas









analyzer (Model 6400, Li-Cor, Inc., Lincoln, NE, USA) between 0700 and 1100 h. Five

newly formed mature leaves from three individuals of each species were measured at 370

[tmol mol-1, (slightly higher than ambient CO2 concentration), and 1500 [tmol m-2 s-1

photosynthetic photon flux density (PFD) provided by a red blue light source (model

6400-02B #SI-710, Li-Cor, Inc.). Only gas exchange values measured at a leaf-to-air

vapor pressure deficit (VPD1) below 1.5 kPa were used because significant stomatal

closure was observed above this value. Gas exchange was measured on leaves growing in

the highest light environment in which the species tends to grow. For canopy trees,

lianas, pioneer trees and some palms, terminal leaves were accessed using the canopy

crane. For monocot herbs and understory palms, measurements were made from the

forest floor.

Following photosynthetic measurements, lamina thickness was measured between

primary and secondary veins with a digital caliper (Mitutoyo, Japan). Leaves were dried

for 48 hours at 650C, weighed for determination of specific leaf area (SLA) and ball

milled to a fine powder. All leaves from the same individual were pooled for chemical

analysis. One composite sample from each individual was analyzed for C and N using an

elemental analyzer (Carlo Erba, Milano, Italy). Leaf toughness was measured in the field

on freshly collected leaves with a penetrometer (Pesola, Switzerland), which measured

the maximum force during punching through the lamina between primary and secondary

veins at a steady, slow rate with a 1 mm diameter plunger. Such measurements do not

replicate the action by which herbivores puncture leaves (fracture toughness) but leaf

toughness was correlated with fracture toughness in a study of 42 tropical tree species

(Choong et al. 1992), so I use it as an index of allocation to structure and potential









resistance to decomposition. Leaf polyphenol concentration was analyzed with the

Prussian Blue Procedure in a subset of the species studied, including canopy trees,

pioneer trees, lianas, and one palm (S. Joseph Wright, unpublished data).

Results

Decomposition rate (k) varied from a minimum of 0.41 in the canopy tree

Vochysiaferruginea to a maximum of 4.58 in the pioneer tree, Piper hispidum. Overall,

there were no significant differences in k among plant growth forms (F=1.76; P= 0.1617;

Table 5-2). Several indices of litter quality varied significantly between the monocots,

(palms and monocot herbs), and dicots (canopy trees, pioneer trees, and lianas). For

example, non-polar extractives were significantly lower in monocot herbs and palms than

other growth forms (F=7.86; P=0.0002). Litter cellulose concentration was significantly

higher in monocot herbs and palms (F=7.16; P=0.0003), and acid detergent fiber (ADF)

was higher in palms than other growth forms (F=6.43; P=0.0007).

Litter lignin concentration was negatively related to decomposition rate (Table

5-3, Figure 5-1). Although litter N concentration was not a significant predictor of

decomposition, the composite variable, litter lignin:N, was the strongest predictor of

decomposition rate (Table 5-3, Figure 5-1). Litter C concentration and C:N were also

negatively related to decomposition (Table 5-3, Figure 5-2). No other indices of litter

quality were related to decomposition except ADF, and this relationship was weak

(Table 5-3).

Several chemical and structural traits of living leaves were significantly related to

decomposition. LeafN and C:N were significantly related to decomposition (Table 5-4,

Figure 5-3). SLA was positively related to decomposition (Table 5-4, Figure 5-4), and

species with tougher leaves tended to decompose more slowly (Table 5-4). LeafN and









SLA, two of the strongest leaf predictors of decomposition were also significant

predictors of photosynthetic rate (Figure 5-5), indicating a relationship between leaf

physiology and decomposition through these structural and chemical traits.

Although lignin and lignin:N were the strongest predictors of decomposition,

there were several species with low lignin concentrations that decomposed slowly. A

general linear model including the effects of lignin and polyphenol concentration

explained more of the variance in decomposition than lignin alone (Table 5-5). This

result suggests that although lignin is probably the best general leaf defensive

characteristic for predicting decomposition, other antiherbivore defenses may play a

similar role in deterring decomposers following senescence (Grime et al. 1996; Wardle et

al. 2002), and when combined with lignin appear to explain a greater proportion of the

variance.

Discussion

Plant Growth Forms and Ecosystem Processes

In high diversity tropical ecosystems, it has been difficult to relate resource use

characteristics of individual plant species to system-wide processes such as nutrient

storage and cycling. This predicament has led to the functional group approach, in which

species with similar resource use characteristics are grouped into a smaller, more

manageable number of functional groups. This approach greatly simplifies modeling of

plant processes and has achieved success in many ecosystems (Chapin et al. 1995; Ewel

and Bigelow 1996; Korner 1993; Tilman et al. 1997). However, it appears that

continuous functions reflecting convergence in leaf physiology can also be used to

simplify the effects of many species. Therefore, my results suggest that for predicting the

effect of a species on an ecosystem process such as litter decomposition, it is more









important to know the value of a continuous variable such as litter lignin:N than any

discrete classification such as growth form, canopy position, successional status, or

phylogenetic association.

Litter Quality and Decomposition

The result that litter lignin:N is the strongest predictor of decomposition reported

in this study is consistent with numerous other decomposition studies (Cornelissen 1996;

Hobbie 1996; Melillo et al. 1982; Ostertag and Hobbie 1999). Similarly, the result that

litter C concentration and C:N are significantly related to decomposition corroborates

much of what exists in the ecological literature. Mechanistically, it is important to note

that the ratio of lignin to N reflects a ratio between leaf defense and photosynthetic

potential. High leaf lignin concentration is a characteristic of species in low resource

habitats and species with long-lived leaves, and is considered a general antiherbivore

defense (Chapin 1980; Coley 1983; Reich et al. 1997). To some degree, litter N is a

reflection of the physiological capacity of the leaf. Therefore, litter lignin:N is a ratio of

defense and structure to physiological capacity, although effects on decomposition are

more strongly driven by lignin. Viewed in this respect, the strongest predictor of litter

decomposition integrates plant allocation along the axes of defense and potentially

carbon gain.

Photosynthesis and Decomposition

Photosynthesis is the primary function of leaves and therefore most of the features

of leaves are under strong selection to maximize carbon gain over the life of the leaf

(Kikuzawa 1991; Mulkey et al. 1995; Reich et al. 1997). Ecologists have identified a

suite of highly interdependent leaf characteristics that occur repeatedly and appear

important in the exploitation of specific habitats (Chapin 1980). Specific leaf area (SLA)









and leaf N are two of these leaf characteristics that are positively related with

photosynthesis across broad groups of plant species (Reich et al. 1992). SLA and leaf N

are also the strongest leaf predictors of decomposition in this study and suggest that

allocation to photosynthetic capacity is related to decomposition through these

correlations (Figure 5-6). Interestingly, leafN was correlated with decomposition rate,

but litter N was not. Leaf N may be related to a host of other leaf characteristics that

influence decomposition. Although photosynthetic rate and decomposition rate may show

a positive correlation, litter does not photosynthesize. Therefore, it is necessary to

consider chemical and structural leaf components that are related to photosynthesis to

understand how decomposition is a product of selection to maximize carbon gain in

leaves.







68


Table 5-1. Summary of study species and leaf litter decomposition rate (k) organized by
growth form. All leaf litter material was collected in lowland wet tropical
forest at Fort Sherman, Panama.


Species
Aspidosperma cruenta
Brosimum utile
Calophyllum longifolium
Carapa guianensis
Dussia mundia
Manilkara bidentata
Nectandra purpurascens
Pourouma bicolor
Simarouba amara
Tapirira guianensis
Vochysiaferruginea
Apeiba membranaceae
Cecropia insignis
Jacaranda copaia
Clidemia octona
Ochroma pyrmidale
Piper hispidum
Trema micrantha
Arrabidaea verrucosa
Cayaponia granatensis
Doliocarpus dentatus
Heisteria scandens
Maripa panamensis
Phryganocydia corymbosa
Pleonotoma variabilis
Tontelea richardii
Costus pulverulentus
Dillenhacl hi pittieri
Heliconia f, g., ,,,ni,,i
Stromanthejacquinii
Zingiber rtic ii/ile
Calyptrogyne costatifrons
Carloduvica palmata
Genoma cuneata
Oenocarpus mapora
Socratea esxorrhiza


Family
Apocynaceae
Moraceae
Clusiaceae
Meliaceae
Leguminosae
Sapotaceae
Lauraceae
Cecropiaceae
Simaroubaceae
Anacardiaceae
Vochysiaceae
Tiliaceae
Cecropiaceae
Bignoniaceae
Melastomataceae
Bombacaceae
Piperaceae
Ulmaceae
Bignoniaceae
Cucurbitaceae
Dilleniaceae
Olacaceae
Convolvulaceae
Bignoniaceae
Bignoniaceae
Hippocrateaceae
Zingiberaceae
Araceae
Heliconiaceae
Maranthaceae
Zingiberaceae
Palmae
Cyclanthaceae
Palmae
Palmae
Palmae


Growth form
Canopy tree
Canopy tree
Canopy tree
Canopy tree
Canopy tree
Canopy tree
Canopy tree
Canopy tree
Canopy tree
Canopy tree
Canopy tree
Pioneer tree
Pioneer tree
Pioneer tree
Pioneer tree
Pioneer tree
Pioneer tree
Pioneer tree
Liana
Liana
Liana
Liana
Liana
Liana
Liana
Liana
Monocot herb
Monocot herb
Monocot herb
Monocot herb
Monocot herb
Palm
Palm
Palm
Palm
Palm


k (yr-)
1.27
0.47
0.46
0.45
1.11
0.66
0.57
0.44
1.01
0.96
0.41
1.05
1.00
0.80
0.69
0.58
4.58
1.55
0.72
0.72
0.44
1.40
0.82
1.24
0.48
0.56
0.64
3.18
0.81
0.80
1.76
0.61
0.90
0.79
0.64
0.46


Dowels









Table 5-2. Leaf litter decomposition rates and initial litter quality of study species from
Fort Sherman, Panama, averaged by growth form. Values are mean 1 SE.
Values with the same letter are not significantly different (P>0.05).


k (yr-1)

C (%)
N (%)
C:N
Lignin (%)
Lignin:N
Non-polar
extractives
(%)
Cellulose
(%)
Fiber-ADF
(%)


Canopy
trees
(n=11)
0.71a + 0.10

47.27a 1.38
0.90a + 0.09
58.42a + 6.42
16.03a + 1.66
19.90a + 3.34
54.23a + 3.49


18.41a + 1.90

45.77a + 3.49


Pioneer trees

(n=7)
1.47a + 0.58

43.23a + 2.66
1.22a + 0.18
39.23a + 5.95
13.67a + 2.71
11.81a + 2.76
55.48a + 5.07


18.01a + 3.00

44.52a + 5.07


Lianas

(n=8)
0.80a 0.13

46.45a + 2.07
1.10a + 0.08
44.42a + 5.57
16.46a + 3.17
16.80a + 4.93
50.35a + 3.01


19.56a + 2.72

49.65a + 3.01


Monocot
herbs
(n=5)
1.44a + .054

40.71a + 1.40
1.02a + 0.06
40.29a + 1.44
11.96a+ 1.62
11.74a+ 1.37
35.81b + 6.33


27.67b + 1.71

64.18a + 6.33


Palms

(n=6)
0.68a + 0.08

41.98a 1.11
1.14a + 0.16
38.69a + 3.89
21.48a + 3.19
19.9a + 4.30
29.59b + 1.65


32.61b 1.82

70.39b+ 1.64









Table 5-3. Regressions of litter chemical parameters and leaf litter decomposition rate (k).
All plant growth forms were combined for analyses. Only statistically
significant regressions (P<0.1) are reported. ADF, acid detergent fiber. (n=36)
Parameter Regression equation r2 Significance
Litter C ln(k)=6.91-1.89-Litter C 0.17 0.0122
Litter C:N ln(k)=2.29-0.66-Litter C:N 0.15 0.0181
Lignin ln(k)=1.66-0.70-Lignin 0.30 0.0005
Lignin:N ln(k)=1.53-0.65-Lignin:N 0.38 <0.0001
ADF ln(k)=1.78-0.55-ADF 0.09 0.0772

Table 5-4. Regressions of leaf chemical and structural parameters and leaf litter
decomposition rate (k). All plant growth forms were combined for analyses.
Only statistically significant regressions (P<0.1) are reported. SLA, specific
leaf area. (n=36)
Parameter Regression equation r2 Significance
Leaf N ln(k)=-0.97+1.15-Leaf N 0.20 0.0059
Leaf C:N ln(k)=2.48-0.85-Leaf C:N 0.19 0.0077
Lamina toughness ln(k)=0.66-0.18-lamina toughness 0.10 0.0575
SLA ln(k)=-2.43+0.48-SLA 0.21 0.0050

Table 5-5. Results of general linear model for combined effects of leaf phenol (measured
with the Prussian Blue Procedure) and litter lignin on litter decomposition
rate. This analysis was conducted on a subset of the study species for which
phenol data was available. The analysis included canopy trees, pioneer trees,
lianas and one palm.
k
Parameter Estimate P
Intercept 2.005 0.0052
Phenols -1.243 0.0589
Lignin -0.839 0.0026
Phenols*Lignin 0.439 0.1146
Overall model p-value 0.0044
Model R2 0.49
n 23














2.0

1.5

1.0

0.5

0.0

-0.5

-1.0

-1.5


2.0

1.5

1.0

0.5

0.0

-0.5

-1.0

-1.5


1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0 3.2 3.4

In litter lignin (%)


2.0


3.5


In litter lignin:N


Figure 5-1. Relationship between leaf litter decomposition rate (k) and (A) initial litter
lignin concentration and (B) initial litter lignin to nitrogen ratio for 36 lowland
tropical wet forest plant species from Fort Sherman, Panama.


r = 0.30
o



V

S o


canopy trees o0 ,
pioneer trees v* e v
lianas
monocot herbs
palms


r2 = 0.38


0


0


0
o0 V
v


S v Os*

















1.5

1.0

0.5

0.0

-0.5

-1.0

-1.5
3.4


3.8 3.9 4.0 4.1


In litter C (%)


In litter C:N


Figure 5-2. Relationship between leaf litter decomposition rate (k) and (A) initial litter
carbon concentration and (B) initial litter carbon to nitrogen ratio for 36 wet
forest species. Symbols as in figure 5-1.


2.0


3.5 3.6 3.7


2
r = 0.17
0
V


0
v
Sv
o 0
0 0

S T .



2.0


1.5

1.0

0.5

0.0

-0.5

-1.0


-1.5


2
r2 =0.15
0




0


0 vo
V V

o S O

































-2
0.0


0.5 1.0


In leaf N (%)


-2
2.0


2.5 3.0 3.5


4.0


In leaf C:N



Figure 5-3. Relationship between leaf litter decomposition rate (k) and (A) nitrogen
concentration of living leaf and (B) carbon to nitrogen ratio of living leaf for
36 wet forest species. Symbols as in figure 5-1.


V
V
*^ S
0
*4U v V 0
0 V V
V 0
**v


r2 = 0.20


2 .19
r = 0.19


V
V

04


0 *O *







74






2 -
r2 = 0.21 o
V


0 v
y v
C ov
0 V
-0i V

-1



-2
3.5 4.0 4.5 5.0 5.5 6.0

In SLA (cm2 g-)


Figure 5-4. Relationship between leaf litter decomposition rate (k) and specific leaf area
(SLA) for 36 wet forest species. Symbols as in figure 5-1.
















0
0

0


r2 = 0.25


0 50 100 150 200 250


300 350


SLA (cm2 g-)


0
1.0


0
0






vo



r2 =0.21


1.5 2.0 2.5 3.0 3.5 4.0


Leaf N (%)




Figure 5-5. Relationship between leaf photosynthetic rate per unit mass (Amass) and (A)
specific leaf area (SLA) and (B) leaf nitrogen concentration for 36 wet forest
species. Symbols as in figure 5-1.


250


200


150


100


50


CT

0)

E

U,
U,
E


.* v


250


200


150


100


50


C,



0)
E

U,
U,
EO
E


I I I I I I






76












SLA


Amass k



LeafN






Figure 5-6. Schematic diagram depicting how photosynthesis is related to decomposition
rate (k) through correlation with specific leaf area (SLA) and leaf nitrogen
concentration.














CHAPTER 6
TEST OF GAS EXCHANGE MEASUREMENTS ON EXCISED BRANCHES OF TEN
TROPICAL TREE SPECIES: A TECHNICAL REPORT

Introduction

Plant physiological parameters that describe CO2 uptake and water loss are used to

model biosphere-atmosphere interactions at local and regional scales (Running and

Coughlan 1988; Williams et al. 1996) and to predict the sensitivity of vegetation to

climate change at regional to global scales (Neilson and Marks 1994). Measurements

from the upper canopy are important for calculations of net primary productivity (NPP)

and direct gas exchange measurements are needed to understand the extent to which NPP

is controlled by plant community type and environmental heterogeneity (Haxeltine and

Prentice 1996; Woodward et al. 1995). Towers, scaffolds, walkways and construction

cranes have been employed to obtain physiological parameters from intact branches in

the upper canopy of mature forest (Mulkey et al. 1996). However, these structures are not

available everywhere and there is a need for reliable estimates of leaf-level processes in

places where it is not practical or feasible to build such structures. Researchers have

measured gas exchange of canopy leaves by cutting or shooting small branches and re-

cutting stems under water to re-establish the xylem water column in the form of a

potometer before measurement (Dang et al. 1997; Koyama 1981; Reich et al. 1995; Reich

et al. 1998). Other researchers have measured gas exchange of montane forest trees

(Gerrish 1992) or conifers (Ginn et al. 1991; Samuelson 1998) by simply detaching a

group of leaves or fascicles and placing them in a cuvette for immediate measurement.









Although numerous physiological observations on excised foliage have been

published, there is evidence that excision may alter gas exchange measurements in

several ways. Even if a stem is double cut to re-establish the xylem water column,

cavitation, changes in xylem pressure, and reductions in hydraulic conductance can result

(Boari and Malone 1993; Stahlberg and Cosgrove 1995). These alterations in the xylem

stream can influence stomatal responses (Sperry et al. 1993; Williamson and Milburn

1995) and inhibit the transport of hormones and nutrients, which are important for

regulating shoot water potential and stomatal control (Tardieu and Davies 1993; Zhang

and Davies 1990). The time scale at which these changes occur is likely to be important

in determining the degree of excision-induced effects on gas exchange. However, these

responses vary widely across the plant kingdom leaving us with little information

regarding the reliability of gas exchange measurements on excised foliage; few studies

provide verification of how leaves on excised stems perform (Dang et al. 1997; Ginn et

al. 1991).

The purpose of this study is to compare gas exchange rates measured on excised

and attached branches of tropical forest canopy trees. I investigated how measurements of

gas exchange and biochemical parameters derived from photosynthetic light and CO2

response curves varied between leaves on excised and attached branches. Our main

objective was to determine if excision causes significant effects on measured gas

exchange rates and parameters, i.e. I wanted to know if the effect of excision on gas

exchange parameters is greater than the range of values observed on intact branches.

Second, I hoped to evaluate whether species differ in their gas exchange responses to

excision, and if there are any characteristics that might be used to predict the degree









excision-induced effects. Although measuring gas exchange on excised branches is a

convenient technique for acquiring data, it seems important to address the limitations of

this approach so as to prevent the collection of spurious data and to interpret the results of

studies that reported gas exchange results from excised stems.

Materials and Methods

Study Sites and Species

The study was conducted from two construction cranes operated by the

Smithsonian Tropical Research Institute (STRI) in the Republic of Panama. Each crane is

equipped with a gondola suspended by cables from a rotating boom that allows coverage

of approximately 0.82 ha of forest. The first crane is located in Parque Metropolitano, a

secondary dry forest on the edge of Panama City that receives approximately 1800 mm of

precipitation annually with a distinct dry season between December and April. The

second crane is located at Fort Sherman, on the Caribbean side of the Panamanian

Isthmus in old growth forest that receives approximately 3100 mm of precipitation

annually with a shorter and less intense dry season.

Ten tree species from 9 families, representing a variety of leaf morphology and

gas exchange rates were investigated. At Fort Sherman, Apeiba membranacea Spruce ex.

Benth. (Tiliaceae), Jacaranda copaia (Aubl.) D. Don (Bignoniaceae) and Vochysia

ferruginea Mart. in Mart. & Zucc. (Vochysiaceae) have relatively high rates of

photosynthesis and stomatal conductance (Table 6-1). Aspidosperma cruenta Woods

(Apocynaceae), Brosimum utile (Moraceae), Manilkara bidentata (Sapotaceae) and

Simarouba amara Aubl. var. typical Cronq. (Simaroubaceae), in contrast, have lower rates

of photosynthesis and stomatal conductance, and tend to have thicker leaves. At Parque

Metropolitan, Anacardium excelsum (Bertero & Balb.) Skeels (Anacardiaceae) has









relatively low rates of photosynthesis and stomatal conductance, whereas Pseudobombax

septenatum (Jacq.) Dug. (Bombacaceae) and Luehea seemannii Tr. & Planch. (Tiliaceae)

have higher rates of photosynthesis and stomatal conductance. All nomenclature follows

(Croat 1978).

Gas Exchange Measurements

Net CO2 assimilation (A) and stomatal conductance (gs) were measured with an

infrared gas analyzer (Model 6400, Li-Cor, Inc., Lincoln, NE, USA) on leaves of

branches with full sun exposure. One gas exchange measurement was taken on a newly

formed mature intact and attached leaf at ambient temperature, 37.26 Pa CO2 (equivalent

to 370 [tmol mol-1, slightly higher than ambient), and 1500 [tmol m-2 s-1 photosynthetic

photon flux density (PFD) provided by a red blue light source (model 6400-02B #SI-710,

Li-Cor, Inc.). The segment of branch was then excised 100 cm from the measured leaf

and immediately shortened to 50 cm by re-cutting under water to re-establish the xylem

water column. Successive gas exchange measurements were taken within 3 min and then

several times up to 60 min after excision.

The response of A to PFD was measured on newly formed fully expanded mature

leaves at a range of PFD from 0 to 1500 [tmol m-2 s-1, 37.26 Pa C02, and ambient relative

humidity. Measurements were made at ambient temperature unless leaf temperature

exceeded 33 C, in which case, peltier plates adjacent to the cuvette were used to

maintain the block temperature at 33 C. I first measured A at 1500 [tmol m-2 s-1 PFD,

then light was decreased in a stepwise fashion for a total of 10 measurement points down

to 0 [tmol m-2 S-1. Once the light response curve was completed, the branch was

immediately re-cut under water as described above and the curve was repeated in the









gondola with the stem in a potometer on a leaf of similar age adjacent to the original leaf.

An attempt was made to maintain the cut stem in sunlight to prevent stomatal closure as a

result of low light availability in the gondola. Trials were performed on two to four

branches per species. Parameters for light response curves were fit to a nonrectangular

hyperbola (Sims and Pearcy 1991), which allows calculation of apparent quantum yield

on the basis of incident photons (0), the light-saturated rate of gross CO2 assimilation at

saturating irradiance (Amax), the radius of curvature (0) and dark respiration (Rd).

The response of A to intercellular CO2 partial pressure (pi) was measured on three

branches per species. I conducted measurements on intact branches by first inducing the

leaf with 1700 [tmol m-2 s-1 PFD and 37.26 Pa CO2. Measurements were then taken at 0

Pa CO2 and the concentration was increased over a total of 14 measurement points to a

final concentration of 151 Pa. The branch was then cut and re-cut as described above and

a second CO2 response curve was conducted on a leaf of similar age adjacent to the

original leaf to avoid feedback inhibition.

Parameters for CO2 response curves were modeled by two equations that describe

CO2-limited and rubisco-limited rates of photosynthesis (Lambers et al. 1998; von

Caemmerer and Farquhar 1981). CO2-limited rate of photosynthesis [A(c)] can be

calculated as


Vc max(p- F)
A(c)= -- Rday (6-1)
p, + Km
Where Vcmax is the rate of CO2 assimilation at saturating pi, F is the CO2

compensation point, Km is the Michaelis-Menten constant for the carboxylation reaction









and Rday is the rate of respiration during photosynthesis. The rubisco-limited rate of

photosynthesis [A(j)] can be calculated as:

Jmax(pi- F)
A(j) = 4J mx(p, Rday
4(p, + 2F) (6-2)
(6-2)

Where Jmax is the maximum rate of electron transport and all other parameters

follow equation 1. Repeated measure analysis of variance (ANOVA) was used to assess

the effect of excision on light and CO2 response curves with branches as a random factor

nested within treatments.

Results and Discussion

Time Course Measurements

After excision, all species showed a reduction in both A and gs within 3 min

(Figures 6-1 and 6-2). However, after 10 min, most species showed fewer fluctuations

and appeared to stabilize for the remainder of the 60 min measurement period. This

pattern indicates that hurrying to take a measurement on an excised branch may

minimize, but not necessarily avoid changes in gas exchange rates caused by excision. At

Ft. Sherman, the three species with the highest A for attached leaves, A. membranacea, J.

copaia, and V. ferruginea (Table 6-1), also maintained the highest A at 60 min after

excision (Figure 6-1). The other four species at Ft. Sherman, A. cruenta, B. utile, M.

bidentata, and S. amara, suffered a greater reduction in A 60 min after cutting (Figure 6-

1) and all produce latex or resin in the stem. A. cruenta, B. utile and M bidentata, are

from plant families well-known for this characteristic: the Apocynaceae, Moraceae,

Sapotaceae, respectively. Simarouba amara has a clear yellow resin, also characteristic of

its family, the Simaroubaceae (Croat 1978). Wounding stimulates latex and resin

production (Kramer and Kozlowski 1979), which acts as a wound response and can









physically clog the xylem and prevent water loss and pathogen attack. However, once an

excised stem is in water, if latex or resin leaks in the vicinity of the cut section of active

xylem it can clog the xylem thus inhibiting water supply to the leaf and reducing A and

gs. Therefore, these data suggest that species with latex or resin do not perform well

following excision.

At Parque Metropolitano, L. seemannii maintained a mean of 86.5% of intact

photosynthetic rates 60 min after excision (Figure 6-1), whereas P. septenatum

maintained only 13.4 and 5.5% of the average intact values for A and gs, respectively,

after excision (Figs. 6-1 and 6-2). A. excelsum, maintained only 26.6 and 48.6% of intact

A and gs, respectively after excision. With the exception of P. septenatum, species at

Parque maintained higher percentages of intact gs than A 60 min after excision (Figs. 6-1

and 6-2). In contrast, all species in the wetter forest at Ft. Sherman maintained a higher

percentage of intact A, relative to g 60 min after excision (Figs. 6-1 and 6-2). This trend

may be indicative of greater stomatal sensitivity in wet forest trees as a result of

acclimation to humid conditions.

Response to Light

Photosynthetic light response curves following excision varied among species

(Figure 6-3). Species such as A. membranacea and L. seemannii showed no differences

between intact and excised branches, whereas B. utile and S. amara showed substantial

differences in light-saturated photosynthesis (Figure 6-3). Although the curve from the

excised branch was always lower, this difference was less than the variation between

leaves from different branches, so excision did not significantly reduce A as a function of

PFD in any of the six species investigated (Table 6-2). Therefore, the light response curve

of any excised leaf is likely to reach lower values than measurements on the same leaf









when it is attached, but these lower values are also likely to fall within the range of values

for attached leaves.

Excision decreased the calculated value of Amax in A. cruenta, B. utile, and S.

amara (Table 6-3) but not in the other three species investigated. Since Amax is probably

the most widely used parameter derived from light response curves and half of the

species tested showed significant reductions upon excision, estimates of Amax derived

from light response curves on excised foliage must be interpreted with caution. Minor

differences between 0, D, and Rd from intact and excised light response curves were also

observed in some cases (data not shown), but these differences were generally within the

95% confidence interval of the parameter in curve fitting.

Overall, most species appeared to perform better during light response curves than

over time course measurements. Maintaining a portion of the leaf in the constant

environment of the cuvette may help maintain consistency. How a leafs ability to

respond to short-term changes in light and VPD is affected following excision is

unknown, but measurements suggest that constant conditions may result in more accurate

measurements. (Reich et al. 1995) kept excised branches in sunlight, which is likely to

have maintained the leaf in an induced state. In that study, attached foliage from pioneer

species was compared with excised foliage from mature canopy species and it was found

that the excised foliage of canopy species produced lower gas exchange rates. Since

excision can negatively affect gas exchange rates in a systematic way, it is difficult to

discern whether the lower gas exchange rates of canopy trees observed by Reich et al.

(1995) are the result of excision-induced effects or represent characteristics that have

evolved to specific environmental regimes. In either case, the species-specific responses









to excision presented in this paper suggest that it is important to consider the sensitivity

to excision when comparing species or functional groups.

Response to CO2

The effect of branch excision on leaf photosynthetic CO2 responses was again

species specific (Figure 6-4). Excision changed the general shape of the curve in A.

cruenta B. utile, and S. amara, whereas in A. membranacea, L. seemannii, and V.

ferruginea, lower maxima were the main effect. L. seemannii and V. ferruginea were the

only species exhibiting a significant effect of excision on photosynthetic responses to pi

(Table 6-2). In the other four species, A as a function of pi was reduced following

excision, but the magnitude of the effect of excision was small compared to the range of

values encountered between branches and therefore, no statistical effect was detected.

The reduction in A during CO2 response curves following excision may be

indicative of non-uniform stomatal closure. Heterogeneous gas exchange over small areas

of the leaf has been shown to occur in response to water stress (Downton et al. 1988).

Heterogeneous stomatal closure has also been implicated in causing errors in estimating

pi and may explain observed reductions in A even when pi does not appear to be limiting

(Mansfield et al. 1990). Cheeseman (1991) demonstrated through simulation models that

the effects of heterogeneous stomatal behaviour can only account for minor effects in

calculating pi, and therefore, on CO2 response curves. Yet, the effects of patchy stomatal

closure on CO2 response curves presented by (Cheeseman 1991) are similar to the

patterns of reduction following excision in this study, suggesting that stem water status is

altered by excision and the reduction in gs is heterogeneous over the leaf surface.

Therefore, non-uniform stomatal behaviour may explain why for several species the main

effect of excision on the CO2 response curve was a slight reduction of the maximum. This









reduction probably reflects changes in stomatal behaviour rather than a reduction of the

biochemical functioning of the photosystem.

Excision reduced the calculated values of both Vcmax and Jmax in L. seemannii and S.

amara whereas in A. membranacea and A. cruenta, only the calculation of VCmax was

significantly affected (Table 6-3). In spite of multi-fold variation in Vcmax and Jmax among

species, and in some cases between treatments, there was strong correlation between

these two parameters (r2=0.89). Excision did not produce a significantly different

relationship (t=-0.174; P>0.25) between Jmax and Vcmax, so a single linear regression was

used. A slightly positive intercept of 13.7 [tmol m-2 s-1 was observed, along with a slope

of 1.63. These results are consistent with those reported in a review of 109 species

(Wullschleger 1993). In that study, Wullschleger (1993) also reported linear regression

between Jmax and Vcmax with a positive intercept and a slope of 1.64. The consistency of

the relationship between Jmax and Vcmax suggests that effects on water transport and

patchy stomatal behavior are responsible for the differences in calculating biochemical

parameters, rather than changes in biochemical activity.

Declining values of A at high pi in CO2 response curves on both attached and

excised branches indicate an apparent limitation by triose phosphate utilization (Sharkey

1985) in L. seemannii and V. ferruginea (Figure 6-4). Limitation of A by TPU was once

thought to occur rarely, but there is now increasing evidence that this limitation may be

more common under conditions of high light and high pi (Harley et al. 1992;

Wullschleger 1993). Tropical forest canopies often experience high light and our data

indicate that TPU limitation can be important for at least two species in these forests.









This finding has implications for modeling the responses of tropical forest to increases in

atmospheric CO2.

Conclusions

I investigated the utility of a technique for measuring gas exchange on excised

foliage in tropical forest canopy trees. Responses to excision were species specific and

there is some evidence that responses are related to phylogeny, potentially allowing

prediction of how a species might perform. All species with latex or resin performed

relatively poorly after excision and I suggest that these types of wound responses can

clog cut xylem ends, thus reducing water transport. Patterns of reduction of A and gs in

time course measurements, light and CO2 response curves appear to be mediated by

stomatal closure as opposed to reductions in biochemical photosynthetic capacity. I

recommend individual trials on all species to be investigated. It appears that anatomical,

morphological, and possibly, phylogenetic differences can influence the response of gas

exchange measurements to excision. Therefore, patterns of response to excision at

taxonomic levels higher than species, and different growth forms, such as trees, vines and

shrubs, should also be considered in broad species comparisons of photosynthetic

measurements on excised branches.









Table 6-1. Light-saturated rate of net CO2 assimilation per unit mass (Amass) and area
(Aarea), leaf diffusive conductance (gs), and specific leaf area (SLA) 1 S.D.
for 10 tropical tree species from the Republic of Panama.
Amass Aarea gs SLA
Species (nmol g-1 s-) ([,mol m-2 -1) (mmol m-2 s-1) (cm2 g1)

Parque Metropolitano
Anacardium exelsum 65.1 + 19.1 7.0 2.0 129.7 37.7 93.1 6.8
Luehea seemannii 114.8 12.6 14.8 + 1.6 592.2 140.0 79.5 11.9
Pseudobombax 146.8 20.3 12.4 + 1.7 347.0 + 57.1 118.6 + 7.7
septenatum

Fort Sherman
Apeiba membranacea 200.4 + 16.1 16.1 + 1.3 659.7 + 103.6 124.8 9.6
Aspidosperma cruenta 104.5 5.3 14.2 0.7 387.3 68.9 74.4 7.3
Brosimum utile 94.5 14.2 10.4 + 1.6 298.2 + 83.8 91.2 + 4.2
Jacaranda copaia 157.8 27.1 14.6 2.5 538.0 212.7 96.6 6.4
Manilkara bidentata 58.1 + 16.6 11.8 3.4 199.3 83.9 49.4 1.7
Simarouba amara 106.3 + 11.4 15.7 + 1.7 443.8 + 113.2 68.1 5.4
Vochysia ferruginea 134.5 7.5 15.2 + 0.8 670.2 + 83.9 89.5 9.7

Table 6-2. Repeated measures analysis of variance for the effects of excision on
photosynthetic light and CO2 response curves for 6 canopy tree species from
the Republic of Panama.
Light CO2
Species SS F P SS F P
Apeiba membranacea 13.20 4.38 0.171 0.22 2.98 0.227
Aspidosperma cruenta 9.27 12.90 0.173 50.93 40.84 0.099
Brosimum utile 10.36 2.46 0.361 76.13 1.35 0.452
Luehea seemannii 3.75 4.96 0.156 16.24 523 0.002
Simarouba amara 13.17 2.21 0.377 404.86 3.47 0.314
Vochysia ferruginea 4.58 18.30 0.146 13.82 4.97 0.038









Table 6-3. Means of model parameters from photosynthetic light and CO2 response
curves on leaves of intact (IN) and excised (EX) branches determined by
least-squared estimates. Values of light-saturated rate of gross photosynthesis
(Amax) from light response curves, and maximum rate of carboxylation (Vcmax)
and electron transport (Jmax) from CO2 response curves followed by the same
letter are not significantly different (P<0.05) based on Duncan's multiple
range test. Comparisons are only valid within species.
Species Method Amax VCmax Jmax
(|jmol m-2 s-1) (|jmol m-2 s-1) (|jmol m-2 S-1)

Apeiba membranacea IN 20.0a 51.9a 102.9a
EX 18.9a 42.5b 86.3a
Aspidosperma cruenta IN 16.0a 56.2a 92.3a
EX 13.8b 45.1b 81.4a
Brosimum utile IN 12.8a 35.0a 64.5a
EX 9.8b 29.2a 54.5a
Luehea seemannii IN 14.9a 38.3a 82.7a
EX 13.3a 27.5b 65.2b
Simarouba amara IN 17.9a 61.4a 108.8a
EX 15.6b 32.9b 59.9b
Vochysiaferruginea IN 16.6a 50.9a 108.1a
EX 14.8a 53.6a 104.0a














100 T ort Sherman

80

60

"C 40

c 20
U)
o --.--.-.--.--.
.c 100< Parque Metropolitano



60 -

40

20 -

0 1
0 10 20 30 40 50 60

Time after cut (min)


Figure 6-1. Percent of maximum photosynthesis of 3-5 leaves from excised branches at
two canopy crane sites: Fort Sherman and Parque Metropolitano, Republic of
Panama. Species at Fort Sherman: (filled circle) Apeiba membranacea, (open
circle) Aspidosperma cruenta, (filled upside down triange) Brosimum utile,
(open upside down triange) Jacaranda copaia, (filled square) Manilkara
bidentata, (open square) Simarouba amara var. typical and (filled diamond)
Vochysiaferruginea. Species at Parque Metropolitano: (open diamond)
Anacardium excelsum, (filled triange) Luehea seemannii, and (open triangle)
Pseudobombax septenatum.