Ecology of Native Oil-Producing Palms and Their Potential for Biofuel Production in Southwestern Amazonia

ECOLOGY OF NATIVE OIL-PRODUCING PALMS AND THEIR POTENTIAL FOR BIOFUEL PRODUCTION IN SOUTHWESTERN AMAZONIA

JOANNA MARIE TUCKER LIMA

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

2010

To my parents who taught me to appreciate and marvel at the natural world.

ACKNOWLEDGMENTS

This work is the culmination of a long journey and the encouragement and support of many people throughout my academic career. As this chapter of my life comes to a close, I wish to extend special thanks to my PhD advisor, Karen Kainer, who was a continuous and consistent inspiration and source of encouragement to me. Her expertise and insight enriched my research from beginning to end. I also thank Evandro

Ferreira for his contagious love for palms and mix of practical and scientific advice. His students, Janice and Anelena, as well as the research team at the Parque Zoobotanico at UFAC, especially Plinio, Lira, and Edir, were always willing to help with field work and logistics, for which I am very grateful. Anelise Regiane and her chemistry students at

UFAC (Thayna, Nubia and Marcia) gave unselfishly of their time to help me process palm fruits and run chemical analysis that I could never have done alone. I am indebted to Francis “Jack” Putz for sharing his curiosity for the natural world and his persistent search for answers to both basic and complex ecological issues that affect our daily lives. His prodding stretched my ideas and encouraged me to look beyond the easy answers. I wish to thank Emilio Bruna and Jane Southworth for their insightful feedback as I developed my dissertation, and for sharing their deep understanding in their fields of expertise. I am also grateful to Meghan Brennan and Christina Staudhammer for their brilliant help with statistical analyses.

I am truly grateful for all the financial support I received during my degree from various sources—NSF Integrative Graduate Education Research and Training Program,

Tropical Conservation and Development Program (UF), International Palm Society, and the Environmental Protection Agency. I cannot forget to mention the debt I owe to the principle figures during my academic formation, including Emilio Moran, Eduardo

Brondizio, Andrew Henderson, and Daniel Zarin. These men inspired me, awakened

my zeal for the Amazon, Brazil, and palm trees, and contributed the foundation, both

academic and personal, that brought me to this point. Friends and family have given

me unconditional support throughout this journey, and I thank them from the bottom of

my heart for their patience, wisdom, listening ears, and encouragement. I especially

thank my parents who unwaveringly stood by me in the good and the difficult times.

Last but not least, I thank my husband, Evandro, without whom this work would have

been impossible. I thank him for his reliability and wisdom as he helped with my

fieldwork, and for sharing his ideas and intimate understanding of Amazonian forests. I

thank him for his patience and willingness to stick it out during the “last year” of

finishing-up, and for supporting me unselfishly during that time. Finally, I thank God, for

bringing all these wonderful people into my life, for never letting me go, and for providing me the strength, hope, perseverance, intelligence, and faith to complete my dissertation.

TABLE OF CONTENTS

page

ACKNOWLEDGMENTS ...... 4

ABSTRACT ...... 12

CHAPTER

1 INTRODUCTION ...... 15

2 FLOWER COLOR VARIATION IN ATTALEA PHALERATA (ARECACEAE) ...... 19

Field Observations ...... 21 Phenology ...... 21 Flower Color Variation ...... 22 Color Polymorphism in Palms ...... 24 Possible Explanations for Color Polymorphism in Attalea phalerata Flowers ...... 27 Final Considerations ...... 28

3 DOES LANDSCAPE CHANGE ALTER REPRODUCTIVE PHENOLOGY, SEX EXPRESSION AND PRODUCTIVITY OF ATTALEA PALMS IN SOUTHWESTERN AMAZONIA? ...... 30

Introduction ...... 30 Materials and Methods ...... 34 Study Area ...... 34 Study Species ...... 35 Site Selection ...... 37 Data Collection ...... 37 Reproductive phenology ...... 37 Size and illumination measurements ...... 38 Palm dissections ...... 38 Statistical Analyses ...... 39 Results ...... 40 Comparison of Flower and Fruit Production in Forest and Pasture ...... 40 Flowering and fruiting phenology ...... 40 Flowering and fruiting frequency ...... 43 Associations between Reproduction, Light Availability, and Tree Size ...... 47 Palm Dissections ...... 48 Discussion ...... 49 Pasture Palms Out-Produced Forest Palms ...... 49 Phenology ...... 49 Hidden Mechanisms Control Sex Expression ...... 50 Sunlight and Plant Size Linked to Sex Expression and Productivity ...... 52 Sex Expression and Male Dominance ...... 53

Pollen Availability may Alter Operational Sex Ratios ...... 55 Conclusions ...... 56

4 RECOVERY OF ATTALEA PHALERATA MART. EX SPRENG. PALM POPULATIONS AFTER SLASH-AND-BURN AGRICULTURE IN SOUTHWESTERN AMAZONIA ...... 58

Introduction ...... 58 Study Area ...... 61 Study species ...... 62 Materials and Methods ...... 63 Research Design ...... 63 Data Collection ...... 64 Statistical Analysis ...... 65 Results ...... 66 Pre-Burn Palm Populations ...... 66 Immediate Effects of Slash-and-Burn ...... 68 Recovery ...... 69 Recruitment and survival ...... 69 Effect of slash-and-burn on A. phalerata populations ...... 71 Palm demography ...... 73 Leaf production and growth ...... 75 Discussion ...... 77 Palm Colonization or Persistence? ...... 77 Impediments to seedling recruitment ...... 78 Mechanisms of persistence...... 79 Continued Post-Fire Recovery ...... 81 Palm growth ...... 82 Plant demography ...... 84 Disturbance promotes palm dominance ...... 86 Ecological Pressures Build Resistance ...... 87

5 CAN EXISTING POPULATIONS OF NATIVE OLEAGINOUS PLANTS CONTRIBUTE TO BIOFUEL PRODUCTION IN AMAZONIA? FRUITING PHENOLOGY AND OIL YIELDS FROM TWO NATIVE ATTALEA PALMS IN SOUTHWESTERN AMAZONIA ...... 89

Introduction ...... 89 Methods ...... 92 Site Description ...... 92 Study Species ...... 92 Site Selection ...... 94 Data Collection ...... 95 Reproductive phenology ...... 95 Fruit bunch collection and processing ...... 95 Chemical Analysis ...... 96 Data Analysis ...... 97

Results ...... 97 Comparison of Fruit Production in Forest and Pasture ...... 97 Inflorescence production ...... 97 Infructescence production ...... 98 Fruiting phenology ...... 99 Productivity and Oil Estimates ...... 101 Fruit morphology and yields ...... 101 Oil production estimates ...... 102 Discussion ...... 104 Pasture Palms Out-Produced Forest Palms ...... 104 Attalea Species Comparison ...... 107 Variation, Predictability and Uncertainty ...... 107 What can Attalea Contribute to Energy Portfolios in Rural Amazonia? ...... 109 Harvest Considerations: Advantages and Disadvantages ...... 111 Local-based consumption versus market sale ...... 113 Conclusion ...... 114

6 CONCLUSION ...... 115

LIST OF REFERENCES ...... 118

BIOGRAPHICAL SKETCH ...... 140

LIST OF TABLES Table page

2-1 Number of yellow and purple-shaded staminate inflorescences of Attalea phalerata observed in anthesis in pasture and old-growth forest during monthly visits...... 25

3-1 Comparison of Attalea phalerata and A. speciosa mean (±1 SE) leaf, inflorescence and infructescence production in old-growth forest and pasture in Acre and Rondônia, Brazil...... 46

3-2 Number of inflorescences (staminate and pistillate) followed through time until fruit maturity, and percentage of pistillate inflorescence of total inflorescence for Attalea phalerata and A. speciosa ...... 47

4-1 Attalea phalerata plot densities and mean treatment densities (± 1 SD) measured in six 50 x 20m control (CTL) and treatment (SLB) plots...... 67

4-2 Densities per hectare (ha) and percent (%) of original A. phalerata palms that died within five weeks of forest clearing and burning (T1) in each control plot (CTL-1, CTL-2, CTL-3) and treatment plot (SLB-4, SLB-5, SLB-6)...... 69

4-3 Number of newly germinated A. phalerata seedlings ha-1 in each control (n=3) and treatment plot (n=3) at 2 (T1), 5 (T2), 9 (T3), and 14 months (T4) after forest clearing and burning...... 70

4-4 Proportion of original A. phalerata individuals surviving per study plot 14 months after slash-and-burn (T5), separated by size class...... 71

4-5 Repeated-measures analysis of variance with fixed effects for density by time and size class using PROC MIXED procedure in SAS ...... 72

4-6 Repeated-measures analysis of variance with fixed effects for relative abundance by time and size class using PROC MIXED procedure in SAS...... 72

5-1 Mean (± SE) infructescence and fruit characteristics based on seven Attalea phalerata and ten Attalea speciosa fruit bunches and fruit sub-samples...... 102

5-2 Estimated Attalea phalerata and Attalea speciosa oil production per infructescence and annually per palm in old-growth forests and pastures in Southwestern Amazonia...... 103

LIST OF FIGURES Figure page

2-1 Color variation from yellow to dark purple in Attalea phalerata staminate inflorescences observed in eastern Acre, Brazil...... 19

2-2 Proportion of Attalea phalerata palms with staminate inflorescence and monthly rainfall (http://www.acrebioclima.pro.br/) from January 2006 until December 2007...... 22

2-3 Purple coloration of petal tips on flowers of an Attalea phalerata pistillate inflorescence in eastern Acre, Brazil...... 23

3-1 A) Attalea phalerata and B) Attalea speciosa at pasture study sites in Acre and Rondônia, Brazil, respectively...... 36

3-2 Monthly staminate (male) and pistillate (female) flowering intensities of Attalea phalerata in A) old-growth forests and B) pastures with monthly rainfall (mm) from January 2006 to December 2007 (INMET 2008)...... 41

3-3 Monthly staminate (male) and pistillate (female) flowering intensities of Attalea speciosa in A) old-growth forests and B) pastures with monthly rainfall (mm) from January 2006 to June 2007 (original data)...... 42

4-1 Size class distributions before slash-and-burn (T0) shown as mean proportion (±1 SD) for control (CTL) (n=3) and treatment (SLB) plots (n=3). ... 68

4-2 Mean relative abundance of Attalea phalerata size classes in 50 x 20 m plots at each census before (T0) and 2 (T1), 5 (T2), 9 (T3), and 14 months (T4) after forest clearing and burning...... 74

4-3 Size class distributions on November 1, 2007 (T4), one year after slash-and- burn disturbance shown as mean proportion (±1 SD) for control (n=3) and slash-and-burn plots (n=3)...... 75

4-4 Proportion of non-reproductive Attalea phalerata individuals that grew at least one new leaf per census interval in slash-and-burn (n=3) and control plots (n=3)...... 76

5-1 A) A. phalerata and B) A. speciosa infructescences from study sites in Acre and Rondonia, Brazil, respectively...... 94

5-2 Comparison of mean female inflorescence and mature infructescence production per palm ±1 SE in old-growth forest and pasture...... 99

5-3 Monthly fruiting intensity of Attalea phalerata (with mature fruit) and monthly rainfall from January 2006 to December 2007 (INMET 2008) ...... 100

5-4 Monthly fruiting intensity of Attalea speciosa (with mature fruit) and monthly rainfall from January 2006 to June 2007 ...... 100

5-5 Estimated 2006 oil yields (L) per 12 palms from A) A. phalerata endosperm (kernel) and mesocarp (pulp) and B) A. speciosa endosperm at each study site in Acre and Rondônia, Brazil, respectively...... 104

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

ECOLOGY OF NATIVE OIL-PRODUCING PALMS AND THEIR POTENTIAL FOR BIOFUEL PRODUCTION IN SOUTHWESTERN AMAZONIA

Joanna Marie Tucker Lima

May 2010

Chair: Karen A. Kainer Major: Interdisciplinary Ecology

In the midst of a worldwide explosion of interest in biofuels, concern has arisen over displaced food crops, greenhouse gas emissions, forest conversion and biodiversity loss. These issues notwithstanding, biofuels are promising as a local fuel source in rural areas of Amazonia, where communities depend on fossil fuels for transportation and electricity generation. This study evaluates the potential of two native Attalea palm species in southwestern Amazonia as a source of oil for biofuel, and generates ecological knowledge to facilitate sustainable management of these species for oil. Attalea palms are common to forests as well as pastures in Amazonia, and their fruits contain copious amounts of oil. I compared flowering and fruiting phenologies of

Attalea phalerata Mart. ex Spreng. and Attalea speciosa Mart. ex Spreng. (Arecaceae) in actively grazed pastures and old-growth forests of Acre and Rondônia, Brazil, over 24 months, and combined these observations with data on palm densities, fruit characteristics and oil contents to assess biofuel potential. To better understand the persistence of A. phalerata palms in disturbed landscapes, we also studied the recovery

of A. phalerata palms after slash-and-burn agriculture using a before-after-control-

impact (BACI) deforestation experiment.

We found greater inflorescence and infructescence production in pastures than in

forests, and generally male-biased inflorescence sex ratios among individuals of both

Attalea species. At the population level, pasture palms demonstrated elevated

reproductive activity, bearing flowers and mature fruits year-round, while reproductive

phenophases of forest palms were more cyclic, with isolated flowering peaks and

consecutive months of inactivity. Light availability was positively correlated with

inflorescence production, and specifically with pistillate inflorescence, although less so

in A. phalerata than in A. speciosa. The size of the leaf crown (i.e., number of leaves)

was positively correlated with reproductive activity for both species in pastures, but not

in forests. Total palm height had no apparent effect on inflorescence production.

Results suggested that female function is more sensitive than male function to

environmental stress and resource limitations. Increased access to light and larger

crown size shifted Attalea sex expression towards increased female function and greater productivity primarily during the early stages of bud initiation and development through the regulation of sex determination and bud abortion.

Just 14 months after deforestation and burning, our analyses showed no overall effect of slash-and-burn on A. phalerata palm population densities or size class relative abundances. Pre-adult palms persisted in the slash-and-burn area through resprouting of new leaves, but there was relatively little germination. Protection of the apical meristem below-ground and the capacity to continue leaf production following damage characterizes resprouting in palms were critical to Attalea survival. Within slash-and-

burn plots, we detected accelerated growth rates in younger palms. Leaf production by surviving A. phalerata seedlings, post-seedlings, and pre-juveniles was so successful in slash-and-burn plots that 73% returned to or surpassed their original number of leaves

14 months after disturbance. Our results confirm that A. phalerata populations quickly recover after slash-and-burn agriculture, and will likely constitute a principal component of the future regenerating forests in the Southwestern Amazon.

In isolated rural Amazonian communities, naturally occurring populations of oleaginous plants represent a largely underexploited energy resource and can serve as an important and immediate source of fuel for electrification and transportation (e.g., river boats, small trucks, or even small airplanes). In a single year, an average A.

phalerata palm produced more than three times as many infructescences in pasture

than in forest, and A. speciosa produced more than twice as many. A. phalerata palms

in old-growth forest produced an average 1.17 L palm-1 yr-1 of oil, and in pasture, 4.55 L

palm-1 yr-1, while A. speciosa growing in forest produced 0.93 L palm-1 yr-1 and in

pasture, 2.59 L palm-1 yr-1. Energy production from locally available plant resources

promises greater independence and improved livelihoods for remote communities, while

assigning greater value to standing forests and diversifying pasture systems already

under use.

CHAPTER 1 INTRODUCTION

Over the past decade, interest in alternative energies has soared and biofuels have quickly expanded into a global commodity (Koh and Ghazoul 2008; FAO 2008;

Tilman et al. 2009). National agendas for energy independence has pushed biofuels to the forefront of political policies, and proponents have lauded biofuels as a clean renewable energy source to substitute waning supplies of petroleum-based diesel and gasoline. At the same time, critics have called for a more careful examination of the total carbon energy balance of biofuel production, arguing that biofuels can actually increase emissions of global warming gases (Fargione et al. 2008; Searchinger et al.

2008), accelerate tropical rain forest destruction and biodiversity loss (Koh and Wilcove

2007; Groom et al. 2008), and inflate world food prices (Escobar et al. 2009; Rathmann et al. 2010).

Against the backdrop of a biofuel boom around the world, this work explores the ecology of two native Amazonian palm species (Attalea phalerata and A. speciosa) that are common across both forested and cleared landscapes in southwestern Brazilian

Amazon, and assesses their potential as sources of locally-derived biofuel. Brazil has risen to the forefront of liquid biofuel production, use, and technology—particularly ethanol and biodiesel (da Costa 2004; Brandon 2005; Morgan 2005; Pousa et al. 2007;

Garcez and Vianna 2009), and provides a useful case for exploring the potential of local and regional biofuel production. In addition to research investments to improve liquid biofuel crops like castor bean, jatropha, soybean, and African oil palm, the Brazilian government has funded basic research to identify native oil-producing plants in the

Amazon with potential for biodiesel production.

Attalea phalerata and A. speciosa occur abundantly in various habitats of southwestern Amazonia and contain copious amounts of oil in their fruits. To evaluate the potential of these species as sources of biofuel, understanding of key ecological aspects is fundamental. Such knowledge can also offer insights into management potential, and help understand the impacts of different extraction practices and intensities on oil yields. These data are largely unavailable for native oleaginous plants in tropical forest regions (but see Feil 1996; Moraes et al. 1996; Lima et al. 2007), where energy access looms as an unfulfilled dream for many rural settlements and isolated communities.

The potential of native Amazonian oleaginous plants, and in particular palms, has been recognized for many years, and has a long history of research and investigation

(Lee 1930; Hodge 1975; Balick 1979; Pesce 1985). Nonetheless, this alternative fuel source is largely underutilized and overlooked by local populations as well as governments. In Amazonia, not only can oils from native plants provide a local and reliable source of fuel, they can improve the livelihoods of those who can ill afford conventional fuels. In the rural areas of Amazonia electricity is rare, although some homes use diesel generators to provide intermittent electricity. Poor road access plagues many areas, and rural towns are often isolated from cultural and commercial centers. Biofuel production from native palm populations could circumvent high fuel transportation costs and supply locally available fuel for riverboats, trucks, and even small airplanes frequently used in very remote Amazonian regions. Biofuels extracted from native plants could bring energy self-sufficiency to remote areas and rural

settlements, as well as offer the benefits of a clean and sustainable energy source at a time when climate change and global warming threatens our planet.

African oil palm plantations are frequently cited as a strategy for bringing energy to remote areas and rural communities in tropical countries. Plantation-based production of biodiesel in remote rural areas faces various obstacles not only due to isolation, but also due to scarce economic resources for necessary chemical inputs, technology and infrastructure, and unpredictable agriculture extension assistance. Plantations also experience a lag time before fruit production begins, whereas native populations are already producing. Forest extraction and silvo-pastoral systems that draw upon existing populations of native palm species can provide additional income to local producers with low additional investment and promote the valuation of standing forests in the

Amazon. Political rhetoric in favor of forest conservation and forest-based development, particularly strong in Acre, Brazil, offers hope that the government will support the extraction of vegetable oil from native palm populations for biodiesel production.

This research is part of a collaborative effort with an on-going research program at the Federal University of Acre, Brazil (UFAC), where researchers are investigating the potential of various native plant populations for the production of biodiesel. The overall objective of my research was to increase understanding of the ecology of native oil- producing palms with potential for biofuel production as a local energy source. To assess biofuel potential, I studied the intricacies of flowering and fruiting phenology of

Attalea phalerata and A. speciosa over two years in pasture and forest—two habitats where the palms often occur abundantly—to evaluate productivity within each environment and better understand the factors affecting reproductive biology and

phenology of these palms (Chapters 2 and 3). I also examined the persistence of A.

phalerata through anthropogenic disturbances, and more specifically, slash-and-burn

agriculture, to understand the regenerative capacity of this species and its

predominance in disturbed landscapes (Chapter 4). Finally, I integrated data on fruiting

phenology, fruit biometry, and oil quantification to estimate oil production potential at the

local scale (Chapter 5). Chapters 2-5 were written as independent, stand-alone

manuscripts for publication, and together they contribute to the overall aim of this work.

CHAPTER 2 FLOWER COLOR VARIATION IN ATTALEA PHALERATA (ARECACEAE)

Over the past few years, I have been conducting a phenology study of the arborescent palm Attalea phalerata Mart. ex Spreng. in southwestern Amazonia.

Although botanical records have consistently reported yellow flowers on A. phalerata inflorescences, I observed multiple cases of non-yellow staminate flowers. Flower color varied from dark purple to violet, or a mixture of yellow-orange to magenta flowers within the same male inflorescence (Figure 2-1; hereafter I refer to non-yellow flowers as purple). This article reports on field observations of flower color polymorphism in A. phalerata and discusses possible explanations for this anomaly.

Figure 2-1. Color variation from yellow to dark purple in Attalea phalerata staminate inflorescences observed in eastern Acre, Brazil.

Figure 2-1. Color from yellow to dark purple in Attalea phalerata, continued.

Field Observations

Phenology

I monitored flowering phenology of Attalea phalerata Mart. ex Spreng. in Acre,

Brazil, between January 2006 and December 2007. Using binoculars, I observed

flowering from the ground at monthly intervals at six study sites (three actively grazed

pastures and three areas of old-growth tropical moist forest). At each site I observed 12

reproductive palms. Between July and December 2007, observations were reduced to

two sites per habitat. For each individual I recorded sex and reproductive phase of all

inflorescences – closed inflorescence buds (bracts), inflorescences in anthesis (open

flowers) and dried post-anthesis inflorescence structures. I also categorized crown

illumination on a scale of zero to five by counting the number of sides of the palm crown

directly exposed to sunlight (four lateral sides plus top) (Bechtold 2003). Within old-

growth forest, A. phalerata is mainly a lower canopy palm.

Attalea species alternate between pistillate, staminate, and sometimes hermaphroditic inflorescences on the same plant. During 24 months of observations of

72 A. phalerata palms, I registered only four instances of hermaphroditic inflorescences.

The remaining inflorescences were either exclusively pistillate or exclusively staminate.

The majority of A. phalerata staminate inflorescences initiated flowering at the beginning of the dry season between May and June, peaked in September at the end of the dry season, and dwindled during the wet season (Figure 2-2). Palms growing in old-growth forest were more likely to suspend inflorescence production for a short period each year between February and April, whereas pasture palms produced inflorescences continuously year-round. Still, pasture palms mimicked the overall seasonal patterns of flowering peaks and lulls in the forest.

2006 2007

Figure 2-2. Proportion of Attalea phalerata palms with staminate inflorescence and monthly rainfall (http://www.acrebioclima.pro.br/) from January 2006 until December 2007 in pastures and old-growth forests in Acre, Brazil.

Flower Color Variation

To my surprise, of 55 male inflorescences observed in anthesis, the majority

(55%) produced purple flowers rather than the familiar yellow flowers. During two years of monthly phenological observations, I recorded 19 purple staminate inflorescences in anthesis in pastures (on 15 different palms) and 11 in forests (on nine different palms).

Over the same two-year period, I observed 14 yellow staminate inflorescences in pastures (on 12 different palms) and 11 in forest (on ten different palms). Some A. phalerata individuals alternated between purple and yellow flowers, while a few palms

(n=5) repeatedly produced purple flowers. Of the 22 A. phalerata individuals with purple flowers, more than half (n=12) also produced the better known yellow flowers either before or after a purple flowering event, indicating phenotypic plasticity within individual palms. In one case, a single palm exhibited one yellow and one purple inflorescence simultaneously. Both yellow and purple inflorescences were observed in anthesis, and the colors remained constant as the flowers developed. I also observed purple

coloration at the tips of creamy yellow petals of A. phalerata pistillate flowers (Fig. 2-3),

but only three pistillate inflorescences were observed in anthesis during the entire study

period.

Figure 2-3. Purple coloration of petal tips on flowers of an Attalea phalerata pistillate inflorescence in eastern Acre, Brazil.

Staminate flower color variation occurred not only within and among individual palm trees, but also in both the wet (November to April) and dry seasons (May to

October), across different habitats (pasture and forest), and on a regional scale dispersed over 100 km2. Purple flowers appeared at various times throughout the year,

although mostly during the dry season, which corresponds to A. phalerata’s peak

flowering season (Figure 2-2, Table 2-1). Most flowering anthesis events occurred

between observation visits, and for these I was unable to determine flower color.

Color Polymorphism in Palms

Flower color polymorphism in plants is common in nature and appears within genera, within species, and even within isolated populations. Several examples from herbaceous and other short-lived plants, both wild and cultivated, exist in the research literature (Armbruster 2002). Larry Noblick (pers. comm.) detected flower color variation between yellow and magenta in Attalea palms near Corumba, Mato Grosso do Sul,

Brazil, in the Pantanal region, but to my knowledge no records of within-species color variation in palm inflorescences have been published.

Palms, such as the lipstick palm, Cyrtostachys renda, with its bright red crown shaft, Geonoma epetiolata with a reddish purple underside on young leaves (Blanco and Martén-Rodríguez 2007), and various palm genera with purple fruits (e.g., Euterpe,

Bactris, Butia, Coccothrinax), testify to widespread anthocyanin production within

Arecaceae. Anthocyanins (a flavonoid sub-group) are responsible for the orange, red, purple, and blue flower colors and occur in almost all vascular plants (Grotewold 2006).

Harborne et al. (1974) found other flavonoid pigments (glycosides) specifically in the flowers of ten different palm species, and a few studies have even identified other types of flavonoids in the vegetative structures of Attalea and other coccosoid palm species

(Williams et al. 1983; Williams et al. 1985). Still, the question remains as to what drives flower color variation in A. phalerata.

Table 2-1. Number of yellow and purple-shaded staminate inflorescences of Attalea phalerata observed in anthesis in pasture and old-growth forest during monthly visits in (a) 2006 when I observed 36 reproductive individuals in pastures and 36 in old-growth forests, and (b) 2007 when I observed 36 individuals in each habitat until June and 24 in each habitat from June to December. (a) 2006 J F M A M J J A S O N D TOTAL

Purple-shaded 0 0 0 0 1 3 0 3 2 0 2 2 13 (no. palms=11)

Yellow (no. palms=9) 1 0 0 0 0 1 1 1 0 2 2 1 9 PASTURE

Purple-shaded 0 0 0 0 0 1 0 2 2 0 1 1 7 (no. palms=6)

Yellow (no. palms=4) 0 0 0 0 0 1 0 1 0 1 1 1 4 FOREST OLD-GROWTH

TOTAL 1 0 0 0 1 5 1 7 4 3 6 5 33

Table 2-1. Continued (b) 2007 J F M A M J J A S O N D TOTAL

Purple-shaded 0 1 1 0 0 0 0 0 3 1 0 0 6 (no. palms=4)

Yellow (no. palms=5) 0 0 0 1 0 0 0 3 1 0 0 0 5 PASTURE

Purple-shaded 0 0 0 0 1 1 0 0 0 0 2 0 4 (no. palms=4)

Yellow (no. palms=7) 0 0 0 0 0 0 0 3 2 1 1 0 7 FOREST OLD-GROWTH

TOTAL 0 1 1 1 1 1 0 6 6 2 3 0 22

Possible Explanations for Color Polymorphism in Attalea phalerata Flowers

To try and uncover the reasons for flower color variation in A. phalerata, I used

Pearson’s Chi-squared Test to examine relationships between flower color and four variables: (1) habitat (old-growth forest versus pasture) (χ2=1.01 , d.f.=1, p=0.32), (2)

season (wet versus dry) (χ2=0.799 , d.f.=1, p=0.37), (3) year (2006 versus 2007)

(χ2=1.84 , d.f.=1, p=0.17), and (4) crown illumination (χ2=10.67, d.f.=5, p=0.06). Results

revealed no significant associations with flower color. Crown illumination, or light

availability, was marginally significant, but the absence of an association between

habitat and flower color precluded any strong linkage between crown illumination and flower color, since the two habitats we compared—pasture and forest—represent two

extremes in light availability.

A common explanation for color variation within and among species is pollinator

selective pressures (Hannan 1981). Pollinators respond to various floral signals—color,

shape, size, fragrance, temperature—and these preferences exert selective pressures

on the plant to optimize reproductive success (Levin and Brack 1995; Meléndez-

Ackerman and Campbell 1998). Studies of A. phalerata pollination are scarce, however

nitidulid beetles from the genus Mystrops are most likely the principal pollinators

(Moraes et al. 1996). Beetles respond to floral signals of increased temperature and

fragrance, rather than color, and Attalea flowers are known to mature quickly, heating

up before anthesis (Henderson 2002). I observed insects, apparently pollinators,

actively feeding on purple inflorescences, so purple flowers do not appear to deter

pollinators. Further research may help determine if purple staminate flowers negatively

affect fertilization, fruit set, and reproductive success in A. phalerata.

Associations between anthocyanins in vegetative organs and flowers and

environmental stresses may also help explain flower color variation in A. phalerata. In vegetative organs, plants manufacture anthocyanins to protect against environmental stresses, such as herbivory (Fineblum and Rausher 1997), photo-damage (Close and

Beadle 2003) and drought (Levin and Brack 1995). Plants also synthesize pigments in

response to extreme environmental conditions, such as cold temperatures (Stiles et al.

2007) and nitrogen deficiency (Bonguebartelsman and Phillips 1995). Finally,

Armbruster (2002) found linkages between anthocyanins in vegetative organs and their

presence in flowers. Selection pressures related to environmental heterogeneity and

stress tolerance may be responsible for plant anthocyanin production in general, helping

maintain flower color polymorphism within and among species (Warren and MacKenzie

2001).

Final Considerations

Until now, A. phalerata inflorescence color in southwestern Amazonia has only been reported as yellow or cream-colored (Evandro Ferreira, pers. comm.).

Anthocyanins present in the plant to various degrees likely account for the various

shades of purple flowers observed. The question remains as to what provokes the

differences in anthocyanins seen in these palms. Flower color is not genetically fixed in

A. phalerata, since yellow inflorescences often followed the production of purple

inflorescences on the same plant. If environmental stress is responsible for flower color

variation in A. phalerata, three sources of stress come to mind: (1) Intermittent cold

fronts pass through the region each year during the early dry season dropping

temperatures into the lower teens (⁰C) (cf. Stiles et al. 2007); (2) A severe drought in

2005 induced soil moisture stress and could have indirectly augmented susceptibility to herbivory or pathogen attack; and (3) Extensive fires during the 2005 drought killed a

large number of pollinators, and palms may have reacted with a different flower color to

attract alternative pollinators. A. phalerata is a broadly distributed species, found

throughout the southern and western periphery of the Amazon region, including Brazil,

Bolivia and Peru, as well as the planalto of Brazil, Bolivia and Paraguay (Henderson et al. 1995). More detailed studies of this species’ flowering phenology and variation in flower color over its geographical range are warranted to better understand what triggers deviations in A. phalerata flower color and how flower color variation affects the ecology of this species.

CHAPTER 3 DOES LANDSCAPE CHANGE ALTER REPRODUCTIVE PHENOLOGY, SEX EXPRESSION AND PRODUCTIVITY OF ATTALEA PALMS IN SOUTHWESTERN AMAZONIA?

Introduction

Throughout human history the immigration of settlers, expansion of agriculture, cattle pastures and mining, and growth of urban centers and transportation networks have resulted in the conversion of almost half of the world’s former forest cover to other land uses (Abramovitz and Mattoon 1999; Williams 2008). Forest conversion not only results in the fragmentation of once vast forested landscapes, but also leaves behind scattered trees and spatially disjointed plant populations. Remnant trees confront a new set of environmental conditions that potentially alter reproductive biology and have far-reaching consequences for reproductive success.

Despite a broad array of studies throughout different ecosystems that consider unique aspects of plant reproductive biology (i.e., pollination, fruit set, breeding systems, phenology, etc.; Freeman et al. 1981; Aizen and Feinsinger 1994; Fenner

1998; Korpelainen 1998; Consiglio and Bourne 2001; Aizen et al. 2002; Lobo et al.

2003; Delph and Wolf 2005; Ramirez 2005;Alonso et al. 2007; Barrett 2010), relatively little attention has been given to the phenology and reproductive success of wild remnant forest trees in converted, human-disturbed landscapes, especially in the tropics

(but see Aldrich and Hamrick 1998). The paucity of studies may partially stem from the expectation of their inevitable demise. Janzen (1986) described most remnant trees as

“the living dead”, with no productive future because of elevated mortality and poor recruitment in apparently hostile environments (Aldrich and Hamrick 1998; Manning

2009). Nonetheless, when landowners clear forests for agricultural use, they often

spare individual trees due to the difficulty of felling very large or dense-wooded trees, to

comply with legal stipulations prohibiting harvest of particular species (e.g., Brazil nut

(Bertholletia excelsa), or to garner benefits that scattered trees provide under the new

land-use system, such as shade, fodder, or building materials (Cajas-Giron and Sinclair

2001; Pulido and Caballero 2006). Furthermore, numerous studies emphasize the ecological value of remnant trees in converted landscapes. They improve soil quality

(Wick et al. 2000), enhance nutrient cycling (Harvey and Haber 1999), facilitate tree regeneration and forest recovery (Nepstad et al. 1991; Vieira et al. 1994; Elmqvist et al.

2001; Carriere et al. 2002; Guevara et al. 2004; Nadkarni and Haber 2009), export

seeds to neighboring forest fragments (Aldrich and Hamrick 1998), increase genetic

connectivity for plant populations (Cascante et al. 2002; Manning et la. 2006), and help

conserve of faunal biodiversity by providing habitat and connectivity for animals within

fragmented landscapes (Guevara and Laborde 1993, Holl 1998; Harvey and Haber

1999; Lumsden and Bennett 2005; Manning et al. 2006; Herrera and Garcia 2009). As

deforestation and land-use change advances in tropical forest regions (Achard et al.

2002), an understanding of how plants adjust their reproductive patterns to changing

landscapes is essential, and offers insights into the flexibility of plant phenology and

regulation of fruit production.

Trees surviving in converted or disturbed forest habitats tend to display elevated

reproductive activity (Anderson 1983; Aldrich and Hamrick 1998; Lamont et al. 2003;

Kelly et al. 2007), but forest conversion can also have profound negative effects on

reproduction. When forests are converted to agriculture or pasture, changes in

fundamental conditions, such as water availability, sun exposure, soil moisture, soil

fertility, species composition, and resource competition (Aldrich and Hamrick 1998)

affect plant physiology and can also disrupt of pollination patterns and dispersal

systems (Cunningham 2000; Cascante et al. 2002; Levin 2009), potentially impairing

the reproductive success of individuals and populations of wild tree species left behind.

Wind-pollinated species would likely benefit (Anderson 1983), whereas insect pollinated

species may suffer from lower visitation rates depending on the pollinators’ habitat

requirements (Pardo 2001; Murren 2002). On the other hand, any trees left standing

potentially benefit from higher sunlight and reduced resource competition from other

tree species (Chazdon 1986; Wright and van Schaik 1994; Schroth et al. 2004).

In Amazonia and many parts of tropical America, palm trees represent a predominant feature in cleared forest landscapes due to their high economic and cultural values (Balick and Beck 1990). Palms have a reputation for quickly ruining chainsaw chains and dulling axes, making their felling costly. Land managers also frequently leave palms standing because they provide shade and food for cattle. In this

paper, we compare the flowering and fruiting phenology, sex expression, and fecundity

of two Amazonian palm species in their natural forested environment and as remnant

trees growing in pastures converted from forest 15-20 years earlier. We selected two

palm species, Attalea phalerata and A. speciosa, that regularly appear scattered across

deforested landscapes of contemporary Amazonia and demonstrate persistence long

after forest conversion. Because both species are common in both forest and pasture,

it provided us with the unique opportunity to study and compare patterns of reproductive

biology across drastically different environmental conditions and between two closely

related species.

Attalea palms are monoecious and exhibit fairly unique reproductive biology, where they alternate between unisexual—entirely pistillate or entirely staminate—and sometimes hermaphroditic inflorescences on the same plant (Henderson 2002). As early as the mid-1900s, botanists speculated about sex determination in monoecious plants, suggesting that sex alteration may be influenced by external environmental conditions (Schaffner 1921, 1925; Joshi 1939; Jones 1947). Since then researchers have identified numerous factors that drive sex allocation in plants, including environmental conditions such as temperature, soil moisture, sunlight, vapor pressure deficit (Charnov and Bull 1977; Freeman et al. 1980, 1981; Charnov 1982;

Bierzychudek 1984; Henson 2000), plant size (Bierzychudek 1984; Burd and Allen

1988; Condon and Gilbert 1988; Klinkhamer et al. 1997; Queenborough et al. 2007), past reproductive history (Bierzychudek 1984; Cunningham 1997), and genetics

(Anderson 1983; Werren and Beukeboom 1998).

Forest and pasture represent two contrasting environmental conditions, especially in regards to light availability, which several researchers identify as a key factor influencing palm fecundity (Piñero and Sarukhan 1982; Anderson 1983; Chazdon

1986). In plants capable of altering their gender, increased plant size and/or light availability often provokes a change in sex allocation towards more female flowers and greater fruit yields (Gregg 1973; Charnov and Bull 1977; Freeman et al. 1980; Bullock and Bawa 1981; Lovett Doust and Cavers 1982; Bierzychudek 1984; Wright and van

Schaik 1994). In Attalea, researchers have reported an exceptionally high proportion of staminate inflorescences during the first years of reproduction and that, as palms age, the ratio of staminate to pistillate inflorescences gradually declines (Voeks 1988;

Anderson et al. 1991; Feil 1996; Moraes et al. 1996). In his study of Attalea funifera,

Voeks (1988) ascribed this shift to greater access to light as the palms grew taller.

In this study we address the following questions: (1) How do reproductive phenology, and flower and fruit production in A. phalerata and A. speciosa differ between forest and pasture? and (2) What effects do light availability and plant size have on flower and fruit production and sex allocation? We expected greater reproductive activity in pastures and stronger male-biased inflorescence ratios in forests. We also hypothesized that increased access to sunlight would result in greater inflorescence and infructescence production. Because increased plant height in forest is typically positively correlated with increased access to sunlight, and crown size is directly related to photosynthesis potential, we expected Attalea inflorescence and infructescence production to increase with increasing height and crown size. Both increased access to sunlight and larger plant size should also be related to increased numbers and proportions of pistillate inflorescences produced by individual palms. To test these predictions we monitored the reproductive effort and flowering and fruiting phenology of A. phalerata and A. speciosa over 24 and 18 months, respectively, in old- growth forests and pastures of the southwestern Brazilian Amazon, and examined associations of flowering and fruiting with sunlight exposure and plant size.

Materials and Methods

Study Area

The bi-state study area stretches from the western tip of Rondônia (09⁰47′ S

066⁰24′ W) to eastern Acre, Brazil (10⁰00′ S 067⁰58′ W). Across a mostly flat landscape, forest intermingles with cleared and managed lands, creating a mosaic of pastures, agricultural lands, and abandoned fields in various stages of succession, as

well as old-growth tropical rainforest, classified as humid, moist tropical forest (Holdrige

1978). In this region, extensive pastures predominate along paved roads.

The rainy season typically extends from October to April with a pronounced dry season between May and September, when monthly precipitation generally remains below 100 mm (www.acrebioclima.pro.br). Average annual rainfall is between 1600 and

2000 mm, and the mean annual temperature is 25⁰C (ZEE 2002). Each year cold fronts pass through the region during the early dry season, bringing bursts of cooler temperatures in the lower teens (⁰C).

Study Species

Attalea phalerata Spreng. ex Mart. (formerly Scheelea phalerata (Mart. ex Spreng)

Burret) is an arborescent, lower canopy palm with a thick stem (25 to 40 cm in diameter) found in forests throughout the southern and western periphery of the Amazon region, including Brazil, Bolivia and Peru, as well as the drier forests and savannas of the planalto in Brazil, Bolivia and Paraguay (Henderson et al. 1995). In old-growth forest, mature palms reach up to 20 m tall, at which height they are typically still beneath the upper forest canopy. Within its natural range, this species persists in pastures and areas of cleared forest throughout eastern Acre and northern Bolivia (Figure 3-1A) and tends to dominate early regrowth forests (Henderson 1995, Carvalho et al. in press). A. phalerata is principally pollinated by nitidulid weevils from the genus Mystrops (Moraes et al. 1996; Henderson et al. 2000), but also by wind. Fava (2010) also observed the following flower visitors to A. phalerata: stingless bees (Trigona spinipes), small flies

(Drosophila sp.) and curculionid beetles.

Attalea speciosa Spreng. ex Mart. (formerly Orbignya phalerata) is an arborescent upper canopy palm with a thick stem (25 to 41 cm in diameter) occurring in the

Amazonian regions of Guyana, Suriname, Brazil, and Bolivia (Henderson et al. 1995).

When mature, palms reach heights of up to 30 m. It is particularly abundant in transition zones between tropical forest and other vegetation types. In addition to occurring at high densities in old-growth forests, this species flourishes in cleared forest areas, such as pastures or abandoned agricultural fields (Figure 3-1B), often forming dense monodominant stands (Anderson 1983; Henderson et al. 1995). Pollination of this species is also by nitidulid weevils (Mystrops sp.) and wind (Anderson et al 1988).

Both species are monoecious, bearing unisexual flowers on predominantly unisexual inflorescences. A. speciosa tends to produce more mixed-sex, or hermaphroditic inflorescences than A. phalerata, but often these inflorescences are functionally unisexual, where either the pollen produced by staminate flowers is sterile, or the pistillate flowers do not set fruit (Anderson 1988).

A B

Figure 3-1. A) Attalea phalerata and B) Attalea speciosa at pasture study sites in Acre and Rondônia, Brazil, respectively.

Site Selection

For each species, we selected six study sites (three actively grazed pastures and

three areas of well preserved old-growth tropical moist forest, loosely paired) based on

the occurrence of our study species and all-season road accessibility. We selected study sites after random farm visits along principal roadways within 300 km of Rio

Branco, Acre, Brazil (A. phalerata), and along a 35 km stretch of Federal Highway, BR-

364 in Extrema, Rondonia, Brazil (A. speciosa). All forest-pasture pairs were within 2

km of each other; A. phalerata and A. speciosa never co-occurred within study sites. All

sampled pastures were cleared from forest 15-30 years ago, and landowners confirmed

that no chemical fertilizers or herbicides were used on any of the pastures.

Data Collection

Reproductive phenology

In January 2006, we identified 12 reproductive palms to represent a sub-

population at each old-growth forest and active pasture site (n=72 for each species). In

pastures, all individuals stood roughly within a 1 ha area. In old-growth forest, where

palms typically grow farther apart, monitored individuals were distributed throughout 2-5

ha. All selected palms showed evidence of reproduction and their leaf crowns were fully

visible.

At monthly intervals, we observed flowering and fruiting from the ground using

binoculars. For each individual palm we recorded sex and reproductive phase of all

inflorescences—closed inflorescence buds, inflorescences in anthesis (open flowers)

and post-anthesis inflorescences. In the case of infructescences, we monitored

infructescence maturation and recorded when fruits ripened and fell to the ground. For

each reproductive event, we recorded the azimuth of each inflorescence with a

compass to follow it continuously through flowering and fruiting phenophases (cf.

Anderson et al. 1988). We conducted monthly phenological observations of A. phalerata from January 2006 to July 2007, except for two pasture and two forest sites that we continued to monitor through December 2007. We extended A. phalerata observations to capture a second flowering season, since inflorescence frequency was unusually low during the initial observation period. We observed A. speciosa phenology at monthly intervals between January 2006 and June 2007, incorporating two flowering peaks.

Size and illumination measurements

During our initial January 2006 phenology observation visits, we counted the total number of leaves and measured each individual’s total height and stem height (from the ground to the base of the oldest green leaf), using a tangent height gauge. To assess leaf production rates, in February 2006, we marked the youngest fully expanded leaf on a sub-sample of individuals from each species with red paint. One year later we counted the number of newly expanded leaves. During monthly visits we also visually estimated crown illumination by counting the number of sides (0-5; four sides plus top) exposed to direct sunlight, imagining the tree crown as a box (Bechtold 2003).

Palm dissections

To improve our understanding of inflorescence sex expression, we dissected the crowns of four randomly selected adult A. phalerata palms growing in pasture (n=2) and forest (n=2). Palms were felled and divided into trunk and leaf crown sections using a chainsaw. We then removed each leaf one by one, starting with the oldest, extracting inflorescence buds encountered within every leaf base. Buds were numbered to keep track of age and order, and we noted bud absence within any given leaf base. We slit

open each developing unit with a knife to examine inflorescence structure and identify

sex.

Statistical Analyses

Flowering and fruiting frequencies were calculated as the number of events

registered annually (cf. Newstrom et al. 1994). Flowering and fruiting intensity refers to

the proportion of individual palms that flowered over a set time period (cf. Herrerias-

Diego et al. 2006). For statistical analyses, we combined hermaphroditic with pistillate

inflorescences due to their active female function (i.e., fruit production). We used SAS software version 9.1.3 for Windows (SAS Institute, Inc. 2002-2003) for all data analyses.

To compare flowering and fruiting frequencies between forest and pasture, we used permutation tests with 10,000 repetitions and a significance level of p<0.05. For both Attalea species, we tested for differences in pistillate and staminate flowering frequencies and fruiting frequency. We ran an asymptotic nonparametric test for multivariate balanced single-factor designs (MANOVA) to test for effects of habitat type

(forest vs. pasture) on the proportion of staminate and pistillate inflorescences per palm simultaneously (Bathke et al. 2008). Pistillate and staminate inflorescence production was negatively correlated in all cases except for A. phalerata in 2006, which showed a tendency toward positive correlation. Following recommendations from Harrar and

Bathke (2008), we used the Lawley-Hotelling-type multivariate test statistic for negatively correlated variables and the ANOVA-type multivariate test statistic for positively correlated variables.

To investigate associations between individual palm flowering and fruiting frequencies and crown illumination or palm physical characteristics, we computed correlations for the following variables: annual number of inflorescences (male and

female); annual number of ripe infructescences; total height; number of leaves; annual

leaf production rate and light availability (i.e., crown illumination). Since distributions of most variables were non-normal, we used Spearman’s correlation coefficients. We also computed habitat-specific Spearman correlations.

Results

Comparison of Flower and Fruit Production in Forest and Pasture

Flowering and fruiting phenology

Flowering of each Attalea species throughout the year varied in magnitude between the two habitats but followed similar seasonal patterns in terms of peak and low reproduction periods (Figures 3-2 and 3-3). Forest palms were more likely to suspend inflorescence production for at least a short period each year, whereas pasture palms produced inflorescences continuously year-round (Figures 3-2 and 3-3).

At old-growth forest sites, A. phalerata initiated flowering at the beginning of the dry season (May–June) and peaked around September, when monthly rainfall fell below

150 mm (Figure 3-2). Pistillate flowering was relatively rare among forest palms during

2006—only 21% of individuals produced female flowers. In contrast, 54% of monitored forest palms produced at least one pistillate inflorescence in 2007. Forest palms largely suspended pistillate flower production during the rainy season, while in pastures we observed pistillate inflorescence in all months but March 2006 (Figure 3-2). Staminate inflorescence phenological patterns closely mimicked those of pistillate flowering in

2006, whereas in 2007, staminate inflorescence production peaked two months later

(September–October) than pistillate inflorescence (July).

2006 2007 B

2006 2007

Figure 3-2. Monthly staminate (male) and pistillate (female) flowering intensities of Attalea phalerata in A) old-growth forests and B) pastures with monthly rainfall (mm) from January 2006 to December 2007 (INMET 2008) in eastern Acre, Brazil.

In contrast to A. phalerata, A. speciosa flowering tended to parallel rainfall patterns, increasing flowering intensity during the wet season (cf. Anderson et al. 1991;

Figure 3-3). We observed no pistillate inflorescences in forest towards the end of the

2006 dry season, while pistillate flowers were present in pastures throughout the year

(Figure 3-3). Staminate inflorescences were produced year-round in both habitats.

Male and female flowering patterns generally overlapped between habitats, although

staminate inflorescence production consistently peaked two months earlier in pasture

than in forest.

2006 2007 B

2006 2007

Figure 3-3. Monthly staminate (male) and pistillate (female) flowering intensities of Attalea speciosa in A) old-growth forests and B) pastures with monthly rainfall (mm) from January 2006 to June 2007 (original data) in western Rondônia, Brazil.

The majority of A. phalerata fruits began developing during the dry season (July–

August) as pistillate inflorescences completed anthesis. Infructescences matured over an average period of 6 months (± 0.17 SE, n=6) in forest and 7 months (± 0.08 SE, n=75) in pasture, and the greatest availability of mature fruit coincided with the wettest months of the year (January–March). In old-growth forest, presence of mature fruits was restricted to 5–6 months per year, while in pasture, we encountered ripe fruits year-

round. Successful fruit set was very high, regardless of habitat, and over the entire

study only 8% (n=2) and 3% (n=3) of observed pistillate inflorescences entirely aborted

prior to fruit ripening in forest and pasture, respectively.

Most A. speciosa infructescences initiated during the wet season and, on average, matured over an eight month period in both forest (± 0.08 SE, n=23) and pasture (± 0.39

SE, n=51), such that mature fruits were available mostly during the late dry and early wet seasons. In forest, infructescences with ripe fruit appeared during most months of the year except February–May, while in pastures we consistently observed mature fruits year-round. Among all A. speciosa pistillate inflorescences we monitored over 18 months, merely 6% (n=3) and 1% (n=1) aborted at some time between anthesis and fruit maturation in forest and pasture, respectively. In both Attalea species, infructescences that aborted almost always flowered outside of the peak flowering season.

Flowering and fruiting frequency

As new leaves form, palms produce an inflorescence bud in each leaf axil, so inflorescence frequency is naturally constrained by the production of new leaves.

Annually, A. phalerata produced slightly more leaves in pasture ( =8.7 ± 1.0 SD) than in

forest ( =7.9 ± 1.1 SD, P<0.05), while A. speciosa, showed no differences in annual leaf

production between the two habitats (pasture: =6.1 ± 1.0 SD; forest: =5.8 ± 1.2 SD,

P=0.3777; Table 3-1). Abortion of developing inflorescence buds regularly interrupted

reproduction, creating a discrepancy between annual leaf production and annual

inflorescence production. For both species, total annual inflorescence production (male

plus female) per palm was consistently greater in pasture than forest (P<0.05), and

palms consistently produced more pistillate inflorescences per individual in pasture than

in old-growth forest (P<0.001; Table 3-1). Similarly, A. phalerata and A. speciosa produced more staminate inflorescences per palm in pasture than in forest in 2006

(P<0.001 and P<0.05, respectively); however, differences in staminate inflorescence production between habitats were nonexistent in 2007 (Table 3-1).

Over the entire observation period, both A. phalerata and A. speciosa consistently

produced more staminate than pistillate inflorescences, regardless of habitat (Table 3-

2). Indeed, more than one-third of monitored palms produced exclusively staminate

inflorescences during the study (A. phalerata: n=15; A. speciosa: n=15), while relatively

few palms produced strictly pistillate inflorescences (A. phalerata: n=6; A. speciosa:

n=1). Otherwise, average annual staminate-to-pistillate inflorescence ratios per palm in

forest and pasture were 5.0:1.0 and 1.4:1.0, respectively, for A. phalerata; and 2.5:1.0

and 1.3:1.0, respectively, for A. speciosa. Rank-based multivariate analysis of variance

repeatedly showed differences between forest and pasture in the proportion of pistillate

and staminate inflorescences per palm for each species [A. phalerata: 2006, P≤0.0001

(Dempster-ANOVA-type); 2007, P=0.0009 (Lawley-Hotelling); A. speciosa: over 18-

months, P=0.0001 (Lawley-Hotelling)].

In A. phalerata, hermaphroditic inflorescences were rare and occurred at only one

pasture site. A. speciosa, on the other hand, produced hermaphroditic inflorescences in

both habitats and across study sites. We encountered only one A. phalerata palm with

an hermaphroditic inflorescence and eleven A. speciosa palms that repeatedly

produced hermaphroditic inflorescences. Hermaphroditic inflorescences varied in their

spatial distribution of male and female flowers. In A. phalerata, inflorescences tended to

display a basal congregation of female flowers with relatively fewer distal male flowers,

while in A. speciosa, the two sexes were usually mixed along the primary inflorescence rachis.

Annually, both Attalea species produced more infructescences in pasture than in forest in 2006 (P<0.001, n=36 for each species) and 2007 (P<0.005; A. phalerata: n=24,

A. speciosa: n=36) (Table 3-1). Fruiting phenology closely reflected pistillate inflorescence patterns, since abortion of post-anthesis infructescences was rare. At any one time, individual palms displayed multiple infructescences in different stages of development. The number of mature infructescences produced in a single year by an

A. phalerata palm ranged between 1 and 11 in pasture (mode=1), while in old-growth forest, only one palm produced as many as two infructescences per annum. Otherwise, most A. phalerata palms in forest produced no more than one infructescence per year

(mode=0). The maximum number of infructescences produced annually by a single A. speciosa palm in pasture was five (mode=1), and in old-growth forest, individual palms produced as many as two infructescences per year (mode=0).

Table 3-1. Comparison of Attalea phalerata and A. speciosa mean (±1 SE) leaf, inflorescence and infructescence production in old-growth forest and pasture in Acre and Rondônia, Brazil. P-values are based on randomization tests for differences between forest and pasture (10,000 iterations). Old-growth Forest Pasture

N mean (±1 SE) N mean (±1 SE) P-value Attalea phalerata Annual leaf 24 7.9 (1.12) 19 8.6 (0.95) 0.0413 production ratea No. of staminate inflorescences per palm 2006 36 1.3 (0.03) 36 3.0 (0.35) 0.001 2007 24 2.8 (0.00) 24 2.6 (0.30) 0.8202 No. of pistillate inflorescences per palm 2006 36 0.3 (0.08) 36 1.7 (0.15) <0.0001 2007 24 0.6 (0.00) 24 2.6 (0.30) 0.0003 No. of mature infructescences per palm 2006 36 0.5 (0.06) 36 1.7 (0.17) <0.0001 2007 24 0.2 (0.10) 24 2.3 (0.50) <0.0001 Attalea speciosa Annual leaf 21 5.8 (1.17) 16 6.1 (0.96) 0.3777 production ratea No. of staminate inflorescences per palm 2006 36 2.1 (0.28) 36 2.9 (0.46) 0.0327 2007 (6-mos.) 36 2.1 (0.27) 36 1.9 (0.47) 0.6497 No. of pistillate inflorescences per palm 2006 36 0.8 (0.12) 36 1.8 (0.57) 0.0004 2007 (6-mos.) 36 0.7 (0.06) 36 1.3 (0.29) 0.0019 No. of mature infructescences per palm 2006 36 0.7 (0.06) 36 1.6 (0.43) <0.0001 2007 (6-mos.)b 36 0.2 (0.10) 36 0.7 (0.28) 0.0038 a Annual leaf production rates are represented by means (±SD). b Peak production of mature fruits did not occur during the observation period, explaining the low infructescence values recorded during 2007.

Table 3-2. Number of inflorescences (staminate and pistillate) followed through time until fruit maturity, and percentage of pistillate inflorescence of total inflorescence for Attalea phalerata and A. speciosa in Acre and Rondônia, Brazil.

Old-growth forest Pasture Staminate Pistillate Staminate Pistillate Attalea phalerata 117 26 (18%) 178 125 (41%) (Jan 2006-Dec 2007) Attalea speciosa 153 52 (25%) 171 113 (40%) (Jan 2006-Jun 2007)

Associations between Reproduction, Light Availability, and Tree Size

For A. phalerata, habitat was correlated with total height (rs=0.59), number of

leaves (rs=0.25), light availability (rs=0.86), and annual leaf production rates (rs=0.32; all

P<0.05). Similarly, habitat was correlated with A. speciosa total height (rs=0.83),

number of leaves (rs=0.42), and light availability (rs=0.75; all P<0.001), but not with

annual leaf production rate. These relationships were confounded due to

multicollinearity between inflorescence production and plant size (i.e., height and number of leaves), and light availability. Forest palms had fewer leaves (t-test, A. phalerata: P=0.03; A. speciosa: P=0.0002) and were consistently taller than pasture palms (t-test, all P<0.001).

To avoid collinearity, we divided the data into forest and pasture subsets. Over 24 months, A. phalerata pistillate inflorescence frequency in forest was positively correlated with illumination (rs=0.37, n=24; P=0.0769), while pistillate inflorescence in pasture was

positively correlated with crown size (leaf number) (rs=0.41, n=24; P=0.0477).

Staminate inflorescence production showed no correlations with light or crown size.

Infructescence production showed marginal positive correlations in forest with leaf

production rates (rs=0.41, n=19; P=0.0832), but neither with light nor crown size.

For A. speciosa, pistillate inflorescence production over 18 months in forest was

also positively correlated with illumination (rs=0.39, n=36; P=0.0182), and pistillate

inflorescence frequency in pastures showed positive correlations with crown size

(rs=0.31, n=36; P=0.0639). Staminate inflorescence production in forest was positively

correlated with both illumination (rs=0.38, n=36; P=0.0242) and crown size (rs =0.37, n=36; P=0.0282), but in pasture showed no correlations. Infructescence production in forest was positively correlated with illumination (rs=0.42, n=36; P=0.0098), while in

pastures, it was positively correlated with crown size (rs=0.34, n=36; P=0.0445) and

negatively correlated with total height (rs=0.38, n=36; P=0.0214).

Palm Dissections

We dissected and sexed discernable inflorescences at different stages of

development along a sequence of up to 23 leaves for each of four dissected A. phalerata palms, ranging from fruiting branches associated with the oldest leaf to tiny inflorescences contained within an 11-cm-long prophyll nestled in the base of the spear leaf. The majority of dissected inflorescences was still enclosed within the peduncular bract and had not yet emerged from the prophyll. We were able to determine the sex of inflorescences as small as 5 cm in length. Including all recognizable inflorescence buds, bracts, and inflorescences, the ratio of staminate to pistillate inflorescences was heavily biased toward male function in both forest and pasture palms. In pasture, we recorded inflorescence sex ratios (male:female) of 3:1 (N=16 buds) and 10:1 (N=22 buds), and in old-growth forest, ratios of 3.3:1 (N=13 buds) and 3.7:1 (N=14 buds). The number of leaf bases with absent or aborted inflorescence buds were 5 of 17 and 2 of

23 in pasture, and 4 of 17 and 7 of 19 in forest.

Discussion

Pasture Palms Out-Produced Forest Palms

As expected, pasture palms overwhelmingly produced more inflorescences than

forest palms, specifically producing more pistillate inflorescences per annum than palms

growing in forest (Table 3-1). In a single year, the average A. phalerata palm produced

more than four times as many pistillate inflorescences in pasture than in forest, and A.

speciosa produced more than twice as many. Similarly, in the eastern Amazon,

Anderson et al. (1991) reported that A. speciosa resource allocation towards

reproductive structures increased almost 100% in pasture compared with forest.

Staminate flowering frequency varied less between the two habitats than pistillate

flowering with no significant difference in staminate inflorescence frequency between

habitats during the second year of observations. Since post-anthesis infructescence

abortion was rare (≤ 8%), substantial differences in pistillate inflorescence between pasture and forest palms translated into similar differences in fruit production.

Phenology

Similar timing of flowering and fruiting peaks between pasture and forest in each

Attalea species was expected, given that the principal cues that trigger phenological patterns in the tropics – photoperiod, irradiance, and rainfall (Rathcke and Lacey 1985;

Wright and van Schaik 1994; Borchert et al. 2004; Stevenson et al. 2008) varied little, if at all, between habitats. A. speciosa flowering peaked two months earlier in pasture populations than in forest populations, pointing to rainfall as a proximate cue for flower anthesis. Flowering in this species peaks during the wet season and the quicker saturation of exposed pasture soils than forest soils may have triggered earlier flowering in pastures (Figure 3-3).

The contrast between continuous flowering and fruiting by pasture palms and distinct seasonal phenophases in forest populations highlighted an important difference between reproduction patterns in the two habitats. Extended duration of population flowering and fruiting in pastures points to a lack of seasonal differences in resource and/or pollinator availability (Rathcke and Lacey 1985). Phenology studies often link pollen limitation with fruit set to explain variation in fecundity (Burd 1994; Cunningham

1995, 1996; Ratsirarson and Silander 1996; Marten and Quesada 2001; Herrerías-

Diego et al. 2006), but in our study species, pollen availability is not likely to be responsible for greater productivity in pastures, because inflorescence fruit set was overwhelmingly successful in both habitats: forest 92% and pasture 98%. Instead,

Attalea fecundity was primarily regulated prior to anthesis, though inflorescence sex determination and the abortion of developing inflorescence buds, which is common among palms (Tomlinson 1990). On average Attalea palms produced 6-8 new leaves annually, but only 2-3 inflorescences in forest and 4-5 in pasture developed through to anthesis. The greater inflorescence-to-leaf ratio in pastures than in forest corroborates higher abortion rates in forest and suggests that pastures better approximate ideal conditions for Attalea reproduction (also see Anderson et al. 1991). In Elaeis guineensis, Hartley (1977) reported that the principal causes of high bud abortion rates include defoliation, high levels of shading, and moisture stress. Shadier conditions and smaller crown sizes in forests compared to pastures may contribute to greater rates of

Attalea bud abortion in the former.

Hidden Mechanisms Control Sex Expression

Linking sex expression and productivity to specific environmental conditions or resource availability is less than straight-forward in the case of Attalea. Environmental

stress can regulate reproductive success during different phases of reproduction: (1)

inflorescence sex differentiation, (2) bud development and (3) fruit ripening (Lloyd

1979). In this study, sex determination and bud abortion were the primary regulators of

sex expression and productivity in Attalea and operated along a continuum from bud

initiation to bud-opening and anthesis –a period of at least two years (Anderson 1983;

unpublished data). Fruit set was very high in both A. phalerata and A. speciosa, as was

successful fruit maturation. In our study, post-anthesis fruit abortion was rare (≤ 8%),

although it tended to be more common in forest than in pasture. In weevil-pollinated A.

phalerata, failed fruit set occurred more often when inflorescences flowered outside of

the peak flowering season, during the wettest months of the year (Scariot et al. 1991,

1995; DeSteven et al. 1987; Marquis 1998). Beetle-pollinators (Mystrops spp.) tend to avoid higher rainfall levels, and diminished pollinator populations at these times likely stifled successful fruit set (Nunez et al. 2005).

Even for the much-studied African oil palm (Elaeis guineensis), the exact moment of sex determination remains elusive (Jones 1997), making connections with biotic and abiotic factors a challenge. In our study species, sex determination occurred well before the expansion of the associated spear leaf and up to 3-4 years prior to flowering

(Anderson 1983; unpublished data). Sex differentiation in monoecious palms may depend on feed-back signals from developing bunches or from existing environment conditions (Jones 1997; Henson and Dolmat 2004). In maize, once inflorescence sex is determined, it remains stable up to anthesis (Irish and Nelson 1989), but Oboh and

Fakorede (1999) suggested that in Elaeis guineensis, unfavorable conditions during bud

development can precipitate a later sex reversal from female to male, or vice-versa,.

Even when factors influencing sex determination can be identified, abortion prior to bud- opening modifies the initial ratio of male to female inflorescences. Hartley (1977) presented evidence that female inflorescence buds in Elaeis guineensis are preferentially aborted under stress. If this is also true of Attalea, phenological observations may provide a biased depiction of initial sex differentiation patterns and mask the early operation of environmental conditions, tree size, or genetic factors on inflorescence sex expression. In our dissections, the sex of the aborted inflorescence buds was indeterminable, but even if all aborted buds had been female, male-biased inflorescence sex ratios in Attalea still persisted.

Sunlight and Plant Size Linked to Sex Expression and Productivity

Of the variables we measured, the principal bottleneck to Attalea reproductive success in forest appeared to be light. Inflorescence production was more tightly linked with crown illumination in A. speciosa than in A. phalerata, both in terms of a stronger positive correlation and its effect on reproduction irrespective of inflorescence sex. For

A. phalerata, only pistillate inflorescence frequency increased with increasing illumination. A. speciosa’s customary position in the upper forest canopy hints that this species may be more light-demanding than A. phalerata, which usually remains beneath the upper canopy.

Although we expected to see a positive relationship between inflorescence production and palm height, we found no significant correlations between these two variables, except for a negative correlation between A. speciosa infructescence production and total height in pastures (rs= -0.38). A possible explanation is that the taller A. speciosa palms in our study represent senescing individuals with diminishing reproductive output.

Open conditions in pasture provided favorable light conditions and reduced

competition for Attalea reproduction, presumably facilitating not only higher annual

inflorescence frequencies but also greater allocation to female function (Charnov and

Bull 1977; Voeks 1998). Our finding greater female reproduction in pasture than in

forest supports these conclusions. Once light requirements were largely saturated in

pastures, crown size became an important influence on inflorescence production, as a

way to augment photosynthesis—a larger crown supported more female inflorescences.

When size is associated with sex expression, males are usually smaller, or younger,

than females (Condon and Gilbert 1988; Voeks 1988; Korpelainen 1998; Cozza 2008).

For palms, crown size, or number of leaves, also represents the upper limit of

inflorescence production, as each leaf bears a single inflorescence bud. In A.

phalerata, the positive association between both total (rs=0.50; P=0.01) and pistillate

inflorescence (rs=0.41; P=0.05) and crown size in pastures suggests that photosynthetic

leaf area limits reproduction in pastures (Cozza 2008). For both species, we also found

a negative correlation between staminate and pistillate inflorescence production over 18

months in pastures, but not in forest (A. phalerata: rs= -0.43; P=0.04, and A. speciosa:

rs= -0.75; P<0.0001), illustrating a clear trade-off between male and female reproduction

under open conditions. The majority of pastures palms must have been close to their maximum reproductive output, in terms of the number of inflorescences (particularly A.

speciosa), since greater allocation to pistillate inflorescence meant fewer staminate

inflorescences, and vice-versa.

Sex Expression and Male Dominance

Individual A. phalerata and A. speciosa palms displayed gender plasticity both

between and within flowering seasons, switching back and forth between male and

female inflorescences throughout the study period. Sex expression variability in

monoecious species is considered adaptive to optimize resource allocation to male and

female functions depending on resource levels or external conditions (Lazaro and

Mendez 2007). Favorable environmental conditions tend to increase femaleness, while

environmental stress (e.g., less-than-optimal light, nutrients, water or weather conditions) often favors maleness (Charnov and Bull 1977; Freeman et al. 1981;

Korpelainen 1998). Male-biased inflorescence sex ratios predominate among dioecious and monoecious plant populations, (Solomon 1985; Newstrom et al. 1994; Meagher and

Delph 2001; Richardson and Clay 2001; Queenborough et al. 2007; Lazaro and

Mendez 2007) as well as within sex labile individuals, like Attalea spp (Anderson et al.

1991; Feil 1996; Nunez et al. 2005). Our results confirmed the predominance of the

male function in Attalea regardless of habitat, although the degree of male bias was

greater in forest than in pasture, again suggesting relatively greater resource limitation

in forest. Among all inflorescences observed in this study, 82% and 75% of

inflorescences in forests, and 59% and 60% in pastures, were male for A. phalerata and

A. speciosa, respectively. Differences in reproductive cost between male and female

function are often quite large, with females incurring substantially greater resource

investments to support fruit maturation (Freeman et al. 1976; Lloyd 1979; Charnov

1982; Queenborough et al. 2007). Indeed, the large oil-rich fruits of our two study species fit this model, requiring high levels of energy resources for fruit development

(Voeks 1987; Anderson 1983). In A. phalerata, the stronger associations of pistillate rather than staminate inflorescence with light and crown size seemed to reflect a

disparate influence on female versus male function, supporting the idea that female

function is relatively more sensitive to resource availability (Voeks 1987).

While the proportion of male versus female inflorescences can offer clues about relative resource limitations between pasture and forest, it is important to remember that even under presumably near-optimal growing conditions in pastures, male inflorescences predominate Attalea reproduction, so female:male inflorescence sex ratios less than 1.0 are not necessarily a sign of resource limitation.

Pollen Availability may Alter Operational Sex Ratios

Another possible explanation for the differences in pistillate and staminate inflorescence between forest and pasture is the population operational sex ratio (OSR), defined as the number of sexually active males in relation to the number of receptive females at a given time (Emlen and Oring 1977). Lopez and Dominguez (2003) found that plants can facultatively adjust patterns of sex allocation in response to variation in pollination intensity. Pollen shortage on stigmas of the monoecious perennial herb,

Begonia gracilis, resulted in a more male-skewed sex allocation, while under relatively pollen-rich conditions, a stronger female-skewed sex allocation pattern arose (Lopez and Dominguez 2003).

Palms growing underneath the main forest canopy, such as A. phalerata, likely rely almost exclusively on insect pollination (cf. Voeks 1987), but in open pastures, entomophily (insect pollination) and anemophily (wind pollination) combine to increase pollen availability (Anderson et al. 1988). Because wind pollination functions better in open pastures than in closed forests, Attalea would experience greater pollen availability, fostering conditions for a more female-biased sex allocation (i.e., increased proportion of pistillate inflorescence), according to Lopez and Dominguez’s (2003)

hypothesis. Moreover, anemophily could facilitate pollination and fruit development

year-round in pastures, helping to explain the observed continuous fruit production in

pastures. Although speculative, OSR could also elucidate the relatively smaller

difference between A. speciosa inflorescence production between forest and pasture, because contrary to A. phalerata, A. speciosa palm crowns emerge through the upper forest canopy where wind pollination occurs more easily (Burd and Allen 1988).

Conclusions

Our aim in this paper was to assess the consequences of forest clearing on the

reproductive biology of surviving populations of remnant A. phalerata and A. speciosa

palms by comparing their phenology and productivity in old-growth forests and actively

grazed pastures. Using these two habitats as a natural experiment, we also examined

the influence of light availability and plant size (total height and crown size) on Attalea

inflorescence and infructescence. In the case of Attalea, forest conversion to pasture

had a considerable and favorable effect on reproduction through greater overall

fecundity, flowering and fruiting phenology, and increased female sex expression.

As predicted, we found greater inflorescence and infructescence production in

pastures than in forests, and a generally male-biased inflorescence sex ratio among

individuals of both Attalea species. As a population, pasture palms demonstrated

elevated reproductive activity, bearing flowers and mature fruits year-round, while

population-level reproductive phenophases of forest palms were more cyclic, with

isolated flowering peaks and consecutive months of inactivity. Light availability was

positively correlated with inflorescence production, specifically of pistillate

inflorescences, confirming our hypothesis, although less so in A. phalerata than in A.

speciosa. The size of the leaf crown (i.e., number of leaves) was positively correlated

with reproductive activity for both species, as expected, but only in pastures, and total palm height had no effect on inflorescence production. Results suggested that female function is more sensitive than male function to environmental stress and resource limitations. Increased access to light and larger crown sizes shifted Attalea sex

expression towards increased female function and greater productivity primarily during

the early stages of bud initiation and development through the regulation of sex

determination and bud abortion.

CHAPTER 4 RECOVERY OF ATTALEA PHALERATA MART. EX SPRENG. PALM POPULATIONS AFTER SLASH-AND-BURN AGRICULTURE IN SOUTHWESTERN AMAZONIA

Introduction

After clearing the forest, many contemporary farmers and ranchers in the tropics use fire to prepare land for agriculture and pasture use (Nepstad et al. 1999, Cochrane

2003, Zarin et al. 2005). Burning temporarily increases soil fertility for cultivated crops,

reduces unwanted vegetation (weeds), and kills agricultural pests (Nye and Greenland

1960; Popenoe 1960; Moran 1980). Pasture management in the tropics also typically

entails regular burning every few years to reduce weedy growth and reinvigorate grasses (Kauffman et al. 1998), although newer no-burn management practices are beginning to catch on. In this paper we assess Attalea phalerata palms’ resistance to

forest clearing and fire, and investigate its mechanisms of palm populations’ recovery

after slash-and-burn agriculture.

Forest clearing and fire are in no way new to the Amazon (Sanford et al. 1985).

Human populations have occupied the Amazon for at least 11,300 years (Roosevelt et al. 1996), and indigenous Amazonians have used fire as a tool for managing vast landscapes for agriculture as well as for game management (Bush et al. 2000; Zarin et

al. 2005; Bush et al. 2007). Charcoal found in the subsoil underneath forested lands

provides unequivocal evidence of widespread fire use in the region (Sanford et al. 1985;

Bush et al. 2007), and mounting archaeological evidence points to complex organized

pre-Colombian societies that extended their influence over thousands of square

kilometers (Denevan 1992; Roosevelt et al. 1996, Heckenberger et al. 2007,

Heckenberger et al. 2009).

Within the last few decades, large-scale forest clearing in Acre and Rondônia,

Brazil, has exposed extensive earthworks that date back to about 1250 C.E. (Ranzi and

Aguiar 2005). Trenches up to 7 m deep outline large geometric shapes (or geoglyphs) across the landscape, including circles, squares, rectangles, hexagons, and circles within squares, that range 100-350 m in diameter (Mann 2008). The origins and function of the geoglyphs remain a mystery, but archaeologists interpret them as proof of dense regionally organized populations in the region prior to European colonization.

To construct geoglyphs, inhabitants would have had to clear vast areas of forest (Ranzi

and Aguiar 2005; Schaan et al. 2007), and likely employed fire as a management tool

(Bush et al. 2007).

One of the most common plant families found at archaeological sites in the

Amazon is Arecaceae (Henderson 1995). Many species, including Astrocaryum

vulgare, Acrocomia aculeata, Elaeis oleifera, Attalea maripa, and Attalea speciosa

occur in high numbers on anciently inhabited sites (Baleé 1988; Marcote-Rios and

Bernal 2001). Palms represent the second most economically important plant family

after grasses (Henderson 2006), and their centrality to human societies over the

centuries (Balick and Beck 1990; Campos and Ehringhaus 2003) has given them

favored status within and around settlements and exposed them to frequent

anthropogenic disturbances, such as forest clearing and fire. Balée (1988, 1989) even

suggested that natural palm forests in Amazonia may be anthropogenic, and Henderson

(1995) speculated that throughout history, human activities likely influenced and even

extended the ranges of certain useful palm species. Even today humans are

encouraging the spread of A. speciosa palms when they clear forests (Mitja and Ferraz

2001), whether purposefully or not, as this species quickly colonizes deforested areas

(Anderson 1983).

In our study region, where over 150 geoglyphs have been identified (Schaan et al.

2007), Attalea phalerata Spreng. ex Mart. attracts attention for its exceptional ability to

persist in pastures and cleared forest areas as well as for its tendency to dominate

secondary forest. A. phalerata’s persistent habit is reminiscent of several other related

congeneric palm species that display similar behavior in other parts of the Amazon

(Baily 1947; Pires-O’Brien 1993; Souza and Martins 2002; Salm et al. 2005).

Remarkably, little attention has been given to the specific role of palms in secondary

forest succession in the humid tropics despite the tendency of some to persist after disturbance (but see Anderson 1983; Anderson et al. 1991; Capers et al. 2005; Salm et

al. 2005), and even predominate after forest clearing and repeated burning (Bondar

1964).

We studied the recovery of A. phalerata populations after slash-and-burn agriculture in Acre, Brazil, and evaluated the resistance of A. phalerata palms to forest

clearing and burning. In southwestern Amazonia, palms represent a particularly

important, and many times predominant, element of the forest ecosystem and play an

integral role in forest regrowth, as is evidenced by their frequent predominance in

secondary forests. As one moves westwards across the Amazon Basin, palm diversity

increases (Kahn and de Granville 1992), and the local vegetation classification for Acre

includes several forest types distinguished by the presence of palms (ZEE 2002). While

most studies of early forest recovery following slash-and-burn agriculture begin after the

forest has been cleared and burned (Uhl et al. 1981; Uhl et al. 1982; Uhl 1987; Miller

and Kauffman 1998; Kammescheidt 1999; Gehring et al. 2005), our study adds another

dimension by comparing post-deforestation palm populations with pre-burn palm

populations. Our research questions are: (1) Is Attalea phalerata’s persistence after deforestation due to the stimulation of seed germination (i.e., development of a largely

new population) or survival of the original population? (2) What are the survival and

growth recovery rates of different size classes in clear and intact forest? (3) What are

the impacts of slash-and-burn agriculture on population structure?

Study Area

Research was carried out on a privately-owned farm approximately 60 km west of

Acre’s state capital, Rio Branco (09⁰54′ S 068⁰20′ W). The regional landscape is a

mosaic of pastures, croplands, abandoned fields in various stages of succession, and mature or lightly-logged forest. Pastures predominate along roadsides, whereas deeper within individual properties, small-scale farmers cultivate rice, corn, beans, and manioc on 1-2 ha of cleared secondary or old-growth forest. Following crop harvest, farmers typically leave the area fallow or convert it to pasture for cattle grazing.

The original vegetation in the region is classified as humid, moist tropical forest

(Holdridge 1978). The old-growth forest studied here has a relatively open canopy, including various deciduous tree species, and a dense understory. The rainy season typically extends from October to April with a pronounced dry season between May and

September when monthly precipitation generally remains below 100-mm

(www.acrebioclima.pro.br). An average annual rainfall in the region varies from 1600 to

2000 mm, and mean annual temperature is 25⁰C (ZEE 2002).

Study species

Attalea phalerata Mart. ex Spreng. (synonym Scheelea phalerata (Mart. ex

Spreng) Burret.; local name = ouricuri) is an arborescent, sub-canopy palm up to 20 m

tall with a thick stem (generally 25 to 40 cm in diameter) found in forests along the

southern and western periphery of the Amazon region, including Brazil, Bolivia and

Peru, as well as the drier planalto forests and tropical savannas in Brazil, Bolivia, and

Paraguay (Henderson et al. 1995). Within its natural range, this species also persists in

pastures and cleared forests throughout eastern Acre and northern Bolivia and often

dominates early regrowth forests (Henderson 1995; Carvalho et al. in press). Local

farmers sometimes treat these palms as pests due to persistence and their apparent

invasion after forest clearing. A. phalerata is chiefly pollinated by nitidulid beetles from

the genus Mystrops (Moraes et al. 1996), and also by wind. Seed germination mode is

remote-tubular (Uhl and Dransfield 1987), and during early establishment the

underground stem develops a “saxophone-shaped axis” that buries the apical meristem

deeper into the soil (Tomlinson and Esler 1973). The saxophone-shaped below-ground

stem facilitates palms’ ability to recover and proliferate on disturbed sites, because the

growth point remains protected below the soil surface, and neither fire nor leaf

destruction usually kills the palm (Voeks 1987; Anderson et al. 1991). Few studies have

been published regarding the ecology of this species (Moraes et al. 1996), and those that have focus mostly on seed dispersal (Hesse and Duffield 2000; Quiroga-Castro and

Roldán 2001; Galetti and Guimarães 2004; Rios and Pacheco 2006).

Attalea phalerata has a rich history of human use in the region (Campos and

Ehringhaus 2003; Paniagua Zambrana 2007). Its leaves are used as thatch by rural

populations, and the fleshy mesocarp is enjoyed as a food source in areas of eastern

Acre and Bolivia (Moraes et al. 1996). In the recent past, residents used the woody endocarps as a source of charcoal for smoking rubber collected from the forest. A. phalerata fruits provide an important food resource for many animal species that also act as seed dispersers, including squirrels (Sciurus sp.), agouti (Agouti paca), coati

(Nasua nasua), tapir (Tapirus terrestris), raptors (Caracara plancus), various parrot species, monkeys, macaws (Ara glaucogularis, Anodorhynchus hyacinthinus) (Johnson et al. 1997; Hesse and Duffield 2000; van Holt 2001; Galetti and Guimaraes 2004; Rios and Pacheco 2006).

Materials and Methods

Research Design

To assess the vulnerability of A. phalerata populations to forest clearing, we monitored about 250 palms from August 2006 to October 2007, using a BACI (before- after control-intervention) research design. We installed six 50 x 20 m study plots within an area of intact old-growth forest along the Transacreana Highway (AC-90) in Acre,

Brazil. Three slash-and-burn treatment plots (SLB) were located within a forested area the landowner planned to clear and burn for agriculture. To install plots, we selected three adult (reproductive) palms within and three outside the area slated for clearing and annual crop cultivation, and extended a 50 m transect in a randomly chosen compass direction from each adult tree. All undisturbed control plots (CTL) were separated by at least 50 m, but SLB plots were separated by a minimum distance of only 5 m due to the limited number of adult palms within the small area to be cleared (2 ha). Natural abundance of adult A. phalerata in forests in the study region is typically no more than 5-8 ha-1.

Data Collection

Inventories of A. phalerata from all size classes were taken before (T0) and after

(T1, T2, T3, T4) forest clearing and burning at approximately 4-month intervals. We

separated individuals into five size classes:

(1) seedlings with no more than two entire (i.e., undivided) leaves <2 m long (P1);

(2) post-seedlings with more than two entire leaves <2 m long (P2);

(3) pre-juveniles with divided leaves <2 m long (P3);

(4) immature juveniles with leaves ≥2 m long (with or without trunk) (J);

(5) reproductive adults (A).

We assumed that by the time a young palm develops three leaves, it no longer relies on seed reserves for subsistence.

We carried out the first survey in August 2006 (T0). Within all 20 x 50 m study plots (n=6), we counted the total number of leaves, measured the longest leaf, and tallied the number of leaflets per longest leaf of A. phalerata seedlings and juveniles.

Due to hyper-abundance of seedlings in the majority of control plots (up to 2570 individuals ha-1), we limited the control survey of individuals with leaves less than 2 m

(P1, P2, P3) to two 25 x 10 m subplots: we systematically positioned one subplot to the

right of the transect line between 0-25 m, and the other to the left of the transect

between 25-50 m. For adult palms we counted the number of leaves, number and sex

of inflorescences, and number of infructescences, and estimated total height. We

assigned (X, Y) coordinates to each palm and tagged the newest leaf of all censused

seedlings and juveniles with numbered flagging. In the treatment plots, we inserted a

small iron rod (approximately 1 cm in diameter) alongside each individual, leaving about

10-15 cm protruding from the soil. Using a metal file, we scored tick marks on each rod

to represent the number of leaves for each individual to keep track of individual palms

after the area was cleared and burned. Iron stakes were also used to mark the two

ends of each transect.

Within a week of our initial survey on August 27, 2006 (T0), the landowner

completely slashed two hectares of old-growth forest. After the downed vegetation had dried sufficiently, on September 19, 2006, he burned the area and cultivated rice and corn, followed by beans 9 months later. The asked the landowner to be careful to avoid cutting palms when weeding. In subsequent visits 2, 5, 9, and 14 months later (T1-T4), we re-censused all palms, each time recording the number of newly expanded leaves, the longest leaf and the number of leaflets on the longest leaf for seedlings and juveniles. We also registered palm mortality and recruitment of new individuals. During the first post-deforestation inventory, individuals in treatment plots were tagged with numbered flagging. To facilitate identification of new leaves and monitor leaf production and increment, we marked the newest fully expanded leaf of all individuals with red paint and tagged the longest leaf with flagging.

Statistical Analysis

For statistical analyses, palm abundances in each sample plot were converted to density per hectare, and averages were calculated for the SLB (n=3) and CTL (n=3) treatment groups at each time step. We also calculated relative abundances for each of the five size class: seedlings (P1), post-seedlings (P2), pre-juveniles (P3), juveniles (J), and adults (A). All statistical analyses were conducted using SAS 9.1.3 (SAS Institute,

2003).

To test for differences between SLB and CTL plots prior to intervention, we constructed a generalized linear mixed model using PROC MIXED procedure to model

the effects of treatment and size class on densities of A. phalerata palms at the

beginning of the study (i.e., prior to slash-and-burn). To compare the densities of

germinated seedlings and surviving palms (proportion surviving from one time step to

the next) between SLB and CTL, we examined tabular results.

The effect of treatment, time, and plant size (size class), and their interactions on

palm density and relative abundance were analyzed using generalized linear mixed

models (PROC MIXED), with time as a repeated measure and plot as a random factor.

According to Underwood (1994; also see Green 1979), a significant time x treatment

interaction should signal an effect of forest clearing and burning on palm populations.

When the interaction term was significant at P≤0.10, we used the SLICE DIFF procedure with PROC GLIMMIX to compare among levels of one factor (or interaction term) at the fixed level of the other factor. Annual leaf production rates and leaf length growth increment between time-steps were also analyzed using mixed models. These data were normally distributed, except for T0-T4 leaf length increment (LL). To normalize the data, we transformed this variable by taking the ratio of each observed value (LL) plus one: 1/(LL+1).

Results

Pre-Burn Palm Populations

At the initiation of our study in August 2006, Attalea phalerata populations presented a range of demographic profiles both within and between treatments. In forest control plots (n=3), we counted 80-2570 palms per hectare, while in treatment plots (n=3), we recorded 240-350 palms per hectare (Table 4-1). A super-abundance of young seedlings in Plot CTL-2, making up 70% of total plot abundance, accounted for the elevated average density in control plots. In all plots, seedlings (P1) and post-

seedlings (P2) composed the bulk of the population, while juvenile palms were sparse.

Each plot contained only one adult palm (Figure 4-1).

Table 4-1. Attalea phalerata plot densities and mean treatment densities (± 1 SD) measured in six 50 x 20m control (CTL) and treatment (SLB) plots prior to experimental intervention (T0) and then 2 (T1), 5 (T2), 9 (T3), and 14 (T4) months after slash-and-burn in an area of old-growth tropical rainforest in Eastern Acre, Brazil. Attalea phalerata density per hectare (all size classes) PLOT NO. T0 T1 T2 T3 T4 CTL-1 580 590 610 590 550 CTL-2 2750 2610 2810 2710 2620 CTL-3 80 80 80 80 80 AVG 1077 1093 1167 1127 1083 (± SD) (1317) (1338) (1448) (1395) (1351)

SLB-4 240 190 180 170 150 SLB-5 240 220 220 220 220 SLB-6 350 250 330 270 250 AVG 277 220 243 220 207 (± SD) ( 64) (30) (78) (50) (51)

Figure 4-1. Size class distributions before slash-and-burn (T0) shown as mean proportion (±1 SD) for control (CTL) (n=3) and treatment (SLB) plots (n=3). P1: Seedlings with no more than 2 entire leaves <2 m long; P2: Post- seedlings with more than 2 entire leaves <2 m long; P3: Pre-juveniles with divided leaves <2 m long; J: Non-reproductive juveniles with leaves ≥2 m long (with or without trunk); and A: Reproductive adults.

Despite a substantial range in values, mixed model analysis (PROC MIXED) comparing the density of A. phalerata palms, distributed across size classes, showed no

significant differences between palm distributions in control and treatment plots prior to

forest clearing (p>0.20).

Immediate Effects of Slash-and-Burn

At the end of October 2006—two months after clearing—fallen charred tree trunks littered the slashed-and-burned treatment area. The farmer had felled all the adult A. phalerata palms using a chainsaw, and all other palms remaining from the additional four size classes surveyed had burned to some degree, although in a few cases, fire scorched only a portion of the leaves. Many individuals lost all their leaves, but new

ones already began emerging within five weeks of fire. In spite of the immediate

depletion of green leaf material, mortality was low (18 of 83 originally surveyed palms

confirmed dead). In comparison, control plot mortality was barely perceivable over the

same time period (Table 4-2).

Table 4-2. Densities per hectare (ha) and percent (%) of original A. phalerata palms that died within five weeks of forest clearing and burning (T1) in each control plot (CTL-1, CTL-2, CTL-3) and treatment plot (SLB-4, SLB-5, SLB-6). Data are separated by size class—P1: Seedlings with no more than 2 entire leaves <2 m long; P2: Post-seedlings with more than 2 entire leaves <2 m long; P3: Pre-juveniles with divided leaves <2 m long; J: Non-reproductive juveniles with leaves ≥2 m long (with or without trunk); and A: Reproductive adults. P1 P2 P3 J A ALL

No. of 0, 40, 0 0, 0, 0 0, 0, 0 0, 0, 0 0, 0, 0 0, 40, 0 palms ha-1

(n=3) Percent

CONTROL (%) of 0, 2, 0 0, 0, 0 0, 0, 0 0, 0, 0 0, 0, 0 0, 2, 0 palms

No. of 0, 10, 60 30, 10, 30 10, 0, 0 0, 0, 0 10, 10, 10 50, 30, 100 palms ha-1

(n=3) Percent (%) of 0, 17, 46 23, 8, 17 20, 0, 0 0, 0, 0 1, 1, 1 21, 13, 29

TREATMENT palms

Recovery

Recruitment and survival

We registered a flush of recently germinated recruits in both slash-and-burn and

undisturbed forest areas at the January (T3) census, corresponding to the rainy

seasonal peak in A. phalerata germination (December–February). Nonetheless,

slashing-and-burning seemed to stifle germination in deforested plots. Over the entire study period, P1 seedling recruitment rates were consistently lower in SLB than in CTL plots (Table 4-3). Newly germinated recruits in the SLB plots failed to compensate

relatively high P1 seedling mortality due to slashing-and-burning (Table 4-2). Between

August 2006 and November 2007, 11 seedlings germinated in SLB plots (total area=3000 m2), and 13 seedlings died. Meanwhile, in the CTL plots (total area=1500 m2), 19 seedlings germinated, but only 10 died.

Table 4-3. Number of newly germinated A. phalerata seedlings ha-1 in each control (n=3) and treatment plot (n=3) at 2 (T1), 5 (T2), 9 (T3), and 14 months (T4) after forest clearing and burning. Control plots: CTL-1, CTL-2, CTL-3, and treatment plots: SLB-4, SLB-5, SLB-6. -1 No. of germinated A. phalerata seedlings ha T1 T2 T3 T4 Control (n=3) 0, 80, 0 20, 240, 0 0, 40, 0 0, 0, 0 Treatment (n=3) 0, 10, 0 0, 10, 90 0, 0, 0 0, 0, 0

The overall balance between recruitment and mortality in A. phalerata slash-and- burn populations was negative: Total palm abundance in SLB plots decreased by 25%.

Over the same time period, undisturbed control plots experienced only a slight decline (-

2%). Although SLB palm populations did not recover to pre-treatment size during the 2- yr observation period, they demonstrated resistance to slash-and burn, mainly through survival of resprouting individuals (Table 4-4) that subsequently recruited into larger size classes, in the case of P1 and P2, or alternatively regressed to smaller size classes, in the case of juveniles. Survival probability increased with plant size in SLB plots, but in

CTL plots size mattered less to longevity (Table 4-4).

Table 4-4. Proportion of original A. phalerata individuals surviving per study plot 14 months after slash-and-burn (T5), separated by size class. Control plots: CTL-1, CTL-2, CTL-3, and treatment plots: SLB-4, SLB-5, SLB-6. Data are separated by size class—P1: Seedlings with no more than 2 entire leaves <2 m long; P2: Post-seedlings with more than 2 entire leaves <2 m long; P3: Pre-juveniles with divided leaves <2 m long; J: Non-reproductive juveniles with leaves ≥2 m long (with or without trunk); and A: Reproductive adults. Proportion of original A. phalerata population surviving at T5 P1 P2 P3 J A ALL Control CTL-1 0.91 1.00 0.88 0.89 1.00 0.91 CTL-2 0.90 0.90 0.83 1.00 1.00 0.90 CTL-3 1.00 1.00 — 1.00 1.00 1.00 Treatment SLB-4 0.75 0.54 0.80 1.00 0 0.63 SLB-5 0.67 0.92 1.00 1.00 0 0.83 SLB-6 0.23 0.72 1.00 1.00 0 0.54

Effect of slash-and-burn on A. phalerata populations

The mixed model of palm densities and relative abundances by size class, revealed no time x treatment interaction, suggesting no substantial effect of forest

clearing and burning on the palm populations (Tables 4-5 and 4-6). In modeling

changes in density over time, size x time was significant, and the treatment x size

interaction also tended towards significance (Table 4-5). More detailed analyses of

these interactions showed effects of treatment only on size class P2 (p=0.0167), and

effects of time on size classes P1 and P2 (p<0.005) (but not for P3, J, or A) for all time

intervals except T0-T1 versus T0-T2 and T0-T3 versus T0-T4. When we sliced the

three-way treatment x size x time interaction (p=0.3728), we found effects of treatment

at the following size x time levels: P1 x T4 (p=0.0185); P2 x T3 (p=0.0103); P2 x T4

(p=0.0004).

Table 4-5. Repeated-measures analysis of variance with fixed effects for density by time and size class using PROC MIXED procedure in SAS Numerator Denominator Effect F-value P-value DF DF Time 3 72 0.20 0.8994 Treatment 1 20 0.21 0.6494 Treatment x Time 3 72 0.04 0.9879 Size 4 20 0.41 0.7972 Size x Time 12 72 4.35 <0.0001 Treatment x Size 4 20 2.41 0.0832 Treatment x Size x Time 12 60 1.11 0.3728

Table 4-6. Repeated-measures analysis of variance with fixed effects for relative abundance by time and size class using PROC MIXED procedure in SAS. Numerator Denominator Effect F-value P-value DF DF Time 3 60 0.00 1.0000 Treatment 1 20 0.00 0.9999 Treatment x Time 3 60 0.00 1.0000 Size 4 20 4.55 0.0089 Size x Time 12 60 29.11 <0.0001 Treatment x Time 4 20 10.21 0.0001 Treatment x Size x Time 12 60 6.09 <0.0001

Changes in relative abundance also showed two- and three-way interactions, but

no interaction between treatment and time (Table 4-6). Slicing of the treatment x time x

size interaction (p<0.0001) revealed that treatment affected the following size x time

interaction levels: P1 x T1 (p<0.0001); P1 x T2 (p=0.0008); P2 x T1 (p<0.0001); P2 x

T2 (p<0.0001); P2 x T3 (p=0.0158); P2 x T4 (p=0.0003); P3 x T3 (p=0.0380); P3 x T4

(p<0.0001).

Palm demography

Over time, shifts of A. phalerata individuals across size classes revealed

remarkably similar trends in slash-and-burn and control plots, but with substantial

variation in the magnitudes of change (Figure 4-2). In slash-and-burn plots, P1 seedling

numbers spiked at T1 as slashing and fire bumped most P2 individuals backwards into

the P1 size class. After T1, a continuous drop in seedling (P1) abundances

corresponded to an increase in the relative abundance of post-seedlings (P2) in all study plots, as P1 individuals recruited into the P2 size class and relatively higher P1 mortality rates drew down seedling numbers, and seedling cohort densities remained relatively low for the remainder of the study (Figure 4-2A). Despite the relative expansion of the P2 cohort, the absolute number of post-seedlings in SLB plots

progressively declined over time, and 14 months after burning, neither the P1 nor the

P2 cohort had recovered its original density. Pre-juvenile (P3) densities fell immediately

after slash-and-burn but afterwards rebounded, increasing to over three times the

original cohort size after 14 months (from 33 to 93 individuals ha-1). Juvenile (J) palms

present prior to forest clearing and burning experienced no mortality during the study

period, although most retrogressed into a smaller class due to the slash-and-burn

treatment. At the end of the study (14 months after clearing and burning), size class

distributions in the SLB plots revealed an unbalanced concentration of individuals in the

P2 and P3 size classes, and minimal representation in other classes (Figure 4-3).

Meanwhile, in undisturbed forest, we observed a similar complementary pattern of decreasing seedling (P1) contributions and increasing post-seedling (P3) contribution to the palm populations between T1 and T4 (Figure 4-2B). The post-seedling (P2) cohort doubled in population size over the study period. The initially small pre-juvenile cohorts

in CTL plots shrunk slightly by the end of the study, and juvenile (J) cohorts remained

relatively stable (Figure 4-2B).

Figure 4-2. Mean relative abundance of Attalea phalerata size classes in 50 x 20 m plots at each census before (T0) and 2 (T1), 5 (T2), 9 (T3), and 14 months (T4) after forest clearing and burning in A) slash-and-burn plots (n=3) and B) control plots (n=3).

Figure 4-3. Size class distributions on November 1, 2007 (T4), one year after slash- and-burn disturbance shown as mean proportion (±1 SD) for control (n=3) and slash-and-burn plots (n=3). Data are separated by size class—P1: Seedlings with no more than 2 entire leaves <2 m long; P2: Post-seedlings with more than 2 entire leaves <2 m long; P3: Pre-juveniles with divided leaves <2 m long; J: Non-reproductive juveniles with leaves ≥2 m long (with or without trunk); and A: Reproductive adults.

Leaf production and growth

According to our mixed model, the slash-and-burn treatment increased annual leaf

production (p=0.0002), and the size x treatment interaction was nearly significant

(p=0.1090). Mean annual leaf production rates were greater in slash-and-burn plots

than in control plots for seedlings (P1) (p=0.0001) and post-seedlings (p2) (p<0.0001),

but not for pre-juveniles (P3) (p=0.4339) or juveniles (J) (p=0.3014). Seedlings and

post-seedlings together annually produced on average 3.35 (±0.06 SE) new leaves in

slash-and-burn plots compared with 1.95 (±0.18 SE) in control plots. One year after the

SLB plots were burned, 73% of surviving P1, P2, P3 individuals returned to or grew

beyond their original leaf number. The proportion of all palms producing at least one

new leaf between censuses was greater in the slash-and-burn than in control plots between January and May 2007 (T2-T3) (p=0.0809; Mann-Whitney-Wilcoxon Z-test). In

SLB plots, leaf production peaked during the rainy season (January-May 2007), and

95% of inventoried palms produced at least one new leaf during this period. Meanwhile, leaf production among palms in control plots tended to slow during the same interval

(Figure 4-4). Otherwise, no differences appeared between treatments.

T1 – T2 T2 – T3 T3 – T4

Figure 4-4. Proportion of non-reproductive Attalea phalerata individuals that grew at least one new leaf per census interval in slash-and-burn (n=3) and control plots (n=3). Time steps refer to 2 (T1), 5 (T2), 9 (T3), and 14 months (T4) after forest clearing and burning.

Leaf length growth over the entire study period (14 months) displayed a tentative positive treatment effect of slash-and-burn (p=0.1008). Differences in leaf growth between SLB and CTL populations were not related to size class (p=0.3822), and there was no interaction effect between the two factors (p=0.3301). Detailed analyses revealed accelerated growth in leaf length in SLB plots compared with CTL plots during

the T2-T3 survey interval due to treatment effect (p<0.0001), with no size (p=0.1381) or

interaction effects (p=0.1693). During the T3-T4 survey interval, there was an

interaction effect (p=0.0095) on leaf length increment. Treatment continued to have a positive effect on leaf growth in size classes P2 (p=0.0075), P3 (p=0.0012), and J

(p=0.0252), but not P1 (p=0.1352). We did not include adults in leaf growth analyses because no leaf length measurements were made for that size class.

Discussion

Despite the destructive nature of forest clearing and burning, A. phalerata populations demonstrated a remarkable capacity to recover from disturbance.

Throughout the mosaic landscape of eastern Acre, Brazil, A. phalerata palms abound in areas of abandoned crop and pasture lands and frequently dominate secondary forests

(Carvalho et al. in press), bearing witness to species’ resistance to human disturbance.

Just 14 months after slashing and burning, our analyses showed no overall treatment

effect on the palm population densities or size class relative abundances, except for the

adults that were cut.

Palm Colonization or Persistence?

Attalea phalerata’s proliferation on abandoned agricultural and pasture lands, and its almost immediate reappearance after slash-and-burn raises questions about the origins of seedlings emerging shortly after the fire. After the farmer deforested and burned the treatment area, the charred ground was practically void of all plants and only small leaf stubs were visible on most surviving palms, if discernable at all. Nonetheless, after five weeks, a relatively dense cover of palm seedlings gave the impression that they had colonized the site. Germination, however, contributed very little to post-fire establishment. Instead, much of the original palm population quickly began to produce

new leaves, and a somewhat compromised but robust A. phalerata population ( 220

±30 SD palms ha-1; n=3) reoccupied the SLB plots. Very few other trees were present

in SLB plots after burning, although grasses, herbs, and vines began to establish along

with the planted rice and corn. Mortality five weeks after slash-and-burn was fairly low

(21% ±8% SD).

Impediments to seedling recruitment

While fire has been reported to stimulate seed germination in other palm species

in Amazonia (Schroth et al. 2004), burning seemed to suppress A. phalerata recruitment

from seed. We found more than twice as many newly germinated seedlings in control

plots than in slash-and-burn plots. Over the entire 14-month study, we counted only 11

germinants within SLB plots (3000 m2), compared to 19 in half the area (1500 m2) of

CTL plots. In comparable studies, Voeks (1987) also witnessed lower Attalea funifera

recruitment from seed in a slash-and-burn area than in undisturbed forest, and

Anderson et al. (1991) reported lower germination of Attalea speciosa after fire

associated with shifting cultivation.

Scarcity of viable seeds and harsh conditions in the slash-and-burn plots may

have limited recruitment of new seedlings. The felling of all reproductive A. phalerata

palms within the slash-and-burn area eliminated immediate seed sources; however, an

old infructescence hung from one of the adult trees present prior to cutting and burning,

indicating fruit production within the last year or two. Fire exacerbates the hot dry

conditions in clearings and can kill palm seeds, even when buried (Anderson et al.

1991, Mitja and Ferraz 2001). The 10 new seedlings that germinated within the SLB

area between January (T2) and May (T3) were probably present in the soil as seeds at

the time of disturbance and withstood the fire.

Even though Attalea palm seeds are protected by a thick woody endocarp, they

depend on burial by dispersal agents for successful seedling recruitment (Anderson

1983). Because palms store endosperm polysaccharides in the form of pure

mannans—a substance incapable of retaining much water— there is no ‘water buffer’ to

prevent desiccation, and direct exposure to sunlight or high temperatures can quickly

dry out the seed, causing desiccation of the embryo and germination failure (Anderson

1983, Pinherio 2008). Forest clearing for slash-and burn usually reduces populations of

animal seed dispersers (Dunn 2004) and hence limits seedling recruitment (Holl 1999,

Wijdeven and Kuzee 2000). Attalea phalerata is mostly dispersed by scatter-hoarding

rodents that tend to avoid large open spaces (Uhl et al. 1988, Aide and Cavelier 1994,

Nepstad et al. 1996, Cardoso da Silva et al. 1996, Holl 1999), but since the clearing was

relatively small (about 2 ha), animals may have dispersed in seed from the bordering

forest. In forest environments, A. phalerata fruits ripen and fall during the rainy season,

(January–April), and seedlings begin to emerge from the soil 10–12 months later. If

dispersal agents successfully carried A. phalerata seeds onto the site after burning,

germination should begin to increase after the new year (2008), and shade from

developing regrowth forest should facilitate germination of incoming seeds and reduce

risks of desiccation. A. phalerata palms growing near the edge of the forest clearing may help counteract scarcity of dispersal agents by producing more seeds, because these palms increase fruit production in response to improved light environments (see

Chapter 3).

Mechanisms of persistence

We found clear evidence that survival and persistence of the extant population

overwhelmingly drove post-disturbance recovery of A. phalerata populations (Anderson

1983; Voeks 1987). More than half of each of the original populations was still alive 14

months later ( =67%; Table 4-4). Researchers attribute widespread palm tolerance to disturbance to an exceptional capacity for resprouting present among some species

(Bondar 1964; Anderson et al. 1991; McPherson and Williams 1998; Henderson 2002).

At germination, the seedling axis of various palm genera (e.g., Attalea, Sabal, Syagrus,

Acrocomia, Allagoptera, Astrocaryum, Chamaedorea, Gastrococos, Dypsis, Hedyscepe,

Kentiopsis, Oenocarpus, Raphia, and Rhopalostylis) burrows into the ground to increasing depth with the production of each new leaf (Bondar 1964; Corner 1966;

Tomlinson 1990). As development proceeds, the axis turns upwards, producing a saxophone-shaped underground stem (Tomlinson and Esler 1973). The apical meristem remains protected beneath the soil surface until stem emergence years or decades later, enabling young palms to repeatedly resprout after cutting and burning

(Corner 1966; Tomlinson 1990; Anderson 1983; McPherson and Williams 1998; Souza et al. 2000; Henderson 2002). The protected position of the growth point also provides access to higher moisture levels, escaping dry conditions at the soil surface, and protects the apical meristem from herbivores (Tomlinson 1960a; Brown 1976).

Excavations of several young stemless A. phalerata palms of various sizes revealed stem axes that extended downwards 25-30 cm into the soil before turning upwards, and a swollen tuber-like structure from which new leaves emerged that increased with increasing palms size (cf. Bondar 1964; Tomlinson 1960b; Tomlinson

1990). In a congeneric species, A. speciosa, Henderson (2002) reported that the growth point eventually reaches 1-1.5 m below the surface. The burial depth of the apical meristem largely determines resistance to forest clearing, slash-and-burn

agriculture, and other anthropogenic disturbances (Charles-Dominique et al. 2003).

While forest disturbance favors species like A. phalerata, it can be detrimental to other palm species (Svenning 1998), many of which fail to regenerate in open areas

(Pedersen 1994; Moraes et al. 1995). The saxophone axis formed during establishment growth provides an extra advantage that ensures the survival of young A. phalerata palms after high-intensity disturbances. Little is known about the extent of “saxophone stem” specialization in palms. Tomlinson (1990) suggested that it may be more widespread than currently appreciated and seems to have evolved independently within different groups (also see Henderson 2002).

Although no A. phalerata with aboveground stems survived in the SLB plots due to felling, when farmers leave such larger palms standing, they likely survive because of persistent leaf bases along the trunk that insulate the apical meristem from high temperatures during fire (Souza et al. 2000; McPherson and Williams 1998). Once a palm grows tall enough for its crown to escape fire, it is virtually immune to burning, as palm stems are highly fire-resistant due to the lack of a vascular cambium (Tomlinson

1979).

Continued Post-Fire Recovery

Results revealed some negative size-dependent effects of slashing-and-burning at our study site. Besides the adult size class, which was completely cut down, the smallest size class (P1) suffered the highest mortality within the slash-and-burn plots

( 45% ±28% SD), likely due to undeveloped root systems, smaller underground carbon stores, and closer proximity of the apical meristem to the soil surface. Also, in remotely germinated palm species, like Attalea, the cotyledonary axis can remain attached to the seed for more than a year at the seedling develops (Pinheiro 2008).

Thus, for young seedlings still attached to the seed, an exposed cotyledonary axis near

the soil surface ran a high risk of being severed from the burn, cutting short the transfer

of endosperm reserves to the emerging plant and leading to death.

As young palms increase in size, their root systems expand and they accumulate

below-ground resources, facilitating new leaf growth after slash-and-burn (Miller and

Kauffman 1998). Since most of the leaf material left after slashing was destroyed in the fire, resprouting A. phalerata palms relied on stored carbohydrate reserves in their underground stem to produce new leaves during the initial phase of resprouting

(Mendoza et al. 1987). The bigger the individual, the larger the stem tuber and the better the chance of resprouting and surviving (McPherson and Williams 1998).

Palm growth

Leaf production of surviving A. phalerata seedlings, post-seedlings, and pre- juveniles was so successful in our slash-and-burn plots that 73% returned to or surpassed their original number of leaves within 14 months after disturbance. Palms demonstrated overall faster growth rates in slash-and-burn plots than in undisturbed control forest, but again, results were size-dependent. Both annual leaf production rates and leaf length increments of seedlings (P1) and post-seedlings (P2) were greater in the slash-and-burn plots than the undisturbed control plots. Greater leaf growth at the slash-and-burn site may be attributed to greater access to sunlight and nutrients than in undisturbed forest (Sanchez et al. 1983; Chazdon 1986; Voeks 1987; Kainer et al. 1998; McGrath et al. 2001).

In contrast, slash-and burn did not favor leaf production in larger pre-juvenile (P3) and juvenile (J) palms. Despite 100% survival of juvenile palms in the SLB plots, none of the juveniles returned to their original size after 14 months, but instead, regressed

into the pre-juvenile size class. Mendoza et al. (1987) also found that complete leaf

removal reduced leaf production rates in stemless juvenile and immature Astrocaryum mexicanum palms. Once palms begin stem elongation, they invest more resources in height growth than in leaf production, so this may have retarded recovery of this cohort.

Compensatory growth also helps explain higher growth rates in SLB than in CTL populations. After slashing-and-burning, palms were also almost entirely denuded of leaves. In many plants, partial or complete defoliation can stimulate leaf production and growth (McNaughton 1979; Mendoza et al. 1987; Oyama and Mendoza 1990; Chazdon

1991; Ratirarson et al. 1996; Anten et al. 2003; Parra-Tabla et al. 2004; Bruna and

Nogueira Ribeiro 2005) due to improved light capture (Chazdon 1986), increased photosynthetic rates in remaining leaves (Anten and Ackerly 2001), or mobilization of stored carbohydrates (McPherson and Williams 1998; Chazdon 1991). In contrast,

Anderson (1983) found no change in growth rates of stemless Attalea speciosa palms after a single defoliation event, and McPherson and Williams (1998a) observed declining leaf regrowth and even death with increasing frequency of defoliation events in

Sabal palmetto. Under scenarios of high frequency disturbances (fire, weeding, and cattle grazing), A. phalerata survival and growth rates would likely diminish, but further study is merited to explore this palm species’ threshold of resistance.

Contrasting temporal patterns in leaf growth appeared between slash-and burn and undisturbed control plots. Beneath the shade of the forest canopy, light resources were presumably further limited by fairly persistent cloud cover during the wet season, and new leaf production and leaf length growth in A. phalerata were lower in forest during this period (Figure 4-4). DeSteven et al. (1987) also observed lower rates of leaf

production during the rainy season for various palm species in Panamanian forests.

Conversely, new leaf production peaked in slash-and-burn plots during the wettest months of the year between January (T3) and May (T4). Full exposure to sunlight in the slash-and-burn area mitigated the diminished light resources from wet-season cloud cover, and wet season rains probably enhanced nutrient availability from burned ash

(Nye and Greenland 1960), stimulating growth.

Plant demography

As expected, the control plots in standing forest maintained relatively constant palm densities and population structure throughout the study period (Figures 4-1 and 4-

3). Slash-and-burn plots underwent only a small drop in total population size, but a detailed assessment revealed more substantial alterations in population structure. In the SLB area, a combination of seedling mortality, steady recruitment of P1 seedlings into the P2 cohort (39%), and few new germinants to replace dead individuals resulted in a continuous but minor decline in A.phalerata seedling numbers and their complete absence in two of three plots at the end of this study. Mortality contributed to P1 decline in undisturbed forest as well, but to a lesser extent (10%), while 54% of seedlings recruited into the P2 size class. Regardless of treatment, relative abundance of seedlings (P1) was lowest at the end of the study, which was immediately prior to the seasonal peak in germination (December–February).

Accelerated growth rates of younger palms in the SLB plots led to more rapid recruitment into P2 and P3 cohorts, resulting in substantial changes in size class relative abundances. Approximately one year after slashing-and-burning, relative abundances of all but the pre-juvenile size class were lower than at the initial inventory: the pre-juvenile class (P3), tripled in size. This increase reflected persistence of the P3

cohort (80% of P3 individuals remained in this class), post-seedling recruitment (30% of

P2 individuals advanced to P3) as well as retrogression of all three juveniles into the P3

size class (100%). Meanwhile, palms in the CTL plots demonstrated a different pattern:

post-seedling (P2) and juvenile (J) size classes expanded, and the pre-juvenile (P3)

size class shrank.

At the end of our study, the disappearance of seedlings from two of the three SLB

plots suggested that the smaller size classes were not being replenished. Nonetheless,

despite zero adult survival, we expect that future germination will repopulate the P1

cohort through successful dispersal of seeds into the cleared area coupled with

anticipated favorable growing conditions beneath the developing regrowth forest (Uhl et

al. 1982; Guevara et al. 1992). Other demographic studies of palms have argued that the seedling cohort—or the smallest size class—tends to have little effect on population growth rates. Instead, recruitment into and survival of post-seedling and immature stages generally has the most influence on population growth and stability (Bullock

1980; Piñero et al. 1984; Enright and Watson 1992; Pinard and Putz 1992; Pinard

1993; Ratsirarson et al. 1996; Barot et al. 2000). Growth of the pre-juvenile cohort and stability of juveniles suggested a strong potential for growth and recovery of recuperating populations, and high light conditions in SLB plots should also help accelerate juvenile maturation (Salm et al. 2005; Kahn 1986). Once local site fecundity is restored, full recovery of A. phalerata is promising, and the species may eventually come to dominate regenerating forest vegetation, as is common in the region (Carvalho et al. in press).

Disturbance promotes palm dominance

Because Attalea is long-lived, its ability to prosper in regenerating forest surpasses

that of typical pioneer species that dominate early successional forests, but usually die

within 20-30 years (Saldarriaga et al. 1988). In the southwestern Brazilian Amazon, A.

phalerata occurs at relatively low densities of 3-8 adults ha-1 in undisturbed forest, while

under more open conditions, such as savannas, transitional forests, and pantanal

wetlands, as well as in disturbed landscapes, such as cattle pastures and secondary

forests, palms densities tend to be greater than 10 adults ha-1. The congeneric

arborescent palm, Attalea maripa, is also generally rare in undisturbed forest, but

abundant in disturbed forest, as well as in secondary forest in southeastern Amazonia

(Salm et al. 2005). Kahn and de Granville (1992) stressed the importance of forest disturbance in determining palm occurrence and distribution (also see Kahn 1987).

Large arborescent palms behave like light-demanding successional species and depend on large disturbance events to become ecologically dominant (Salm et al.

2005).

Although this research represents a single case study, together with works by

Anderson (1983) and Voeks (1987), we clearly establish a connection between survival by vegetative sprouting capacity and persistence and abundance of Attalea spp. in disturbed landscapes. Although many trees growing in the tropical forests of the

Amazon resprout and survive after low intensity disturbance (Uhl and Jordan 1984), increasing severity of disturbance, such as more frequent and high intensity fires, diminishes sprouting capacity in most tropical tree species (Uhl et al. 1981; Putz and

Brokaw 1989; Kauffman 1991; Kennard et al. 2002; Hooper et al. 2004). Trees tend to have a lower survival threshold of disturbance intensity than many palms. Young

stemless palms that maintain their apical meristem buried below ground, including

Attalea spp., can more easily escape fire’s destructive path, and the lack of vascular cambium in palm trunks makes them highly resistant to fire after they have grown taller.

While woody dicotyledonous plants exhibit various modes of resprouting from root collars, underground stems, roots, and the tree trunk or branches (Del Tredici 2001), palms and other monocotyledons—with the exception of clonal and multi-stemmed species—rely on a single apical meristem for all growth and re-establishment following disturbance. Protection of this growth point is critical for survival, and the capacity to continue leaf production following damage characterizes resprouting in palms.

Ecological Pressures Build Resistance

In a recent examination of the phylogenies of three large fire-adapted shrubby tree genera, Bond and Midgley (2003) concluded that, at least as a fire survival strategy, sprouting is not a conserved trait, and that sprouters seem to have evolved from non- sprouters various times. Their conclusion challenged Wells’ (1969) long-standing hypothesis that sprouting is an ancestral trait among woody dicotyledons (Schwilk and

Ackerly 2005). Bond and Midgley (2003) argued that sprouting is highly labile (also see

Vesk and Westoby 2004) and proposed that strong ecological factors have led to the adaptation of sprouting in dry, fire-prone environments at various times and various places across the globe.

Attalea phalerata seems to be particularly adapted to human-induced disturbances, such as deforestation and fire. Widespread natural disturbances are uncommon in recent Amazonian history with the exception of fairly frequent and large blowdowns in some parts of the region (Nelson et al. 1994). If sprouting is indeed highly labile, the long and continuing history of palm use by Amazonian human

populations has likely shaped the distributions and abundances of palm species with resprouting capacity and mechanisms to protect the apical meristem, like A. phalerata.

In response consistent use, to planting, and repeated exposure to disturbances, such as

fires, these palm populations can extend their range and increase dominance within

disturbed landscapes. Correlations between resistance, palm resprouting capacity,

saxophone-shaped growth axes, and population densities and distributions with

historical settlement patterns in the Amazon have yet to be thoroughly explored. Today, forest clearing and human disturbance across the Amazonian landscape is driving and will continue to drive the spread of disturbance-resistant palm species, like Attalea. Our results confirm that A. phalerata populations quickly recover after slash-and-burn agriculture, and will likely constitute a principal component of the future regenerating forests in the Southwestern Amazon.

CHAPTER 5 CAN EXISTING POPULATIONS OF NATIVE OLEAGINOUS PLANTS CONTRIBUTE TO BIOFUEL PRODUCTION IN AMAZONIA? FRUITING PHENOLOGY AND OIL YIELDS FROM TWO NATIVE ATTALEA PALMS IN SOUTHWESTERN AMAZONIA

Introduction

Worldwide, sovereign nations seek to achieve energy independence and without

reliance on imported fossil fuels, like oil, natural gas, and coal. This has recently led to

an explosion of interest and investments in biofuels (Calvin 1983; Koh and Ghazoul

2008). Amid growing concerns over climate change, biofuels have also emerged as

both a savior and a villain (Dauvergne and Neville 2009). While some argue that

biofuels are a clean sustainable alternative energy source that will reduce greenhouse

gas emissions (Farrell et al. 2006; Pacala and Socolow 2004), others insist biofuels will

actually increase emissions of global warming gases (Fargione et al. 2008;

Scharlemann and Laurance 2008; Searchinger et al. 2008), accelerate tropical rain

forest destruction and biodiversity loss (Koh and Wilcove 2007; Groom et al. 2008), and

drive up world food prices (Escobar et al. 2009; Rathmann et al. 2010). Most biofuel

research and incentive programs have focused on large-scale production from non-

native plantation-grown crops, such as soybean, African oil palm, rapeseed, corn, jatropha, sugar cane, and sunflower. Meanwhile, the potential of small-scale, locally-

based biofuel production has been largely overlooked (but see Coello 2000; Lima et al.

2006; Gmunder et al. 2009; Ansley et al. 2010).

In many developing countries, populations living in the remote countryside depend

on imported and largely unaffordable diesel and gasoline for their energy needs, both for electrification and transportation. With the global explosion of interest in biofuels, energy-poor rural communities and researchers alike are advocating a closer look at

locally abundant native plants as a potential energy source (Fletcher et al. 2010). Oil

extracted from native forest species can be used directly in diesel engines with little or

no modification, or it can be converted to biodiesel through transesterification, or

catalytic pyrolysis (Demirbas 2009).

Use of abundant native oleaginous plants for bioenergy bypasses the need for

land conversion or forest clearing, as is necessary for non-native plantation-grown

biofuel crops, like jatropha, soybean or African oil palm. Native sources are already producing fruits that provide oil, so there is no lag for plants to reach the reproductive phase. Furthermore, native plants require fewer infrastructure investments and inputs,

such as fertilizer, pesticides, and irrigation and the associated technologies and

maintenance. On the other hand, disadvantages of native oils include poor

understanding of native species’ ecology, uneven distributions, difficult access in remote

locations, high yield variability, unpredictability, and generally lower oil yields per unit of

land compared with plantations, and make potential producers wary of investing in

native species.

Our aim was to explore the biofuel potential of two native palms in Amazonia,

where people rely heavily on diesel to fuel electricity generators and out-board motors

used in riverboat travel, especially during the rainy season when many roads become

impassable. A locally-derived, affordable energy source would be very welcome by

local people living in remote forested areas, such as extractive reserves, as well as in

rural agricultural settlements.

Amazonian rainforests support a diversity of native species with biofuel potential

(Balick 1979; Calvin 1983; Pesce 1985). Among tropical tree taxa, the palm family

(Arecaceae) boasts numerous Neotropical species with oil-rich fruits (e.g., Oenocarpus

bataua, Attalea spp., Astrocaryum spp., Mauritia flexuosa, Acrocomia aculeata, Elaeis oleifera, Syagrus spp.; Hodge 1975; Balick 1979; Pesce 1985). Certain palm species

naturally occur in monodominant stands, facilitating extraction (e.g., Mauritia flexuosa,

Euterpe spp.), while others exhibit an extraordinary capacity to thrive on deforested and

degraded lands, at times forming dense agglomerations of palms after disturbance

(e.g., Attalea spp.).

Attalea phalerata Mart. ex Spreng. and Attalea speciosa Mart. ex Spreng.

(Arecaceae) have oil-rich fruits (60-70% lipid content) and are abundant throughout

Amazonia in both old-growth and converted/disturbed forest habitats. In spite of their

potential importance, limited information is available on A. phalerata ecology or oil

production (Moraes et al. 1996), although a handful of published studies discuss seed

dispersal (Quiroga-Castro and Roldán 2001; Hesse and Duffield 2000; Galetti and

Guimarães 2004; Rios and Pacheco 2006). On the other hand, A. speciosa has been

more broadly studied for its ecology (Anderson 1983; Anderson et al. 1991; Barot et al.

2005), socio-economic value (Teixeira 2008), and political ecology (Porro 2005), with

few studies on its use for oil (Lee 1930), conversion to biodiesel (Lima et al. 2007) and

electricity generation (Teixeira and Carvalho 2007; Teixeira 2008). Practically, all

studies on A. speciosa, however, have been conducted in the eastern Amazon.

In this study in southwestern Amazonia, our aims were: (1) to assess the potential

of A. phalerata and A. speciosa as sources of oil for the production of biofuel; (2) to

generate ecological knowledge to improve estimates of oil production of these species;

and (3) facilitate sustainable management. We accomplish this by comparing A.

phalerata and A. speciosa in two very different environments where they frequently

occur: old-growth forest and pasture. Our research focuses on key ecological aspects

relevant to oil production, including fruiting phenology and variability of fruit production at various scales, within and between environments and species, and among populations and individuals. We then integrate fruiting phenology records with Attalea fruit biometry data and chemical analyses of fruit-derived oils. To complete the story, we explore how biofuel production based on native palms could be feasible in both remote and relatively accessible rural areas in Amazonia.

Methods

Site Description

We conducted our research in two main areas located in western Rondônia

(09⁰47′ S 066⁰24′ W) and in eastern Acre, Brazil (10⁰00′ S 067⁰58′ W). In the region, moist tropical forest coexists with a mosaic of pastures, agricultural lands, and abandoned fields in various stages of succession over a relatively flat landscape.

Throughout the region, extensive cattle-grazing predominates along paved roads, often forming continuous pastures that stretch for miles.

Mean annual rainfall varies from 1600 to 2000 mm. The rainy season extends from

October to April with a pronounced dry season between May and September, when

monthly precipitation rarely surpasses 100-mm (INMET 2008). Mean annual

temperature is 25⁰C (ZEE 2002), but cold fronts sweep through the region during the

early dry season, driving temperatures down into the lower teens (⁰C).

Study Species

Attalea phalerata Spreng. ex Mart. (common names=“uricuri” in Brazil; “motacú” in

Bolivia) is an arborescent palm that occurs in forests throughout the southern and

western periphery of the Amazon Basin, including Brazil, Peru, and Bolivia, as well as

the drier tropical forests and savannas of Brazil, Bolivia, and Paraguay (Henderson et

al. 1995). Within its range, this species flourishes in pastures and tends to dominate

disturbed and secondary forests (Henderson 1995, Carvalho et al. in press). A. phalerata palms grow up to 20 m tall and 25 to 40 cm in diameter, but in old-growth forest, leaf crowns of mature palms usually remain beneath the upper canopy.

Individuals usually produce up to four infructescences of 350-800 fruits per year (Figure

5-1A), and each fruit consists of an oily mesocarp (pulp) and the seed consisted of 1-6 endosperms, or kernels.

Attalea speciosa Spreng. ex Mart. (common names=“babassu” in Brazil; “cusi” in

Bolivia) occurs at variable densities in old-growth forests across the Amazonian regions of Guyana, Suriname, Brazil, and Bolivia, and is especially abundant in transition zones between tropical forest and other vegetation types. It thrives in cleared forest areas, like pastures or abandoned agricultural fields, frequently forming dense monodominant stands (Anderson 1983; Henderson et al. 1995). A. speciosa is an arborescent upper canopy palm that reaches heights of up to 30 m in forest and has a thick stem, usually between 25 and 41 cm in diameter. Most palms produce up to four infructescences of

200-400 fruits per year (Figure 5-1B), and each fruit usually contains 1-4 oil-rich kernels.

Contrary to A. phalerata, its mesocarp is dry and powdery.

Both Attalea species exhibit a relatively unique reproductive biology, where palms

alternate among inflorescences that are exclusively pistillate (only female flowers),

staminate (only male flowers), and sometimes hermaphroditic (mixed female and male

flowers) on the same plant. The ratio of pistillate, staminate, and hermaphroditic

inflorescences has a considerable effect on fruit production and, hence, is critical to understanding oil production in these species (see Chapter 3).

A B

Figure 5-1. A) A. phalerata and B) A. speciosa infructescences from study sites in Acre and Rondonia, Brazil, respectively.

Site Selection

We selected six study sites (three actively grazed pastures and three areas of well

preserved old-growth tropical forest) for each palm species after random farm visits

along principal roadways within 300 km of Rio Branco, Acre, Brazil (A. phalerata), and

along a 35 km stretch of the BR-364 Federal Highway in Extrema, Rondônia, Brazil (A.

speciosa). Study sites were scattered broadly across the region to capture spatial variation. The two Attalea species never co-occurred within study sites. All sampled pastures were cleared from forest 15-30 years ago, and no chemical fertilizers were

reportedly used on any of the pastures.

Data Collection

Reproductive phenology

In January 2006, we selected 12 reproductive palms at each old-growth forest and pasture site for phenological observations (n=72 for each species). At each forest site, we observed palms within areas of approximately 2-5 ha. In pastures, palms typically grew closer together, and we observed all 12 individuals within a roughly 1 ha area at each site. We only selected palms with evidence of reproduction and leaf crowns that were fully visible from the ground.

Using binoculars, we conducted monthly flowering and fruiting observations of both Attalea species from January 2006 to June 2007. For A. phalerata we continued phenology observations through December 2007 due to limited flowering frequency during the initial 18 months. Methodological details of flowering phenology observations are presented in Chapter 3. For fruiting phenology, we recorded the azimuth of each infructescence, or fruit bunch, with a compass and followed it continuously through four stages of fruit development: (1) immature—fruit initiating development with flower bracts still present, (2) green—small fruit with soft endosperm, (3) unripe—fruit similar in size to ripe fruit but endosperm still in liquid form, and (4) ripe—hardened endosperm with fruit falling already falling to the ground.

Fruit bunch collection and processing

We randomly selected two individuals at each of the six phenology sites, from which we collected a ripe infructescence using a long-handled pruner. We then weighed the entire infructescence, counted the fruits, randomly selected a sub-sample of 30 fruits, weighed each fruit, and measured its diameter and length. Forest palms

with mature fruit were difficult to find, so we sometimes removed infructescences with ripened fruits already beginning to fall, or collected recently fallen fruits from the ground.

We peeled and de-pulped Attalea fruits from each sub-sample using a knife, and broke open the endocarp with a machete to remove the kernels, or endosperm. The epicarp and endocarp were discarded, and for each fruit we weighed the mesocarp

(pulp) and endosperm (kernel) separately. Mesocarp and endosperm samples were dried at 50⁰ C for at least 72 hours and then weighed again at the Food Technology

Research Laboratory (UTAL) at the Federal University of Acre. Since A. speciosa

produces a dry non-oily mesocarp, only the endosperm was measured using the same

procedure.

Chemical Analysis

To analyze the mesocarp and endosperm for lipids (i.e., oil) content, we grouped

dry samples by study site and ground them into a coarse flour using a heavy-duty cast

iron corn grain grinder (Pro-line Trading Corp.). Ground samples were then frozen in

sealed plastic bags for later chemical analysis. Lipids were determined using the

Soxhlet extraction procedure described in the chemical analytical methods and rules of

the Instituto Adolfo Lutz (1985), using three 5-g replications of each composite sample.

To calculate oil yields per palm, we used the following model,

OIL PALM-1 (kg) = (No. of infructescences palm-1) * (no. of fruits infructescence-1) * [(mesocarp mass fruit-1 (g) * % oil in mesocarp) + (endosperm mass fruit-1 (g) * % oil in endosperm)] / 1000

Next, to convert kilograms of oil to liters, we divided the number of kilograms by the specific gravity (g mL-1 at 15°C) of the oil for each species: A. phalerata=0.9231 (Pesce

1985) and A. speciosa=0.868 (Lee 1930).

Data Analysis

Fruiting frequencies were calculated as the number of events registered annually

(cf. Newstrom et al. 1994). Fruiting intensity refers to the proportion of individual palms

that fruited over a set time period (cf. Herrerias-Diego et al. 2006). We tested for

differences in A. phalerata and A. speciosa fruiting frequency between forest and

pasture, using permutation tests with 10,000 repetitions. To test for differences in fruit

biometry measures (total fruit mass, mesocarp, and endosperm mass) between

environments, we used a Satterthwaite-adusted t-test for normally distributed samples

and a non-parametric Mann-Whitney-Wilcoxon test for variables with non-normal

sample distributions (Quinn and Keough 2002). We conducted all statistical analyses

using SAS 9.1.3 for Windows (SAS Institute, Inc. 2002-2003).

Results

Comparison of Fruit Production in Forest and Pasture

Inflorescence production

Leading up to fruit production, Attalea phalerata and A. speciosa both produced significantly more inflorescences per palm in pasture than in old-growth forest, and specifically produced more female inflorescences in pasture (P<0.001) (Figure 5-2; see

Chapter 3). Over the entire observation period, both Attalea species also produced

more male than female inflorescences, regardless of environment: average annual

male to female inflorescence ratios per palm in forest and pasture were 5.0:1.0 and

1.4:1.0, respectively, for A. phalerata, and 2.5:1.0 and 1.3:1.0, respectively, for A.

speciosa. Both female and hermaphroditic inflorescences bore fruit. In A. phalerata,

hermaphroditic inflorescences were rare and occurred in only one pasture. A. speciosa,

in contrast, produced hermaphroditic inflorescences in both environments and across

study sites.

Over the entire study period (24 months), we followed a total of 26 A. phalerata

pistillate (female) inflorescences in old-growth forests and 125 in pastures, through

anthesis to fruit set, fruit development, and fruit ripening. Over 18 months, we observed

52 A. speciosa female inflorescence in forests and 113 in pastures.

Infructescence production

Annual infructescence production was greater in pasture than in forest for both

Attalea species in 2006 (P<0.0005, n=36; Figure 5-2) and 2007 (P<0.002; A. phalerata:

n=24, A. speciosa: n=36; see Chapter 3). Fruit production closely reflected female

inflorescence patterns, since infructescence failure prior to fruit maturation was rare.

Over the entire study, only 3% of A. phalerata female inflorescences observed in

pasture (n=3) and 8% of forest inflorescences (n=2) aborted prior to fruit maturation.

Among A. speciosa in pasture, just one infructescence aborted before its fruits ripened

(1%) and only three in old-growth forest (6%), during 18-months of observation.

At any one time, individual palms bore multiple infructescences in different stages

of development. In pasture, A. phalerata palms individually produced between 1 and 11 mature infructescences each year (mode=1), while in, only one palm produced as many as two infructescences per year. Most forest palms produced just one infructescence per year, or none at all (mode=0). A single A. speciosa palm in pasture produced five

mature infructescences, at most, over one year (mode=1), while, individual palms in

forest produced no more than two infructescences per annum (mode=0).

A B

Figure 5-2. Comparison of mean female inflorescence and mature infructescence production per palm ±1 SE in old-growth forest and pasture January- December 2006 for A) Attalea phalerata in Acre, Brazil (n=36), and B) A. speciosa in Rondônia, Brazil (n=36). Forest and pasture were significantly different in all cases (P<0.0005) based on randomization tests (10,000 iterations).

Fruiting phenology

The majority of A. phalerata fruits began developing during the dry season (July–

August) as female inflorescences completed anthesis. Infructescences matured over an average period of 6 months (±0.17 SE, n=6) in forest and 7 months (±0.08 SE, n=75) in pasture, and the greatest availability of mature fruit coincided with the wettest months of the year (December-May) (Figure 5-3). In old-growth forest, presence of mature fruits was restricted to 5-6 months per year, while in pasture, we encountered ripe fruits year-round.

Most A. speciosa infructescences initiated development during the wet season and on average, matured over eight months in both forest (±0.08 SE, n=23) and pasture

(±0.39 SE, n=51), such that mature fruits were most abundant during the late dry and early wet seasons (Figure 5-4). In old-growth forest, infructescences with ripe fruit were present during most months of the year except between February and May, while in pastures we observed mature fruits year-round.

2006 2007

Figure 5-3. Monthly fruiting intensity of Attalea phalerata (with mature fruit) and monthly rainfall from January 2006 to December 2007 (INMET 2008) in pastures and old-growth forests in eastern Acre, Brazil.

2006 2007 Figure 5-4. Monthly fruiting intensity of Attalea speciosa (with mature fruit) and monthly rainfall from January 2006 to June 2007 in pastures and old-growth forests in western Rondônia, Brazil.

100

Productivity and Oil Estimates

Fruit morphology and yields

Infructescence size (number of fruits per fruit bunch) varied more in A. phalerata than in A. speciosa (Table 5-1). We did not capture intact mature A. phalerata infructescences because of field visit timing and thus were unable to quantify forest infructescence sizes for this species. Nonetheless, we did collect mature forest fruits from infructescences in the process of fruit fall. From personal observation of young infructescences, we found no reason to suspect substantial differences in fruit numbers between environments. Individual A. phalerata fruits, as well as their mesocarp and endosperm components, were larger in forest than pasture (P<0.0001). Mesocarp oil content was lower in forest than in pasture (although highly variable), while endosperm oil content was nearly identical between the two environments (Table 5-1). A. speciosa fruits displayed few differences between the two environments, although endosperm oil content was slightly higher in pasture than in forest (Table 5-1).

101

Table 5-1. Mean (± SE) infructescence and fruit characteristics based on seven Attalea phalerata and ten Attalea speciosa fruit bunches and fruit sub-samples (n=30) randomly selected from each infructescence. Lipids (oil) content is expressed as % of dry weight and based on composite samples from each study site. Differences between forest and pasture were tested using a randomization test (10,000 iterations) for number of infructescences per palm, and Satterthwaite’s adjusted t-test, or non-parametric Mann-Whitney-Wilcoxon test (np) for all other variables.

n Old-growth forest n Pasture P-value Attalea phalerata No. of mature Infructescences per palm 36 0.5 (0.06) 36 1.7 (0.15) <0.0001 (2006) No. of fruits per 0 No data 4 569 (227) — Infructescencea

Fruit mass (g) 2 67.48 (1.99) 5 55.47 (7.36) <0.0001

Mesocarp (dry)(g) 2 6.16 (0.94) 5 4.65 (1.40) <0.0001 (np) % lipidsa 2 38.0 (15.1) 2 54.5 (19.1) —

Endosperm (dry)(g) per fruit 2 3.53 (1.09) 5 1.98 (0.20) <0.0001 (np) % lipidsa 2 67.9 (1.7) 2 67.2 (2.1) — Attalea speciosa No. of mature Infructescences per palm in 36 0.7 (0.06) 36 1.6 (0.43) <0.0001 2006 No. of fruits per 2 256 (46) 6 296 (74) — Infructescencea

Fruit mass (g) 4 192.38 ( 5.81) 6 201.06 (15.19) 0.2703

Endosperm (dry)(g) per fruit 4 6.92 (1.61) 6 6.90 (0.55) 0.1563 (np) % lipidsa 2 65.4 (1.4) 3 68.6 (3.9) — a Number of fruits per infructescence and percent lipids (%) represented by means ±SD.

Oil production estimates

A. phalerata oil yields per infructescence were, on average, greater in forest than in pasture (Table 5-2). Conversely, despite larger mesocarp and endosperm mass in old-growth forest, A. phalerata growing in pastures still out-produced palms growing in

102

forests in terms of individual annual oil yields due to much higher average infructescence frequency in pasture. The same was true for A. speciosa but to a lesser

degree. For both species, estimated mean oil production from pasture palms was more

than double the production levels from forest palms.

Table 5-2. Estimated Attalea phalerata and Attalea speciosa oil production per infructescence and annually per palm in old-growth forests and pastures in Southwestern Amazonia. Using mean values from Table 5-1, we calculated Liters of oil based on the formula below (*), and then divided by the specific density for each species (0.9231 and 0.868, respectively).

Attalea phalerata Attalea speciosa Old-growth forest Pasture Old-growth forest Pasture

Oil per infructescence (L) 2.92 2.39 1.32 1.62

Oil per palm yr-1 (L) 1.17 4.55 0.93 2.59 * OIL PALM-1 (L) = {(No. of infructescences palm-1) * (no. of fruits infructescence-1) * [(mesocarp mass fruit-1 (g) * % oil in mesocarp) + (endosperm mass fruit-1 (g) * % oil in endosperm)]} /1000.

Between study sites, population level variation in oil yields reflected palm level

differences, with pasture populations generally supplying more oil than forest

populations (except for A. speciosa study site P4; Figure 5-5). A. phalerata mesocarp

oil yields varied more between forest and pasture than endosperm oil yields, and A.

speciosa endosperm in two of the three pasture sites demonstrated dramatically greater

yields than in forest (Figure 5-5).

103

A B

Figure 5-5. Estimated 2006 oil yields (L) per 12 palms from A) A. phalerata endosperm (kernel) and mesocarp (pulp) and B) A. speciosa endosperm at each study site in Acre and Rondônia, Brazil, respectively. F1-F6 are old-growth forest sites and P1-P6 are pasture sites. Yield calculations are based on site-level averages using the formula from Table 5-2.

Discussion

Our aim in this paper was to explore the possibility of exploiting native oil palms for biofuel production in Amazonia. To do this, we compared Attalea phalerata and A. speciosa fruiting phenology, productivity and oil yields between old-growth forests and pastures. Both Attalea species exhibited promise as a source of oil for use as biofuel.

Pasture Palms Out-Produced Forest Palms

With regards to oil yields, pasture palms overwhelmingly out-produced forest palms. Even though A. phalerata bore larger fruits and yielded more oil per infructescence in forest than in pasture, annual oil yields per palm were substantially lower in forest than in pasture due to relatively low fruiting frequencies. And, while A. speciosa palms produced similar amounts of fruit and oil across both environments, pasture palms also out-produced forest palms for the same reason. In a single year, an average A. phalerata palm produced more than three times as many infructescences in

104

pasture than in forest, and A. speciosa produced more than twice as many. Peak

availability of ripened fruits in each Attalea species fell around the same time regardless of environment, but continuous production of flowers and fruits by pasture palms bolstered population-level fruit production, while reproductive activity in forests was largely seasonal with isolated fruiting peaks and consecutive months of inactivity

(Figures 5-3 and 5-4).

Greater fecundity in open pastures can be attributed to various factors, including changes in pollination regimes (Burd 1994; Herrerías-Diego et al. 2006), increased resource availability, or release from environmental stresses and competition from other trees (Aldrich and Hamrick 1998; Schroth et al. 2004) Attalea palms exploit both insect and wind pollination to fertilize female flowers (Henderson et al. 2000). Under open pastures conditions, wind pollination should augment pollen availability compared to forest, where dense vegetation obstructs wind (Anderson et al. 1988). Nonetheless, in

Chapter 3 we reported that the principle bottleneck to fruit production in both environments was successful development of female inflorescences to anthesis, and not pollination. Female inflorescences that blossomed had a very high pollination success rate: successful fruit set from female inflorescences was very similar between the two environments (< 90%).

Stark differences in resource availability between pasture and forest likely drove much of the differences in Attalea fecundity. In pasture, less competition for resources, like light, water and nutrients, likely boosted fruit production. Access to sunlight is positively related to female inflorescence and infructescence production in both A. phalerata and A. speciosa, giving a clear advantage to pasture palms (see Chapter 3).

105

In the eastern Amazon, Anderson et al. (1991) reported a large increase (almost 100%) in resource allocation towards A. speciosa reproductive structures in pasture compared with forest. Certain “super-producers” in pasture developed as many as 11 mature infructescences during a single year (A. phalerata), but such high production rates were rare. The best A. speciosa individual in pasture produced only five mature infructescences in a single year. Most palms in pasture, regardless of species, produced two or fewer infructescences per annum, and in forest, infructescence frequencies were even lower and never more than one (A. phalerata) or two (A. speciosa). Palms without infructescences were common, especially in forest, where approximately 50% of palms from each species produced no infructescences in 2006.

Voeks (1988) similarly recorded very low fruit production in adult Attalea funifera palms growing in forests of northeastern Brazil; only seven of the 86 palms produced viable fruit during 18 months of observations.

We speculate that the noticeably lower A. phalerata infructescence peaks and overall inferior fruit production in 2007, compared with 2006 (Figure 5-3), was triggered by an severe drought in 2005, when precipitation dropped more than 60% beneath normal levels during the driest part of the year (INMET 2008). Increased water stress likely provoked greater than usual bud abortion rates in the study populations (Hartley

1977), leading to lower inflorescence frequency the following year (2006) and a subsequent decline in mature infructescences during 2007. Kainer et al. (2007) observed a similar drop in Brazil nut (Bertholletia excelsa) fruit production in Acre following the 2005 drought.

106

Attalea Species Comparison

In comparing the two Attalea palm species, A. phalerata was more productive than

A. speciosa, even though the former suspended infructescence production for a longer period each year (6 months) than A. speciosa (2-3 months). Oil contributions from both

the endosperm and the mesocarp were responsible for greater total oil yields from A.

phalerata, whereas A. speciosa only produced oil from its endosperm. A. phalerata also

exhibited wider differences than A. speciosa in infructescence frequencies, fruit

characteristics and estimated oil yields between forest and pasture (Table 5-1). A. speciosa fruit biometry and oil content were statistically indistinguishable between the two environments. These disparities may be partially explained by each species’ natural crown position within the forest canopy. A. phalerata grows beneath the upper forest canopy, while A. speciosa reaches into the upper canopy, sometimes as an emergent. We speculate that forest conversion to pasture causes a more dramatic change in light resources for A. phalerata than for A. speciosa, stimulating a stronger

response in productivity of the former.

Variation, Predictability and Uncertainty

Broad variation in Attalea fruiting patterns at all levels of analysis—oil content, fruit

biometry, fruit bunch size and infructescence frequency—resulted in inconsistent oil

yields from one palm to another, as well as between environments (see Chapter 3). In

A. phalerata the number of fruits per infructescence, the amount of mesocarp per fruit,

and mesocarp oil contents each varied as much as four-fold among individuals. A.

speciosa exhibited more consistent infructescence and fruit characteristics between

forest and pasture, although endosperm mass oscillated by a magnitude of four.

Variation in reproductive output is naturally broad among wild species populations

107

(Freese 1997; Schroth et al. 2004). Climate, disease, herbivory, light, water and

nutrient availability, soils, tree size and age, disturbance history, and genetic factors,

can all produce variability in plant reproduction (Kainer et al. 2007).

Whether in pasture or forest, Attalea palms, especially A. phalerata, also exhibited uneven distributions across the landscape, making accurate estimates of yields per hectare particularly difficult. Densities changed quickly over short distances, with some pastures void of palms and others with dense palm coverage or dotted with dense clumps of multiple palms. Near our study sites in Rondônia, A. speciosa occurs at densities of 10-15 reproductive individuals per hectare in both forests and pastures, but in other parts of Rondônia, Moret (2007) reported average densities of 37 palms per hectare. Moraes (1996) reported dense A. phalerata stands of at least 20 reproductive individuals per hectare in savanna and transitional forest areas of Bolivia, but in eastern

Acre, A. phalerata palms normally occur at densities around 10 reproductive individuals per hectare or less in pastures. In old-growth forests in Acre, A. phalerata densities

ranged from none to as many as 35 reproductive palms per hectare (M. Silveira

unpublished data). High densities of A. phalerata were rarely sustained over large

areas, although we repeatedly observed predominance of this species within young

secondary forest (Carvalho et al. in press). Although not studied here, these habitats

represent another promising source of Attalea oil.

Although uneven distributions coupled with observed variation in individual and

population-level fruit production made oil yields somewhat unpredictable at a broader

scale, it strongly suggests management interventions could increase yields. For

example, protection and tending of new recruits, pruning the forest canopy to increase

108

access to light, and strategic planting of seeds from high producers could improve

Attalea oil yields. After planting both species take 7-8 years to reach reproductive

maturity (Anderson 1983; Moraes 1996). Still, one of the benefits of the extractive

system is no need for greater inputs to obtain satisfactory yields. If rural inhabitants

adopted a decentralized, small-scale model for energy production based on native palm

oils, those familiar with the landscape and its forests could concentrate fruit harvests on

more productive palms and in areas with higher palm densities. Producers can also

monitor female inflorescence production. Because fruit abortion was relatively rare in

both species, female inflorescence production can serve as a good early predictor of

fruiting patterns and oil yields.

What can Attalea Contribute to Energy Portfolios in Rural Amazonia?

Small producers in rural Amazonia use diesel fuel for electricity generation and

transportation to sell their products and buy goods in town, transport children to school,

and attend to medical emergencies. In the Amazon, diesel can be as much as three

times more expensive than the national average (Da Costa 2004). In isolated rural

communities, energy production from locally available plant resources promises greater

independence and improved livelihoods, while assigning greater value to standing

forests and diversifying pasture systems already under use. Substitution of petroleum-

based diesel with vegetable oils and biodiesel will also reduce greenhouse gas

emissions and air and water pollutants. Two recent oil spills from boats transporting

diesel to Acre polluted local waterways in 2008 and 2009 and attest to the negative

consequences of mineral diesel distribution in the Amazon (O Globo 2009).

Large amounts of oil can be extracted from both A. phalerata and A. speciosa fruits (Figure 5-5) to support electricity expansion and transportation in rural Amazonia.

109

Small producers can extract about 800 L of palm oil annually from approximately 100 A. phalerata infructescences harvested in forest (~100 palms) plus 200 A. phalerata infructescences harvested in pasture (~100 palms). That is enough to supply a community of ten small homes with electricity for 2-3 lamps, a television, and a freezer for four hours each night, for about one year. If each of ten households harvested this amount of oil from 10 pasture palms and 10 forest palms, collectively, they would have enough oil to generate electricity year-round for the community (J. Evandro Santos

Lima, pers. comm.).

In Rondônia, Brazil, communities from the Rio Ouro Preto Extractivist Reserve

(201,334 ha) installed a slightly modified diesel mill that generates electricity directly from oil extracted from A. speciosa endosperm (www.gpers.unir.br). Mean annual oil consumption for this community with approximately 500 inhabitants was 4313 L (Moret

2007). Using their regional estimate of A. speciosa palm densities (37 reproductive individuals ha-1), an estimated 125 ha of forest or 45 ha of pasture would meet this

demand.

Endosperm oil from both A. phalerata and A. speciosa can be use directly in diesel

engines, or with slight adjustments. An iodine index of less than 25 is required for long

term use of pure vegetable oil in unmodified diesel engines (Calais and Clark 1999 in

Ghen 2005). Ghen (2005) reported low iodine values for A. phalerata (19.96), and A. speciosa has a value of 16.83 (Pesce 1985). In contrast, African oil palm’s (Elaeis guineensis) pure oil iodine value ranges from 35 to 61 (Espadafor et al. 2009). A. phalerata mesocarp oil, meanwhile, has an iodine value of 51, which precludes its effective use directly as biofuel, due to risks of polymerization when combusted. This

110

obstacle can be overcome by heating the vegetable oil prior to injection into a diesel

engine (GPERS 2007).

Harvest Considerations: Advantages and Disadvantages

Socio-economic and technical studies are still needed to complement the ecological information presented here to further assess the feasibility and practicality of harnessing native palm oil species for biofuel production in rural Amazonian communities. Fruit processing technologies will need to address a range of issues, including more efficient mesocarp and endosperm removal techniques and oil extraction. Difficulty in cracking the hard woody endocarp of Attalea fruits to remove oil- bearing seeds presents a considerable challenge to oil extraction. Local populations in the Brazilian state of Maranhão practice an effective, albeit dangerous, manual technique for opening A. speciosa endocarps to extract kernels using a short-handled axe (Anderson et al. 1991). With greater attention to the specific challenges of kernel extraction, more modern extraction techniques can greatly improve efficiency and safety

(Lee 1930, Teixeira 2008). Once the pulp and kernel are removed, oil can be extracted using a manual press. Compared with the chemical extraction process we utilized, pressing yields somewhat relatively less oil (Anelise Regiana, pers. comm.) but represents a more feasible solution for rural populations interested in extracting oil from palms fruits. Research into more effective and efficient pressing techniques is needed.

Investments in technical assistance and capacity building will also be necessary to establish micro-processing mills to use native palm oil to generate electricity directly or convert it to biodiesel (Teixeira 2008). Recently, Teixeira and Carvalho (2007) proposed a promising steam cogeneration system to generate electricity from A.

speciosa oil. Within remote rural settings, communities could benefit from smaller-

111

scale, less costly processing facilities that can be strategically located near oil sources

(Ansley et al. 2010). To reduce costs and concentrate maintenance needs, landowners could bring their Attalea fruit crop to a central location with pressing equipment for oil extraction and go home with oil to supply their daily energy and fuel needs. A similar system is locally used to process rice using a rice sheller mill centrally available to surrounding residents.

Opposite seasonal availability of mature fruits differentiates A. phalerata from A. speciosa, and presents specific challenges in more remote areas. In A. phalerata, which peaks during the wet season, the fruits’ mesocarp is more perishable than the endosperm and quickly turns rancid, limiting the utility of mesocarp oil in isolated communities, unless harvesters can extract oil from the pulp immediately. Palm kernels, on the other hand, can be stored for several months. Collection of A. speciosa fruits during the dry season would facilitate storage, but is more likely to compete with other labor-intensive small producer activities, such as clearing and burning forest and planting crops (August-October).

The year-round availability of mature infructescences in pastures confers an advantage over forests, where palms suspend fruit production for 2-3 months or more each year. Although such results may provide incentive to clear forest and install more pasture, these signals are less likely to affect small rural producers who would benefit most from locally-harvested palm oils. Within Amazonia, many small-scale rural communities are not trying to maximize one output from their landholdings, but rather diversify their production systems for multiple products and services. Furthermore,

Brazilian forest laws deter many small landholders from cutting more forest (Fearnside

112

2003), and internally negotiated limits on forest clearing in extractive reserves curtail

large-scale conversion of forest to pasture. Native palm oil could join the suite of other

timber and non-timber forest resources garnered from intact forests and the larger

regional landscape, as well as encourage diversification into silvo-pastoral production

systems.

Local-based consumption versus market sale

This work was not designed to address the question of whether oil extraction from

native palms can make a significant contribution to national or even state-wide biofuel supplies. Industrial biofuel production would probably benefit little from naturally occurring populations of oil-producing palms in forests and pastures, because large- scale, genetically improved crops and plantation level production would easily out- compete native species in most cases. In Brazil, for example, African oil palm (Elaeis guineensis) plantations annually produce 3500-5000 kg of oil per hectare, and soybeans, 350-450 kg per hectare (Da Costa 2004), but such high productivity also brings trade-offs—forest conversion, loss of biodiversity, high costs of chemical inputs, planting and maintenance, infrastructure, technology, research, and development.

Avoidance of these costs at least partially compensates for the lower yields derived from naturally occurring native oleaginous species as a source of biofuel (Ansley et al.

2010). And, in rural areas, oil extracted from native palms complements other income sources garnered from the forest or within an agro-pastoral system. Besides the fuel potential of mesocarp and endosperm oils, the woody endocarp of Attalea fruits can be used to make charcoal (Emmerich and Luengo 1996), while residues left over from peeling the fruit and pressing the pulp and kernel can supplement animal feed (Pinheiro and Frazão 1995), adding to the benefits garnered from these palms.

113

Conclusion

In rural agricultural settlements distant from city centers, as well as in isolated forest and riverine communities, naturally occurring populations of oleaginous plants represent a largely underexploited energy resource and can serve as an important and immediate source of fuel for electrification and transportation (e.g., river boats, small trucks, or even small airplanes). In many areas of rural Amazonia, rivers serve as the principal mode of transportation and navigable road networks are sparse. Use of oil from the fruits of native plants can promote greater energy self-sufficiency for remote communities, greatly reduce fuel costs, and even generate jobs and income at the small scale. A small-scale, decentralized approach to biofuel production that responds directly to regional needs and draws upon local renewable forest resources can bring much needed socio-economic, environmental, and energy benefits to rural communities, as well as enhance the competitive economic value of standing forest versus conversion to other uses and diversify current pasture management systems.

114

CHAPTER 6 CONCLUSION

The aim in this research was to assess the potential of Attalea phalerata and

Attalea speciosa palms as local sources of biofuel in the Southwestern Amazon. My

evaluation was primarily based on an ecological study to hone in on the key aspects

related to fruit production, while comparing two habitats where they occur most—forest

and pasture.

In Chapter 2, I report on a novel finding of color polymorphism in Attalea phalerata flowers in Acre, Brazil. Normally, these palms exhibit yellow or cream-colored flowers, but during 24 months of phenology observations I noted that over half of the staminate inflorescences observed in anthesis displayed flowers that were purple in color, ranging from magenta to deep violet. No previously published records of variation in A. phalerata flower color exist. I speculated that the purple flowers could be an indication of environmental stress, such as short-term cold spells, a dryer than usual dry season, or even disease, herbivory or parasitism, that provokes an increase in anthocyanins in the palms, leading to purple flowers in this species.

In Chapter 3, I found that Attalea’s unique reproductive biology introduced high variability into fruit and oil production at the tree level, but also created flexibility in sexual expression that led to greatly increased inflorescence and infructescence production in pastures, where resource availability was high. Overall, pasture palms out-produced forest palms, in terms of both inflorescence and infructescence production, even though inflorescence sex ratios remained predominantly male-biased.

Greater light availability and larger leaf crowns were associated with increased productivity, as well as increased female function in both A. phalerata and A. speciosa.

115

Results from Chapter 4, showed that A. phalerata palms quickly recuperated after forest clearing and burning at our study site. Prior to above-ground stem elongation, the protection of A. phalerata’s apical meristem belowground benefited palm’s survival and

persistence in disturbed landscapes. Just 14 months after deforestation and burning,

our analyses showed no overall effect of slash-and-burn on palm population densities or

size-class relative abundances, except for the adults, which the farmer had cut. Pre-

adult palms persisted in the slash-and-burn area through resprouting of new leaves, and

there was relatively little germination. Growth of young A. phalerata palms was

accelerated in the cleared area compared with intact forest, leaving the smallest size

class nearly empty due to low seedling recruitment over the 14 month study period, but

strong survival and persistence of post-seedling size classes will ensure A. phalerata

population recovery at our study site. Attalea palms’ remarkable capacity to survive

forest clearing and fire, and their tendency to dominate disturbed areas and

regenerating secondary forests means that this species will prevail across the

landscape and across diverse habitats—pastures, old-growth forest, degraded lands,

and secondary forests—for years to come.

Finally, Chapter 5 demonstrated that both A. phalerata and A. speciosa represent

a promising local energy source for rural areas of Amazonia. If adopted, the small-scale

use of these native palm oils for both electricity generation and transportation in remote

communities and rural settlements can improve livelihoods, promote energy self-

sufficiency, reduce pollution originating from petroleum-based fuels, help increase the value of standing forests, and diversify current pasture management systems.

116

This work increases our knowledge of Attalea ecology and presents evidence of strong potential for A. phalerata and A. speciosa as a local source of biofuel in rural

Amazonian communities. Still, a myriad of questions and directions for future research emerged from this study. For example, the ecological consequences of harvesting

Attalea fruits for oil will need to be addressed (Ticktin 2004). We know remarkably little about the long-term productivity of Attalea palms, and how long they continue to reproduce before entering senescence. Investigations into possible management practices to increase Attalea fruit and oil yields are needed (e.g., enrichment planting could augment palm densities and increase oil yields), and more experimental research is needed to understand the mechanisms as well as the biotic and abiotic factors affecting sex determination in Attalea. Finally, an analysis of Attalea palm densities using remotely sensed images could greatly expand the ability to locate high oil production areas and better estimate oil production on a broader scale. These recommendations include a few of the possible future research directions that emerged from this work.

117

LIST OF REFERENCES

Abramovitz JN, Mattoon AT (1999) Reorienting the forest products economy. In State of the world (Starke L, ed) pp. 60-77,206-213. WW Norton and Co., New York

Achard F, Eva HD, Stibig H-J, Mayaux P, Gallego J, Richards T, Malingreau J-P. (2002) Determination of deforestation rates of the world’s humid tropical forests. Science 297:999-1002

Aide TM, Cavelier J (1994) Barriers to lowland tropical forest restoration in the Sierra Nevada de Santa Marta, Colombia. Restoration Ecology 2:219-229

Aizen MA, Feinsinger P (1994) Habitat fragmentation, pollination, and plant reproduction in a Chaco dry forest. Ecology 75:330-351

Aizen MA, Ashworth L, Galetto L (2002) Reproductive succession fragmented habitats: Do compatibility systems and pollination specialization matter? Journal of Vegetation Science 13:885-892

Aldrich PR, Hamrick JL (1998) Reproductive dominance of pasture trees in a fragmented tropical forest mosaic. Science 281:103-105

Alonso C, Mutikainen P, Herrera CM (2007) Ecological context of breeding system variation: Sex, size and pollination in a (predominantly) gynodioecious shrub. Annals of Botany 100:1547-1556

Anderson, AB (1983) The biology of Orbignya martiana (Palmae), a tropical dry forest dominant in Brazil. PhD dissertation, University of Florida, Gainesville

Anderson AB, Anderson ES (1983) People and the palm forest: biology and utilization of babassu palms in Maranhão, Brazil. Final Report to Forest Service, U.S. Department of Agriculture, Washington, D.C

Anderson AB, Overal WL, Henderson A (1988) Pollination ecology of a forest- dominant palm (Orbignya phalerata Mart.) in Northern Brazil. Biotropica 20:192-205

Anderson AB, May PH, Balick MJ (1991) The Subsidy from nature: palm forests, peasantry, and development on an Amazon frontier. Columbia University Press, New York

Ansley RJ, Mirik M, Castellano MJ (2010) Structural biomass partitioning in regrowth and undisturbed mesquite (Prosopis glandulosa): Implications for bioenergy uses. GCB Bioenergy, doi:10.1111/j.1757-1707.2010.01036.x

118

Anten NPR, Ackerly DD (2001) Canopy-level photosynthetic compensation after defoliation in a tropical understorey palm. Functional Ecology 15:252-262

Anten NPR, Martinez-Ramos M, Ackerly DD (2003) Compensatory growth in a tropical understory palm subjected to repeated defoliation events. Ecology 84:2905-2918

Armbruster WS (2002) Can indirect selection and genetic context contribute to trait diversification? A transition-probability study of blossom-colour evolution in two genera. Journal of Evolutionary Biology 15:468-486

Augspurger C (1984) Seedling survival of tropical tree species: interactions of dispersal distance, light-gaps, and pathogens. Ecology 65:1705-1712

Baily L (1947) Palmae indigenae Trinitenses et Tobagenses. Gentes Herb. 7:353-445

Balcazar Vargas MP (2004) Demography of Attalea butyracea in the Bolivian Amazon. MSc thesis, Utrecht University, Netherlands

Balée W (1988) Indigenous adaptation to Amazonian palm forests. Principes 32:47-54

Balée W (1989) The culture of Amazonia forests. Advances in Economic Botany 7:1-21

Balick MJ (1979) Amazonian oil palms of promise: A survey. Economic Botany 33:11-28

Balick M, Beck H (1990) Useful palms of the world: a synoptic bibliography. Columbia University Press, New York

Barot S, Gignoux J, Vuattoux R, Legendre S (2000) Demography of a savanna palm tree in Ivory Coast (Lamto): population persistence and life-history. Journal of Tropical Ecology 16:637-655

Barrett SCH (2010) Understanding plant reproductive diversity. Philosophical Transactions of the Royal Society B 365:99-109

Bathke AC, Harrar SW, Madden LV (2008) How to compare small multivariate samples using nonparametric tests. Computational Statistics and Data Analysis 52:4951-4965

Bechtold WA (2003) Crown position and light exposure classification – an alternative to field-assigned crown class. Northern Journal of Applied Forestry 20:154-160

119

Bierzychudek P (1984) Determinants of gender in Jack-in-the-Pulpit: The influence of plant size and reproductive history. Oecologia 65:14-18

Bjorholm S, Svenning JC, Skov F, Balslev H (2005) Environmental and spatial controls of palm (Arecaceae) species richness across the Americas. Global Ecology and Biogeography 14:423-429

Blanco MA, Martén Rodríguez S (2007) The stained-glass palm, Geonoma epetiolata. Palms 51:139-146

Boll T, Svenning JC, Vormisto J (2005) spatial distribution and environmental preferences of the piassaba palm Aphandra natalia (Arecaceae) along the Pastaza and Urituyacu rivers in Peru. Forest Ecology and Management 213:175-183

Bond WJ, Midgley JJ (2003) The evolutionary ecology of sprouting in woody plants. International Journal of Plant Sciences 164:S103-S114

Bondar G (1964) Palmeiras do Brasil. Instituto de Botânica, Secretária da Agricultura do Estado de S. Paulo, São Paulo, Brasil

Bonguebartelsman M, Phillips DA (1995) Nitrogen stress regulates gene- expression on enzymes in the flavinoid biosynthetic pathway of tomato. Plant Physiology and Biochemistry 33:539-546

Borchert R, Meyer SA, Felger RS, Porter-Bolland L (2004) Environmental control of flowering periodicity in Costa Rican and Mexican tropical dry forests. Global Ecology and Biogeography 13:409-425

Brandon H (2005) A study in contrasts: Brazil sets pace on alternate fuels. Southeast Farm Press, Nov 16

Brown KE (1976) Ecological studies of the Cabbage Palm Sabal palmetto. III. Seed germination and seedling establishment. Principes 20:98-115

Bruna EM, Nogueira Ribeiro MB (2005) The compensatory responses of an understory herb to experimental damage are habitat-dependent. American Journal of Botany 92(12):2101-2106

Bullock SH (1980) Demography of an undergrowth palm in littoral Cameroon. Biotropica 12:247-255

Bullock SH, Bawa KS (1981) Sexual dimorphism and the annual flowering pattern in Jacaratia dolichaula (D. Smith) Woodson (Caricaceae) in a Costa Rican rain forest. Ecology 62:1494-1504

Burd M (1994) Bateman’s principle and plant reproduction: the role of pollen limitation in fruit and seed set. Botanical Review 60: 83-139

120

Burd M, Allen TFH (1988) Sexual allocation strategy in wind-pollinated plants. Evolution 42:403-407

Buschbacher R, Uhl C, Serrao EAS (1988) Abandoned pastures in eastern Amazonia. II. Nutrient stocks in the soil and vegetation. Journal of Ecology 76:682-699

Bush MB, Miller MC, Oliveira PE, Colinvaux PA (2000) Two histories of environmental change and human disturbance in eastern lowland Amazonia. The Holocene 10:543-553

Bush MB, Silman MR, Toledo MB, Listopad C, Gosling WD, Williams C, Oliveira PE, Krisel C (2007) Holocene fire and occupation in Amazonia: records from two lake districts. Philosophical Transactions of the Royal Society B 362:209-218

Cajas-Giron YS, Sinclair FL (2001) Characterization of multistrata silvopastoral systems on seasonally dry pastures in the Caribbean Region of Colombia. Agroforestry Systems 53:215-225

Calvin M (1983) New sources for fuel and materials. Science 219:24-26

Camargo JTC, Kapos V (1995) Edge effects on soil moisture and microclimate in Central Amazonian forest. Journal of Tropical Ecology 11: 205-221

Campbell DR (2000) Experimental tests of sex-allocation theory in plants. Trends in Ecology and Evolution 15:227-231

Campos MT, Ehringhaus C (2003) Plant virtues are in the eyes of the beholders: a comparison of known palm use among indigenous and folk communities of southwestern Amazonia. Economic Botany 57:324-344

Capers RS, Chazdon RL, Brenes AR, Alvarado BV (2005) Successional dynamics of woody seedling communities in wet tropical secondary forests. Journal of Ecology 93:1071-1084

Caplat P, Anund M (2009) Effects of disturbance frequency, species traits and resprouting on directional succession in an individual-based model of forest dynamics. Journal of Ecology 97:1028-1036

Cardoso da Silva, JM, Uhl F, Murria G (1996) Plant succession landscape management and the ecology of frugivorous birds in abandoned Amazonian pastures. Conservation Biology 10:491-503

Carriere SM, Letourmy P, McKey DB (2002) Effects of remnant trees in fallows on diversity and structure of forest regrowth in a slash-and-burn agricultural system in southern Cameroon. Journal of Tropical Ecology 18:375-396

121

Carvalho AL, Ferreira EJL, Tucker Lima JM (in press) Comparações florísticas e estruturais entre comunidades de palmeiras em áreas de floresta primária e secundária da Área de Proteção Ambiental Raimundo Irineu Serra – Rio Branco, Acre, Brasil. Acta Amazonica

Cascante A, Quesada M, Lobo JJ, Fuchs EA (2002) Effects of dry tropical forest fragmentation on the reproductive success and genetic structure of the tree Samanea saman. Conservation Biology 16:137-147

Charles-Dominique P, Chave J, Dubois MA, de Granville JJ, Riera B, Vezzoli C (2003) Colonization front of the understorey palm Astrocaryum sciophilum in a pristine rain forest of French Guiana. Global Ecology and Biogeography 12:237-248

Charnov EL (1982) The theory of sex allocation. Princeton University Press, Princeton, NJ

Charnov EL, Bull J (1977) When is sex environmentally determined? Nature 266: 828-830

Chazdon R (1991) Effects of leaf and ramet removal on growth and reproduction of Geonoma congesta, a clonal understorey palm. Journal of Ecology 79(4):1137-1146

Chazdon R (1986) Light variation and carbon gain in rain forest understory palms. Journal of Ecology 74:995-1012

Cierjacks A, Salgado S, Wesche K, Hensen I (2008) Post-fire population dynamics of two tree species in high-altitude Polylepis forests of central Ecuador. Biotropica 40:176-182

Clark DA, Clark DB (1987) Temporal and environmental patterns of reproduction in Zamia skinneri, a tropical rain forest cycad. Journal of Ecology 75:135- 149

Clarkson JR, Clifford HT (1987) Germination of Jedda multicaulis J.R. Clarkson (Thymelaeaceae). An example of cryptogeal germination in the Australian flora. Australian Journal of Botany 35:715-720

Close DC, Beadle CL (2003) The ecophysiology of foliar anthocyanin. The Botanical Review 69:149-161

Cochrane MA (2003) Fire science for rainforests. Nature 421:913-919

Coello J, Acosta F, Velásquez J (2000) Options for small-scale biodiesel production to self-supply the energy needs of isolated communities in Amazonia. Boiling Point 56:12-15

122

Condon MA, Gilbert LE (1988) Sex expression of Gurania and Psiguria (Cucurbitaceae): Neotropical vines that change sex. American Journal of Botany 75:875-884

Consiglio TK, Bourne GR (2001) Pollination and breeding system of a neotropical palm Astrocaryum vulgare in Guyana: A test of the predictability of syndromes. Journal of Tropical Ecology 17:577-592

Corner EJH, (1966) The natural history of palms. University of California Press, Berkeley, CA

Cozza J (2008) Sex expression in a rainforest understory herb, Begonia urophylla. PhD dissertation, University of Miami, Coral Gables, FL

Cruden RW (1988) Temporal dioecism: systematic breadth, associated traits, and temporal patterns. Botanical Gazette 149:1-15

Cunningham SA (1995) Ecological constraints on fruit initiation by Calyptrogyne ghiesbreghtiana (Arecaceae): floral herbivory, pollen availability and visitation by pollinating bats. American Journal of Botany 82:1527-1536

Cunningham SA (1996) Pollen supply limits fruit initiation by a rain forest understory palm. Journal of Ecology 84:185-194

Cunningham SA (1997) The effect of light environment, leaf area, and stored carbohydrates on inflorescence production by a rain forest understory palm. Oecologia 11:36-44

Cunningham SA (2000) What determines the number of seed produced in a flowering event? A case study of Calyptrogyne ghiesbreghtiana (Arecaceae). Australian Journal of Botany 28:659-665

Da Costa RC (2004) Potential for producing bio-fuel in the Amazon deforested areas. Biomass and Bioenergy 26:405-415

Dauvergne P, Neville KJ (2009) The changing North-South and South-South political economy of biofuels. Third World Quarterly 30:1087-1102

Delph, LF (1986) Factors regulating fruit and seed production in the desert annual Lesquerella gordonii. Oecologia 69:471-476

Delph LF, Wolf DE (2005) Evolutionary consequences of gender plasticity in genetically dimorphic breeding systems. New Phytologist 166:119-128

Del Tredici P (2001) Sprouting in temperate trees: A morphological and ecological review. Botanical Review 67:121-140

123

Demirbas A (2009) Progress and recent trends in biodiesel fuels. Energy Conservation and Management 50:14-34

Denevan WM (1992) The pristine myth: the landscape of the Americas in 1492. Annals of the American Association of Geographers 82:369-385

DeSteven D, Windsor DM, Putz FE, de Leon B (1987) Vegetative and reproductive phenologies of a palm assemblage in Panama. Biotropica 19:342-356

Dunn RR (2004) Recovery of faunal communities during tropical forest regeneration. Conservation Biology 18:302-309

Elmqvist T, Wall M, Berggren AL Blix L, Fritioff A, Rinman U (2001) Tropical forest reorganization after cyclone and fire disturbance in Samoa: remnant trees as biological legacies. Conservation Ecology 5, http://www.consecol.org/vol5/iss2/art10/

Emlen ST, Oring LW (1977) Ecology, sexual selection, and the evolution of mating systems. Science 197:215-223

Emmerich FG, Luengo CA (1996) Babassu charcoal: A sulfurless renewable thermo-reducing feedstock for steelmaking. Biomass and Bioenergy 10:41-44

Enright NJ, Watson AD (1992) Population dynamics of the Nikau palm, Rhopalostylis sapida (Wendl. et Drude), in a temperate forest remnant near Aukland, New Zealand. New Zealand Journal of Botany 30:29-43

Escobar JC, Lora ES, Venturini OJ, Yáñez EE, Castillo EF, Almazan O (2009) Biofuels: Environment technology and food security. Renewable and Sustainable Energy Reviews 13:1275-1287

Espadafor RJ, García MT, Villanueva JB, Gutiérrez JM (2009) The viability of pure vegetable oil as an alternative fuel for large ships. Transportation Research Part D 14:461-469

FAO (2008) The state of food and agriculture. Biofuels: prospects, risks and opportunities. Food and Agriculture Organization of the United Nations, Rome

Fargione J, Hill J, Tilmand D, Polasky S, Hawthorne P (2008) Land clearing and the biofuel carbon debt. Science 319:1235-1238

Farrell AE, Plevin RJ, Turner BT Jones AD, O’Hare M, Kammen DM (2006) Ethanol can contribute to energy and environmental goals. Science 311:506-508

124

Fava WS (2010) Attalea phalerata e Bactris glaucescens (Arecaceae, Arecoideae): fenologia e biologia reprodutiva no Pantanal de Mato Grosso do Sul. Masters thesis, Universidade Federal de Mato Grosso do Sul, Mato Grosso do Sul

Fearnside PM (2003) Deforestation control in Mato Grosso: A new model for slowing the loss of Brazil’s Amazon forest. Ambio32:343-345

Feil JP (1996) Fruit production of Attalea colenda (Arecaceae) in coastal Ecuador – an alternative oil resource? Economic Botany 5:300-309

Fenner M (1998) The phenology of growth and reproduction in plants. Perspecitves in Plant Ecology, Evolution and Systematics 1:78-91

Fineblum WL, Rausher MD (1997) Do floral pigmentation genes also influence resistance to enemies? The W locus in Ipomoea purpurea. Ecology 78:1646-1654

Fletcher RJ, Robertson BA, Evans J, Doran PJ Alavalapati JRR, Schemske DW (2010) Biodiversity conservation in the era of biofuels: Risks and opportunities. Frontiers in Ecology and the Environment, doi:10.1890/090091

Freeman DC, Klikoff LG, Harper KT (1976) Differential resource ultilization by the sexes of dioecious plants. Science 193:597-599

Freeman DC, Harper KT, Charnov EL (1980) Sex change in plants: Old and new observations and new hypotheses. Oecologia 47:222-232

Freeman DC, McArthur ED, Harper KT, Blauer AC (1981) Influence of environment on the floral sex ration on monoecious plants. Evolution 35:194-197

Freese CH (1997) Harvesting wild species: Implications for biodiversity conservation. Johns Hopkins University Press, Baltimore, MD

Galetti M, Guimaraes PR (2004) Seed dispersal of AttaleAba phalerata (Palmae) by crested caracaras (Caracara plancus) in the Pantanal and a review of frugivory by raptor. Ararajuba 12:133-135

Gehring C, Denich M, Vlek PL (2005) Resilience of secondary forest regrowth after slash-and-burn agriculture in central Amazonia. Journal of Tropical Ecology 21:519-527

Gemerden BS, Olff H, Parren MPE, Bongers F (2003) The pristine rain forest? Remnants of historical human impacts on current tree species composition and diversity. Journal of Biogeography 30:1381-1390

125

Ghen G (2005) Attalea phalerata and biodiesel: Implications for local and regional sustainability. Senior thesis, University of Hawaii, Manoa, HI

Gmunder SM, Zah R, Bhatacharjee S, Classen M, Mukherjee P, Widmer R (2009) Life cycle assessment of village electrification based on straight jatropha oil in Chhattisgarh, India. Biomass and Bioenergy 34:347-355

Green RH (1979) Sampling design and statistical methods for environmental biologists. Wiley Interscience, Chichester, England

Gregg KB (1973) Studies on the control of sex expression in the genera Cycnoches and Catasetum, subtribe Catasetinae, Orchidaceae. PhD dissertation, University of Miami, Coral Gables, FL

Groom MJ, Gray EM, Townsend PA (2008) Biofuels and biodiversity: Principles for creating better policies for biofuel production. Conservation Biology 22:602-609

Grotewold, E (2006) The genetics and biochemistry of floral pigments. Annual Review of Plant Biology 57:761-780

Grupo de Pesquisa Energia Renovável Sustentável (GPERS) (2007) Orbignya barbosiana, o babaçu: Fonte de energia para comunidades isoladas. http://www.gpers.unir.br/index.html

Guevara S, Laborde J (1993) Monitoring seed dispersal at isolated standing trees in tropical pastures: Consequences for local species availability. Vegetatio 108:319-338

Guevara S, Meave J, Moreno-Casasola P, Laborde J (1992) Floristic composition and structure of vegetation under isolated trees in Neotropical pastures. Journal of Vegetation Science 3:655-664

Guevara S, Laborde J, Sanchez-Rios G (2004) Rain forest regeneration beneath the canopy of fig trees isolated in pastures of Los Tuxtlas, Mexico. Biotropica 36:99-108

Hannan GL (1981) Flower color polymorphism and pollination biology of Platystemon californicus Benth. (Papaveraceae). American Journal of Botany 68:233-243

Harborne JB, Williams CA, Greenham J, Moyna P (1974) Distribution of charged flavones and caffeylshikimic acid in Palmae. Phytochemistry 13:1557- 1559

Harrar SW, Bathke AC (2008) Nonparametric methods for unbalanced multivariate data and many factor levels. Journal of Multivariate Analysis 99:1635-1664

126

Hartley CWS (1977) The oil palm. John Wiley and Sons Inc, New York, NY

Harvey CA, Haber WA (1999) Remnant trees and the conservation of biodiversity in Costa Rican pastures. Agroforestry Systems 44:37-68

Heckenberger MJ, Russell JC, Toney JR, Schmidt M (2007) The legacy of cultural landscapes in the Brazilian Amazon: implications for biodiversity. Philosophical Transactions of the Royal Society B 362:197-208

Heckenberger MJ, Neves EG (2009) Amazonian archaeology. Annual Review of Anthropology 38:251-266

Henderson A (2002) Evolution and ecology of palms. The New York Botanical Garden Press, Brooklyn, New York

Henderson A (1995) The palms of the Amazon. Oxford University Press, New York

Henderson A, Fischer B, Scariot A, Whitaker Pacheco MA, Pardini R (2000) Flowering phenology of a palm community in a Central Amazon forest. Brittonia 52:149-159

Henderson A, Galeano G, Bernal R (1995) Field Guide to the Palms of the Americas. Princeton University Press, Princeton, New Jersey

Henderson FM (2006) Morphology and anatomy of palm seedlings. The Botanical Review 72:273-329

Henson IE (2000) Modeling the effects of ‘haze’ on oil palm productivity and yield. Journal of Oil Palm Research 12: 123-134

Henson IE, Dolmat MT (2004) Seasonal variation in yield and developmental processes in an Oil Palm density trial on a peat soil: 1. Yield and bunch number components. Journal of Oil Palm Research 16:88-105

Herrera CM (1991) Dissecting factors responsible for individual variation in plant fecundity. Ecology 72:1436-1448

Herrera JM, Garcia D (2009) The role of remnant trees in seed dispersal through the matrix: Being alone is not always so sad. Biological Conservation 142:149-158

Herrerias-Diego Y, Quesada M, Stoner KE, Lobo JA (2006) Effects of forest fragmentation on phenological patterns and reproductive success of the tropical dry forest tree Ceiba aesculifolia. Conservation Biology 20:1111- 1120

127

Hesse AJ, Duffield GE (2000) The status and conservation of the blue-throated macaw Ara glaucogularis. Bird Conservation International 10:255-275

Hodge WH (1975) Oil-producing palms of the world—A review. Principes 19:119- 136

Holdrigde LR (1978) Life zone ecology. Tropical Science Center, San José, Costa Rica

Holl KD (1998) Do bird perching structures elevate seed rain and seedling establishment in abandoned tropical pasture? Restoration Ecology 6:253- 261

Holl KD (1999) Factors limiting tropical rain forest regeneration in abandoned pastures: seed rain, seed germination, microclimate and soil. Biotropica 31:229-242

Hooper ER, Legendre P, Condit R (2004) Factors affecting community composition of forest regeneration in deforested, abandoned land in Panama. Ecology 85:3313-3326

Inkrot D, Sattler D, Geyer C, Morawetz W (2007) Flowering and fruiting phenology of Normanbya normanbyi (W. Hill) L.H. Bailey (Arecaceae), a palm endemic to the lowland tropical rainforest of north-eastern Australia. Austral Ecology 32:21-28

Instituto Nacional de Meterologia (INMET) (2008) Federal University of Acre, Brazil. http://www.inmet.gov.br

Instituto Adolfo Lutz (1985) Normas Analiticas do Instituto Adolfo Lutz, Métodos quimicos e fisicos para análise de alimentos (2nd edition). Sao Paulo, pp 159, 189, 190

Irish EE, Nelson T (1989) Sex determination in monoecious and dioecious plants. The Plant Cell 1:737-744

Jackson G, (1974) Cryptogeal germination and other seedling adaptations to the burning of vegetation in savanna regions: the origin of the pyrophytic habit. New Phytologist 73:771-780

Janzen DH (1986) The future of tropical ecology. Annual Review of Ecology and Systematics 17:305-324

Johnson MA, Tomas WM, Guedes NMR (1997) On the Hyacinth macaw’s nesting tree: density of young manduvis around adult trees under three different management conditions in the Pantanal wetland, Brazil. Ararajuba:185-188

128

Jones LH (1997) The effects of leaf pruning and other stresses on sex determination in the Oil Palm and their representation by a computer simulation. Journal of Theoretical Biology 187:241-260

Jones KL (1947) Studiees on Ambrosia. IV. Effects of short photoperiod and temperature on sex expression. American Journal of Botany 34:371-377

Joshi AC (1939) A note on the determination of sex of plantain flowers. Journal of Genetics 38:353-356

Kahn F (1986) Life forest of Amazonian palms in relation to forest structure and dynamics. Biotropica 18:214-218

Kahn F, de Granville JJ (1992) Palms in forest ecosystems of Amazonia. Springer-Verlag, New York

Kainer KA, Wadt LHO, Staudhammer CL (2007) Explaining variation in Brazil nut fruit production. Forest Ecology and Management 250:244-255

Kainer KA, Duryea ML, Costa de Macedo N, Williams K (1998) Brazil nut seedling establishment and autecology in extractive researves of Acre, Brazil. Ecological Applications 8:397-410

Kammescheidt L (1999) Forest recovery by root suckers and above-ground sprouts after slash-and-burn agiculture, fire and logging in Paraguay and Venezuela. Journal of Troical Ecology 15:143-157

Kauffman JB (1991) Survival by sprouting following fire in tropical forests of the eastern Amazon. Biotropica 23:219-224

Kauffman JB, Cummings DL, Ward DE (1998) Fire in the Brazilian Amazon. 2. Biomass, nutrient pools and losses in cattle pastures. Oecologia 113:415- 427

Kelly BA, Fleury SG, Bouvet JM (2007) Impact of agroforestry practices on the flowering phenology of Vitellaria pardoxa in parklands in southern Mali. Agorforestry Systems 71:67-75

Kennard DK, Gould K, Putz FE, Fredericksen TS, Morales F (2002) Effect of disturbance intensity on regeneration mechanisms in a tropical dry forest. Forest Ecology and Management 162:197-208

Klinkhamer PGL, de Jong TJ, Metz H (1997) Sex and size in cosexual plants. Trends in Ecology and Evolution 12:260-265

Koh LP, Ghazoul J (2008) Biofuels, biodiversity, and people: understanding the conflicts and finding opportunities. Biological Conservation 141:2450-2460

129

Koh LP, Wilcove DS (2007) Oil palm: Disinformation enables deforestation. Trends in Ecology and Evolution 24:67-68

Korpelainen H (1998) Labile sex expression in plants. Biological Review 73:157- 180

Lamont BB, He T, Enright NJ, Krauss SL, Miller BP (2003) Anthropogenic disturbance promotes hybridization between Banksia species by altering their biology. Journal of Evolutionay Biology 16:551-557

Lazaro A, Mendez M (2007) Variation in sexual expression in a monoecious shrub Buxus balearica at different scales. Plant Biology 9:736-744

Lee AP (1930) The Brazilian babassu: tropical forests natural storehouses for great quantities of high quality oil. Oil and Fat Industries, March:91-107

Levin DA (2009) Flowering-time plasticity facilitates niche shifts in adjacent populations. New Phytologist 183:661-666

Levin DA, Brack ET (1995) Natural selection against white petals in Phlox. Evolution 49: 1017-1022

Lima IND, Condere ACO, Moret AS (2006) Potencial econômico do babaçu (Orbignya speciosa) na RESEX Rio Ouro Preto. International Agroenergy Conference, Oct 18-20, Londrina, Paraná, Brasil

Lima JRO, Silva RB, Silva CCM, Santos LSS, Santos JRS, Moura EM, Moura CVR (2007) Biodiesel de babaçu (Orbignya sp.) obtido por via etanólica. Quimica Nova 30:600-603

Lloyd DG (1979) Parental strategies of angiosperms. New Zealand Journal of Botany 17:595-606

Lobo JA, Quesada M, Stoner KE, Fuchs EJ, Herrerias-Diego Y, Rojas J (2003) Factors affecting phenological patterns of Bombacaceous trees in seasonal forests in Costa Rica and Mexico. American Journal of Botany 90:1054-1063

Lopez S, Dominguez CA (2003) Sex choice in plants: facultative adjustment of the sex ratio in the perennial herb Begonia gracilis. Journal of Evolutionary Biology 16:1177-1185

Lovett Doust J, Cavers PB (1982) Sex and gender dynamics in jack-in-the-pulpit, Arisaema triphyllum (Araceae). Ecology 63:797-808

Lumsden LF, Bennett AF (2005) Scattered trees in rural landscapes: Foraging habitat for insectivorous bats in south-eastern Australia. Biological Conservation 122:205-222

130

Mann CC (2008) Ancient earthmovers of the Amazon. Science 321:1148-1152

Manning AD, Fischer J, Lindenmayer DB (2006) Scattered trees are keystone structures – Implications for conservation. Biological Conservation 132:311-321

Manning AD, Gibbons P, Lindenmayer DB (2009) Scattered trees: A complementary strategy for facilitating adaptive responses to climate change in modified landscapes? Journal of Applied Ecology 46:915-919

Marcote-Rios G, Bernal R (2001) Remains of palms (Palmae) at archaeological sites in the New World: A review. The Botanical Review 67:309-350

Marquis RJ (1988) Phenological variation in the Neotropical understory shrub Piper arielanum: Causes and consequences. Ecology 69:1552-1565

Marten S, Quesada M (2001) Phenology, sex expression, and reproductive success of the rare Neotropical palm Geonoma epetiolata. Biotropica 33:596-605

McGrath DA, Smith CK, Gholz HL, Oliveira FA (2001) Effects of land-use change on soil nutrient dynamics in Amazonia. Ecosystems 4:625-645

McNaughton SJ (1979) Grazing as an optimisation process: grass ungulate relationships in the Serengeti. American Naturalist 113:691-703

McPherson K, Williams K (1998) The role of carbohydrate reserves in the growth, resilience, and persistence of cabbage palm seedlings (Sabal palmetto). Oecologia 117:460-468

McPherson K, Williams K (1998a) Fire resistance of cabbage palms (Sabal palmetto) in the southeastern USA. Forest Ecology and Management 109:197-207

Meagher TR, Delph LF (2001) Individual flower demography, floral phenology and floral display size in Silene latifolia. Evolutionary Ecology Research 3: 845-860

Mejia K (1988) Utilization of palms in eleven mestizo villages of the Peruvian Amazon (Ucayali River, Department of Loreto). Advances in Economic Botany 6:130-136

Melendez-Ackerman E, Campbell DR (1997) Adaptive significance of flower color and inter-trait correlations in an Ipomopsis hybrid zone. Evolution 52:1293- 1303

131

Mendoza A, Piñero DM, Sarukhán J (1987) Effects of experimental defoliation on growth, reproduction and survival of Astrocaryum mexicanum. Journal of Ecology 75:545-554

Miller PM, Kauffman JB (1998) Seedling and sprout response to slash-and-burn agriculture in a tropical deciduous forest. Biotropica 30:538-546

Mitja D, Ferraz IDK (2001) Establishment of babassu in pastures in Pará, Brazil. Palms 45:138-147

Moraes MR (1996) Diversity and distribution of palms in Bolivia. Principes 40:75- 85

Moraes MR, Borchsenius F, Blicher-Mathiesen U (1996) Notes on the biology and uses of the motacú palm (Attalea phalerata, Arecaceae) from Bolivia. Economic Botany 50:423-428

Moraes M, Sarmiento J, Oviedo E (1995) Richness and uses in a diverse palm site in Bolivia. Biodiversity and Conservation 4:719-727

Moran EF (1980) Developing the Amazon. Indiana University Press, Bloomington. pp. 114-143

Moret AS (2006) Óleo vegetatl como combustível para energia elétrica em pquenos aglomerados de Rondônia como forma de geração de renda. Congresso Brasileiro de Planejamento Energético, Brasilia (DF), Brazil

Morgan, D (2005) Brazil biofuel strategy pays off as gas prices soar. Washington Post, June 18

Murren CJ (2002) Effects of habitat fragmentation on pollination: Pollinators, pollinia viability and reproductive success. Journal of Ecology 90:100-107

Nadkarni NM, Haber WA (2009) Canopy seed banks as time capsules of biodiversity in pasture-remnant tree crowns. Conservation Biology 23:1117-1126

Neill C, Piccolo MC, Steudler PA, Melillo JM, Feigl BJ, Cerri CC (1995) Nitrogen dynamics in soils of forests and active pastures in the Western Brazilian Amazon Basin. Soil Biology and Biochemistry 27:1167-1175

Nelson et al. (1994) Forest disturbance by large blowdown in the Brazilian Amazon. Ecology 75:853-858

Nepstad DC, Uhl C, Serrao EAS (1991) Recuperation of a degraded Amazonian landscape: Forest recovery and agricultural restoration. Ambio 20:248-255

132

Nepstad D, Uhl C, Pereira CA, da Silva JMC (1996). A comparative study of tree establishment in abandoned pastures and mature forest of eastern Amazonia. Oikos 76:25-39

Nepstad D, Verissimo A, Alencar A, Nobre C, Lefebvre P, Schlesinger C, Potter C, Moutinho P, Lima E, Cochrane M, Brooks V (1999) Large-scale impoverishment of Amazonian forests by logging and fire. Nature 398:505- 508

Newstrom LE, Frankie GW, Baker G (1994) A new classification for plant phenology based on flowering patterns in lowland tropical rain forest trees at La Selva, Costa Rica. Biotropica 26:141-159

Nunez LA, Bernal R, Knudsen JR (2005) Diurnal palm pollination by mystropine beetles: Is it weather-related? Plant Systematics and Evolution 254:149- 171

Nye PH, Greenland DJ (1960) The soil under shifting cultivation. Technical Communication No. 51. Harpenden, U.K., Commonwealth Bureau of Soils

O Globo (2009) Barco vira e derrama 28 mil litros de oleo diesel em rio do Acre. O Globo, Jan 20. http://oglobo.globo.com/pais/mat/2009/01/20/barco-vira- derrama-28-mil-litros-de-oleo-diesel-em-rio-do-acre-754059725.asp

Oboh BO, Fakorede MAB (1999) Effects of weather on yield components of the oil palm in a forest location in Nigeria. Journal of Oil Palm Research 11:79-89

Oyama K, Mendoza A (1990) Effects of defoliation on growth, reproduction, and survival of a neotropical dioecious palm, Chamaedorea tepejilote. Biotropica 22:119-123

Pacala S, Socolow R (2004) Stabilization wedges: Solving the climate problem for the next 50 years with current technologies. Science 305:968-972

Paniagua Zambrana NYP, Byg A, Svenning JC, Moraes M, Grandez C, Balslev H (2007) Diversity of palm uses in the western Amazon. Biodiversity and Conservation 16:2771-2787

Pardo M (2001) Estrutura genética de Castanha do Brasil (Bertholletia excelsa H.B.K) em floresta e em pastagens no leste do estado do Acre. Master’s thesis, Universidade de São Paulo, Piricicaba, Brasil

Parra-Tabla V, Rico-Gray V, Carbajal M (2004) Effect of defoliation on leaf growth, sexual expression and reproductive success of Cnidoscolus aconitifolius (Euphorbiaceae). Plant Ecology 173:153-160

133

Pedersen HB (1994) Management, extractivism and commercial use of wild palms in Ecuador. In (M. Rios and H.B. Petersen, eds) Las plantas y el hombre. Quito, Ecuador: ABYA-YALA, pp 12-22

Pereira de Souza RP, Valio IFM (2001) Seed size, seed germination, and seedling survival of Brazilian tropical tree species differing in successional status. Biotropica 33:447-457

Pesce, C (1985) Oil palms and other oil seeds of the Amazon. Reference Publications Inc, Algonac, Michigan

Pinard MA, Putz FE (1992) Population matrix models and palm resource management. Bulletin de l’Institut Francais d’Etudes Andines 21:637-649

Pinard M (1993) Impact of stem harvesting on populations of Iriartea deltoidea (Palmae) in an extractive reserve in Acre, Brazil. Biotropica 25:2-14

Piñero DM,Sarukhan M (1982) Reproductive behavior and its variability in a tropical palm, Astrocaryum mexicanum. Journal of Ecology 70:461-472

Piñero DM, Martinez-Ramos M, Sarukhan J (1984) A population model of Astrocaryum mexicanum and a sensitivity analysis of its finite rate of increase. Jounal of Ecology 72:977-991

Pinheiro CUB (2008) Germination strategies of palms: the case of Schippia concolor Burret in Belize. Brittonia 53:519-527

Pinheiro CUB, Frazão JMF (1995) Integral processing of babassu palm (Orbignya phalerata, Arecaceae) fruits: Village level production in Maranhão, Brazil. Economic Botany 49:31-39

Pires-O’Brien M (1993) Local distribution and ecology of “palha preta””—A pioneer and invasive palm in Jari, Lower Amazon. Principes 37:212-215

Popenoe H (1960) Effects of shifting cultivation on natural soil constituents in Central America. PhD dissertation, University of Florida, Gainesville

Porro R (2005) Palms, pastures, and swidden fields: The grounded political ecology of “agro-extractive/shifting-cultivator peasants” in Maranhão, Brazil. Human Ecology 33:17-56

Pulido MT, Caballero J (2006) The impact of shifting agriculture on the availability of non-timber forest products: The example of Sabal yapa in the Maya lowlands of Mexico. Forest Ecology and Management 222:399-409

Putz FE, Brokaw NVL (1989) Sprouting of broken trees on Barro Colorado Island, Panama. Ecology 70:508-512

134

Queenborough SA, Burslem DFP, Garwood NC, Valencia R (2007) Determinants of biased sex rations and inter-sex costs of reproduction in dioecious tropical forest trees. American Journal of Botany 94:67-78

Quinn GP, Keough MJ (2002) Experimental design and data analysis for biologists. Cambridge University Press, New York

Quiroga-Castro VD, Roldan AI (2001) The fate of Attalea phalerata (Palmae) seeds dispersed to a tapir latrine. Biotroipca 33:472-477

Ramirez N (2005) Plant sexuals systems, dichogamy, and herkogamy in the Venezuelan Central Plain. Flora 200:30-48

Ranzi A, Aguiar R (2005) Geoglifos da Amazônia, perspectiva aérea. Starprint Gráfica, Florianopolis, Brasil

Rathcke B, Lacy EP (1985) Phenological patterns of terrestrial plants. Annual Review of Ecology and Systematics 16:179-214

Rathmann R, Szklo A, Schaeffer R (2010) Land use competition for production of food and liquid biofuels: An analysis of the arguments in the current debate. Renewable Energy 35:14-22

Ratsirarson J, Silander JA (1996) Reproductive biology of the threatened Madagascar triangle palm: Neodypsis decaryi Jumelle. Biotropica 28:737- 745

Ratsirarson J, Silander JA, Richard AF (1996) Conservation and management of a threatened Madagascar palm species, Neodypsis decaryi Jumelle. Conservation Biology 10:40-52

Richardson CR, Clay K (2001) Sex-ratio variation among Arisaema species with different patterns of gender diphasy. Plant Species Biology 16:139-149

Rios RS, Pacheco LF (2006) The effect of dung and dispersal on postdispersal seed predation of Attalea phalerata (Arecaceae) by bruchid beetles. Biotropica 38:778-781

Roosevelt AC, Lima da Costa M, Lopes Machado C, Michab M, Mercier N, Valladas H, Feathers, J, Barnett W, Imazio da Silveira M, Henderson A, Silva J, Chernoff B, Reese DS, Holman JA, Toth N, Schick K. (1996) Paleoindian cavedwellers in the Amazon: the peopling of the Americas. Science 272:373-384

Saldarriaga JG, West DC, Tharp ML, Uhl C (1988) Long-term chronosequence of forest succession in the Upper Rio Negro of Colombia and Venezuela. Journal of Ecology 76:938-958

135

Salm R, Jalles E, Schuck-Paim C (2005) A model for the importance of large arborescent palms in the dynamics of seasonally-dry Amazon forests. Biota Neotropica 3:1-6

Sanchez PA, Villachica JH, Bandy DE (1983) Soil fertility dynamics after clearing a tropical rainforest in Peru. Soil Sci. Soc. Am. J. 47:1171-1178

Sanford RL, Saldarriaga J, Clark KE, Uhl C, Herrera R (1985) Amazon rain-forest fires. Science 227:53-55

Scariot AO, Lleras E, Hay JD (1991) Reproductive biology of the palm Acrocomia aculeata in Central Brazil. Biotropica 23:12-22

Scariot AO, Lleras E, Hay JD (1995) Flowering and fruiting phenologies of the palm Acrocomia aculeata: Patterns and consequences. Biotropica 27:168- 173

Schaffner JH (1921) Influence of environment on sexual expression in hemp. Botanical Gazette 71:197-219

Schaffner JH (1925) Experiments with various plants to produce change of sex in the individual. Bulletin of the Torrey Botanical Club 52:35-47

Schaan DP, Ranzi A, Parssinen M, eds (2007) Arqueologia da Amazônia Ocidental: os geoglifos do Acre. Editora da Universidade Federal do Pará, Belém, Pará, Brasil

Scharlemann JPW, Laurance WF (2008) How Green are biofuels? Science 319:43-44

Schroth G, da Mota MSS, Lopes R, de Freitas AF(2004) Extractive use, management and in situ domestication of a weedy palm, Astrocaryum tucuma, in the central Amazon. Forest Ecology and Management 202:161-179

Schwilk DW, Ackerly DD (2005) Is there a cost to resprouting? Seedling growth rate and drought tolerance in sprouting and nonsprouting Ceanothus (Rhamnaceae). American Journal of Botany 92:404-410

Searchinger T, Heimlich R, Houghton RA, Dong F, Elobeid A, Fabiosa J, Tokgoz S, Hayes D, Yu T-H (2008) Use of U.S. croplands for biofuels increases greenhouse gases through emissions from land use change. Science 319:1238-1240

Solomon BP (1985) Environmentally influenced changes in sex expression in an andromonoecious plant. Ecology 66:1321-1332

136

Souza AF, Martins FR (2002) Spatial distribution of an undergrowth palm in fragments of the Brazilian Atlantic Forest. Plant Ecology 164:141-155

Souza AF, Martins FR, Silva Matos DM (2000) Detecting ontogenetic stages of the palms Attalea humilis in fragments of the Brazilian Atlantic forest. Canadian Journal of Botany 78:1227-1237

Stevenson PR, Castellanos MC, Cortes AI, Link A (2008) Flowering patterns in a seasonal tropical lowland forest in Western Amazonia. Biotroipca 40:559- 567

Stiles EA, Cech NB, Dee SM, Lacey EP (2007) Temperature-sensitive anthocyanin production in flowers of Plantago lanceolata. Physiologia Plantarum 129:756-765

Svenning JC (2001) On the role of microenvironmental heterogeneity in the ecology and diversification of neotropical rain-forest palms (Arecaceae). The Botanical Review 67:1-53

Svenning JC (1998) The effect of land-use on the local distribution of palm species in an Andean rain forest fragment in northwestern Ecuador. Biodiversity and Conservation 7:1529-1537

Tan KT, Lee KT, Mohamed AR, Bhatia S (2009) Palm oil: addressing issues and towards sustainable development. Renewable and Sustainable Energy Reviews 13: 420-427

Teixeira MA (2008) Babassu-A new approach for an ancient Brazilian biomass. Biomass and Bioenergy 32:857-864

Teixeira MA , Carvalho MG (2007) Regulatory mechanism for biomass renewable energy in Brazil, A case study of the Brazilian Babassu oil extraction industry. Energy 32:999-1005

Tilman D, Socolow R, Foley JA, Hill J, Larson E, Lynd L, Pacala S, Reilly J, Searchinger T, Somerville C, Williams R (2009) Beneficial biofuels—the food, energy, and environment trilemma. Science 325:270-271

Tomlinson PB (1960a) Essays on the morphology of palms I. Germination and the seedling. Principes 4:56-61

Tomlinson PB (1960b) Seedling leaves in palms and their morphological significance. Journal of the Arnold Arboretum 41:414-428

Tomlinson PB (1979) Systematics and ecology of the Palmae. Annual Review of Ecology and Systematics 10:85-107

137

Tomlinson PB (1990) The structural biology of palms. Oxford University Press, New York

Tomlinson PB, Esler AE (1973) Establishment growth in woody monocotyledons native to New Zealand. New Zealand Journal of Botany 11:627-644

Tucker JM, Brondizio ES, Moran EF (1998) Rates of forest regrowth in eastern Amazonia: A comparison of Altamira and Bragantina Regions, Pará State, Brazil. Interciencia 23:64-73

Uhl C (1987) Factors controlling succession following slash-and-burn agriculture in Amazonia. Journal of Ecology 75:377-407

Uhl C, Jordan C (1984) Vegetation and nutrient dynamics during the first five years of succession following forest cutting and burning in the Rio Negro region of Amazonia. Ecology 65:1476-1490

Uhl C, Buschbacher R, Serrão EAS (1988) Abandoned pastures in eastern Amazonia. I. Patterns of plant succession. Journal of Ecology 76:633-681

Uhl C, Clark K, Clark H, Murphy P (1982) Early plant succession after forest cutting and burning in the Upper Rio Negro region of the Amazon Basin. Journal of Ecology 69:631-648

Uhl N, Dransfield J (1987) General palmarum. L.H. Bailey Hortorium and Allen Press, Lawrence, Kansas

Underwood AJ (1994) On beyond BACI: sampling designs that might reliably detect environmental disturbances. Ecological Applications 4:3-15

Van Holt T (2001) The influences of the Scheelea phalerata palm and landscape patterns on the terrestrial mammalian and avian communities of forest islands in the Brazilian Pantanal. Master’s Thesis, University of Florida, Gainesville

Vesk PA, Westoby M (2004) Sprouting ability across diverse disturbances and vegetation types worldwide. Journal of Ecology 92:310-320

Vieira ICG, Uhl C, Nepstad D (1994) The rold of the shrub Cordia multispicata Cham. as a ‘succession facilitator’ in an abandoned pasture, Paragominas, Amazonia. Vegetatio 115:91-99

Voeks RA (1987) A biogeography of the piassava fiber palm (Attalea funifera Mart.) of Bahia Brazil. PhD Dissertation, University of California, Berkeley

Voeks RA (1988) Changing sexual expression of a Brazilian rain forest palm (Attalea funifera Mart.). Biotropica 20:107-113

138

Vormisto J, Tuomisto H, Oksanen J (2004) Palm distribution patterns in Amazonian rainforests: what is the role of topographic variation? Journal of Vegetation Science 15:485-494

Warren J, MacKenzie S (2001) Why are all colour combinations not equally represented as flower-colour polymorphisms? New Phytologist 151:237- 241

Wells PV (1969) The relation between mode of reproduction and extent of speciation in woody genera of the California chaparral. Evolution 23:264- 267

Werren JH, Beukeboom LW (1998) Sex determination, sex ratios, and genetic conflict. Annual Review of Ecology and Systematics 29:233-261

Wick B, Tiessen H, Menezes RSC (2000) Land quality changes following the conversion of the natural vegetation into silvo-pastoral systems in semi- arid NE Brazil. Plant and Soil 222:59-70

Wijdeven SMJ , Kuzee ME (2000) Seed availability as a limiting factor in forest recovery processes in Costa Rica. Restoration Ecology 8:414-424

Williams CA, Harborne JB, Glassman SF (1983) Flavonoids as taxonomic markers in some cocosoid palms. Plant Systematics and Evolution 142:157-169

Williams CA, Harborne JB, Glassman SF (1985) Further flavonoid studies on Attalea species and some related cocosoid palms. Plant Systematics and Evolution 149:233-239

Williams M (2008) A new look at global forest histories of land clearing. Annu. Rev. Environ. Resourc. 33:345-367

Wright SJ, van Schaik CP (1994) Light and the phenology of tropical trees. American Naturalist 143:192-199

Zarin DJ, Davidson EA, Brondizio E, Vieira ICG, Sá T, Feldpausch T, Schuur EAG, Mesquita R, Moran E, Delamonica P, Ducey MJ, Hurtt GC, Salimon C, Denich M (2005) Legacy of fire slows carbon accumulation in Amazonian forest regrowth. Frontiers in Ecology and Environment 3:365- 369

ZEE (Zoneamento Ecologico-Economico do Acre) (2002) Recursos naturais e meio ambiente, vol. I. Secretaria de Estado de Ciencia, Tecnologia e Meio Ambiente, Rio Branco, Acre, Brazil

139

BIOGRAPHICAL SKETCH

Joanna Marie Tucker Lima was born in 1972, in Toledo, Ohio. Her fascination with the natural world began at an early age with the encouragement of her parents who made visits to local and national parks a regular family activity. The youngest of four children, she grew up in Waterville, Ohio, graduating from Anthony Wayne High School in 1990. She then moved to Bloomington, Indiana, to study Music and Environmental

Science at Indiana University, and was awarded a Bachelor of Science in Public Affairs in December 1995. Shortly thereafter, Joanna then moved to Tucson, Arizona, where she continued her studies, receiving a Master’s degree in Latin American Studies with specializations in Brazil Studies and Ecology from the University of Arizona in

December 1998.

Joanna’s interest in tropical ecology stems from long-term involvement in forest ecology research in Amazonia. When she first travelled to Brazil in 1994 with a research team from the Indiana University, she participated in field work that aimed to better understand the influence of land use histories and soil characteristics on the rate of tropical forest succession on abandoned agricultural land. This work fueled her passion for tropical ecology and secondary forests and led to her first peer-reviewed article (Tucker et al. 1998). Over the following years, work in other parts of the Brazilian

Amazon, provided her the opportunity to learn new skills in forest inventory methodologies, community and institutional analyses and GIS, and taught her to appreciate the rich cultural and biological diversity of the region. During her master’s studies, she switched her focus to the social sciences to gain a clearer picture of the inter-relationships between the social and biological sciences and each area’s dependence on the other to successfully resolve conservation and development issues.

140

After completing her Master’s degree, Joanna worked as a research assistant for Dr.

Daniel Zarin at the University of Florida (1999–2003), conducting both laboratory and

field work with two US-Brazil collaborative research projects that examined the ecological processes of forest regrowth after deforestation in the Brazilian Amazon.

She now brings to a close one more chapter of her academic life by completing her PhD in Interdisciplinary Ecology at the University of Florida in May 2010. She hopes to continue to work as a plant ecologist specialized to better understand the impacts of land-use change on plant populations, and wants to use her skills to tackle applied questions and to help reconcile the conflicts between tropical forest conservation and development.

141