This item is only available as the following downloads:
HERBIVORY IN THE AMAZON LOWLAND TROPICAL RAINFOREST, WITH A SPECIAL FOCUS ON MYRMECOPHYTE S (GENUS TOCOCA ) BY MEGHAN MCAVOY A thesis Submitted to the Division of Natural Sciences New College of Florida in partial fulfillment for th e requirements for the degree Bachelor of Arts Under the sponsorship of Dr. Margaret Lowman Sarasota, Florida May, 2009
ii A CKNOWLEDGMENTS To my parents, who have always been there, my pups, whose cuddles make all things better, Meg Lowman, who made the study of insects and plants interesting, DC Randall, funds to return to the Amazon, Ricardo, who provided endless knowledge and compassion, Duff Cooper, without wh om my data would never have been analyzed, Jenn Creighton, whose shimmies hypnotized me into productivity, Elaine Fritschie, whose energy kept mine up, Michael Marazzi, who assured me I could do it and dealt with so much crazy, and the W clan, who gracious ly put up with absurd behavior.
iii TABLE OF CONTENTS ACKNOWLEDGMENTS ii TABLE OF CONTENTS iii LIST OF TABLES iv LIST OF FIGURES v ABSTRACT vi FOREWARD vii PREFACE viii INTRODUCTION 1 LITER ATURE REVIEW 2 Location of Field Work 2 Herbivory 2 Microclimate 3 Myrmecophytes 4 Hypotheses 8 METHODS 10 RESULTS 18 DISCUSSION 25 APPENDIX 29 REFERENCES 33
iv LIST OF TABLES Table 1. Ant domatia structure s from different plant families 6 Table 2. Temperature comparisons at different levels in canopy 20 Table 3. Percentage herbivorous insects found using beating tray 21 Table 4. Organisms viewed during plant observation 23
v LIST OF FIGURES Figure 1. Temperatures at AC TS, 21 June, 2002 3 Figure 2. Close up of Myrmecophyte ( Tococa ) showing domatia 5 Figure 3. Map showing lodges where research was conducted 10 Figure 4. Aerial view of E xplorNapo Lodge 11 Figure 5. Aerial view of ACTS canopy walkway 11 Figure 6. View of Beating tray from underside 13 Figure 7. Sampled myrmecophyte leaves 15 Figure 8. Leaf on graph paper to calculate leaf area 15 Figure 9. Penetrome ter side and top view 17 Figure 10. Penetrometer in use 17 Figure 11. Results of Broad Surveys of Herbivory 18 Figure 12. Comparisons of foliage loss for myrmecophyte 19 and non myrmecophyte zon during dry and 25 wet seasons Figure 14. Leaf mining 26 29 Figure 16. Study area near lodge 30 Figure 17. Amazon in wet season 30 Figure 18. View from upper canopy at ACTS 31 F igure 19. Explorama Lodge 31 Figure 20. Sunset from canopy 32
vi HERBIVORY IN THE AMAZON LOWLAND TROPICAL RAINFOREST, WITH A SPECIAL FOCUS ON ANT PLANTS (GENUS TOCOCA ) Meghan McAvoy ABSTRACT This thesis examines several aspects of herbivory with a specific focus on m yrmecophyte s from the genus Tococa in the Amazon lowland tropical rainforest. Field work was conducted at several different sites in the Peruvian Amazon Rainforest, selected for their proximity to the Amazon Conservatory of Tropical Studies (ACTS) canopy walkway This enabled research at different rainforest canopy levels I analyzed the percentage of leaves with signs of herbivory versus those with none. Of 2,000 leaves surveyed, 92.5% show si gns of herbivory while only 7.5% had not been fed upon I compared herbivory between two species of plants from the Tococa genus: one that houses ants and one that does not. My hypothesis was that m yrmecophyte s (i.e. ant plants or plants with bene ficial relationships with ants) will show fewer signs of herbivory and may have a greater array of foliag e because the ants defend the leaf tissue. Myrmecophytes showed statistically significant lower levels of herbivory than non myrmecophytes and had a l arger proportion of tougher, older, and larger leaves. Dr. Margaret Lowman Division of Natural Sciences
vii F OREWORD I have long been interested in studying the natural environment. I remember clearly sitting in 6 th grade earth science and hearing tales adventures around the world. She taught us about tectonic plates, clouds, and marine life. We dissected a squid and wrote our names with its back bone and ink sack. She was also the first person I met who had traveled to all seven continents and was the inspiration for my goal to do the same. I thought about studying the environment after first arriving at New College, but was unsure until my second year. In the interim, I took classes that interested me from all disciplines. I traveled to Antarctica for my second independent study project (ISP) and after my experience there chose environmental studies as my area of concentration. Whereas before I was interested in the environment, now I felt an urgency to study it and protect i t for future generations. Having experienced this nearly untouched continent, I felt an obligation to preserve it and other pristine locations like it. Six months later I had the opportunity to explore the Amazon Rain Forest as part of a field course wit h Dr. Meg Lowman and jumped at the opportunity to experience another wild frontier. With the assistance of Dr. Lowman, I decided to look at herbivory on a broad scale and study insect plant relationships in a complex tropical forest canopy. During my first week in the Amazon, I examined approximately 2,000 leaves, including assessment of herbivory and surveys of insect diversity I hope that with increased studies, people will be more aware of the benefits of differing ecosystems such as tropical rainfores ts (or the Antarctic) and conserve them.
viii PREFACE "The old river in its broad reach rested unruffled at the decline of day, after ages of good service done to the race that peopled its banks, spread out in the tranquil dignity of a waterway leading to the uttermost ends of the earth." Joseph Conrad, Heart of Darkness Part 1 The journey to the Amazon Rainforest presented some challenges not entirely Heart of Darkness (although in all honesty, and to my great relief, not enti value if it were not for the adversities faced therein. Travel time from my home to Peru was about 12 hours. It was another 2 hour plane ride from the capital city (Lima) to Iqui tos. Then we had a 4 hour boat ride to arrive at our first lodge. From the ACTS lodge to the canopy walkway we walked for about 30 minutes along the trail. In my travels, I have found the journey home to feel disproportionate to the rest of the trip. In t he case of my first trip to the Amazon, a delay in Miami Airport increased that time before I was back in a land of my native tongue (for those who have never flown through MIA, beware that some knowledge of Spanish language will be useful). On my second j ourney I arrived home at the planned time, but my luggage took another few hours to make it back. The lack of electricity and running water was something of a concern before the trips, but turned out to be one of the most wonderful aspects. The water tha t provided our showers was pumped straight from the river to our shower stall. If one wished to take a warmer shower, they would bathe in the middle of the day, when the sun had warmed the
ix apparatus holding the water. I found the showers harsh at first, bu t eventually found the cold water to be so refreshing on a humid day (and all days in the Amazon are humid). The absence of electricity enabled us to connect to nature in fabulous ways. Before falling asleep below our mosquito netting, we could hear the ca ll of countless insects. Prior to rising in the morning, it was bird sounds that inspired us to wake and prepare for the day. The lodges each had a generator that could be turned on for about an hour of lighting or phone charging and something that ran con sistently so that soft drinks and alcohol could be served cold. The later provided a venue for which we could get acquainted with one another. Many stories were told by the light of oil lamps and laughter shared in hammock houses after other guests had gon e to sleep. The arsenal of vaccinations necessary before leaving turned out to be the most painful preparation. I received 4 shots in a day (less than some of my fellow travelers who were not up to date on all of theirs) and found lifting my arms to be pai nful for a few days. I would gladly have taken many more shots in the arm though, had they relieved me of mosquito bite pain along the way. On my second trip to the rainforest (the wet season), I counted over 120 mosquito bites on my skin. This was the lea st enjoyable portion of the trips, but provided entertainment in the form of a wager with a fellow student. In the end, she had more bites (and so I bought her a drink the price of our bet), though I still stand that I had more in a smaller space of time (I received most of mine during a 1 hour hike, whilst she had received hers over the course of 3 days). There was also the added benefit of finding a reliable solution to the itchiness caused by bites (though unfortunately, I did not have the foresight to Blood is derived from a plant in the rainforest and applied to the skin directly. It gets its
x name from its color and works wonders on itchy skin. Our fabulous guide Ricardo provided some for all o f us who were unfortunate or foolish enough to be bitten. I do not know when I will next be able to return. I feel fortunate now to have not only gone, but to have done so twice in a year. The experience has shaped me significantly. I hope this glorious pl ace will remain for many years to come and inspire future environmentalists and travelers towards greatness.
1 INTRODUCTION (Gurevitich et al. 2002). My thesis examines some of the effective mechanisms plants use to defend agains t herbivory Prior to my first Amazon trip, I did a pilot study at Myakka River State Park, Sarasota, Florida. In this Myakka pilot study, I tested my survey method by using a counter and a data sheet to count 30 leaves and find what percentage of them sho wed signs of herbivory. This proved an effective and efficient method so I subsequently used it in Peru. I was able to survey leaves on the canopy walkway quickly and eas ily I began my research in the Peruvian Amazon rainforest by examining the proporti on of leaves that showed signs of herbivory. During my first journey to the Amazon, one plant genus in particular, that of Tococa caught my interest due to its abundance in the understory adjacent to our jungle lodge and its apparent interaction with inse cts. I studied two different plants from the Tococa genus Tococa guianensis (Michelangeli 2003) those that have bulbous stems that provide habitats for ants (domatia), and an unidentified Tococa spp a neighboring species without the domatia and ants. I w ill refer to the domatia bearing plants as m yrmecophyte s
2 LITERATURE REVIEW 1. ) Location of field work The tropical Amazon Rainforest consis ts of an area about the size of the continental United States and extends across eight countries. The study site was about 3 degrees south latitude in Peru and receives 210 inches of rain each year which results in 30 to 40 feet difference between the wet and dry season ( Castner 2000). The biodiversity in the Amazon may be the greatest in the world. Previous studies have found 350 species of ants from 71 genera in Rio Yuyapichis and 43 species of ants from 26 genera on a single tree at Tambopata Reserve (H lldobler and Wilson 1990). The Amazon Conservatory of Tropical Studies (ACTS) canopy walkway is the largest in the world stretching a quarter of a mile through the treetops (Lowman and Rinker 2004) It consists of 14 host trees supporting platforms that a re connected by narrow bridges (Castner 2000). 2. ) Microclimate Wellington (1957) stressed the importance of measuring climatic factors as they influence the abundance of insects. Trees in the rainforest can exceed 80 meters in height enabling a change i n conditions from base to treetop. The stratification creates a system of microclimates where floor level temperatures tending to be about 0 .5 C to 2.5 C cooler than those at the upper canopy due t o increased sunlight and radiation (Madigosky 2004). Figur e 1 illustrates temperatures at three levels at the ACTS canopy walkway over 24 hours.
3 Figure 1. Temperatures at 4, 24, and 32 meters at AC TS, 21 June, 2002 (Madigosky 2004). 3. ) Herbivory Herbivores and plants encompass about half of all organis ms on earth (Strong et al. 1984; Coley 1998). Defoliation can vary significantly depending on numerous factors such as different strata of the canopy, leaf age, and amount of light a leaf receives Lowman et al. (1994) found leaves higher in the canopy had le ss herbivory than those in the lower portion of the canopy. Herbivory rates in the tropics are higher for young leaves ( Coley 1980; Lowman 1983; 1985; Coley and Aide 1991; Coley and Barone; 1996, Coley 1998), whereas in the temperate region, rates are high er for mat ure leaves (Reichle et al. 1973; Coley 1998). Edwards and Wratten (1983) observed that insects avoided grazing on previously damaged leaves and that hole punched leaves sustained less herbivore damage than control leaves (Edwards and Wratten 1 985; Fowler and Lawton 1985). This may result
4 from the chemical defenses of a plant or the avoidance of predators or parasitoids that are often attracted to damaged lea ves (Hassell and Southwood 1978; Heinrich and Collins 1983; Fowler and Lawton 1985 ; Agra wal et al. 1998 ). Suckling et al. (1996) determined that very small or mobile species are not easily caught with the beating tray, but that it was an efficient form of sampling when comparing time spent in field and laboratory. Optimal efficiency of this method requires 10 15 samples (Suckling et al. 1996). P revious studies have used t racings of leaves onto graph paper to determine leaf area loss due to herbivores (Bray 1961 ; Lowman 1985) Bray (1961) estimated the original leaf perimeter in situations when a portion of the leaf edge was detached times using a digital penetrometer. The punches were aime d at leaf tissue, not veins, so as to obtain an analysis that accurately reflects how difficult it is for herbivorous insects to chew through leaves. Similar techniques were used by Grubbs and Cummins (1994), Cherrett (1968), and and Webster (1992 ). Lowman et al. (1994) used a penetrometer to evaluate toughness during the Jason Project by testing each leaf at least three times. 4. ) Myrmecophytes As indicated earlier, the plants that live in association with ants and possess domatia are referred to as myrmecophytes (Fonseca 1999). Symbiotic relationships between plants and ants have been observed for over a century (von Wettstein 1889; McNaughton 1983; Michelangeli 2003) and their coevolution is theorized to start as far as 60 million years ago (H lldobler and Wilson 1990).The se symbioses occur in about
5 141 plant genera and house ant species from 29 genera (Davidson and McKey 1993, Fonseca 1999). As an evolutionary indicator of this symbiosis, more than 60% (30 of 47 species) of Tococa contain dom atia on petiole with a passage to the outside along the main vein (Roth 1976; Svoma and Morawetz 1992; Michelangeli 2000) (Figure 2). As further evidence of being an evolutionary adaptation, Ridley (1920) grew myrmecophytes in greenhouses and found that do matia are formed regardless of the absence or presence of ants (Beattie 1985). Different families of myrmecophytes have different types of domatia. Table 1 depicts some of these. Figure 2 Close up of Myrmecophyte ( Tococa ) showing domatia and an ts.
6 Table 1. Ant domatia structures from different plant families (Beattie 1985). Myrmecophyte ant interactions are largely beneficial to both the plants and ants ( Huxley 1986; Fiala et al. 1989; Fonseca 1994; Janzen 1967 ; Fonseca 1999). Ants can eliminate several million herbivores in a year (Taylor 1937; Horstmann 1972, 1974; Beattie 1985) and may protect against competing plants in an area by creating physical damage to the leaf via biting and subsequent poisonous spray (Janzen 1967, 1969; Morawetz et. al. 1992; Jolivet 1996; Letourneau 1998; Renner and Ricklefs 1998; de la Fuente and Marquis 1999; Michelangeli 2003). Ants consume a variety of arthropod prey on leaf surfaces including beetles, caterpillars, tr ue bugs and sawfly larvae (Hlldobler and Wilson 1990). The plants may provide Beltian bodies as food for ants (Michelangeli 2003). Beltian bodies are structures on an Acacia leaf that provide protein and lipids for their
7 mutualistic partners (Rickson 1 975). Ant exclusion experiments on Tococa found that myrmecophytes deprived of their resident ants displayed an increased herbivory as compared to when the ants were present (Michelangeli 2003). Ants living in myrmecophytes benefit the plants not only by defending the plants from herbivores, but also via depositing nutrients into the domatia through their frass. This benefit can be observed directly as trees with ants produce more fruit than those without (McCook 1882; Groff and Howard 1925; Beattie 1985). Further, it seems ants provide the best protection when a plant species gives the greatest incentives with one example being Pseudomyrmex in Acacia The Acacia gives a home, nectar and food bodies while the Pseudomyrmex displays aggressive defense behavio r against herbivores and other plants (Beattie 1985). The resident ants in Tococa may be Azteca Crematogaster, or Solenopsis (Hlldobler and Wilson 1990 Cabrera and Jaffe 1994). Davidson et. al (1989) found Azteca and Crematogaster competing in Tococa As noted by Michelangeli (2003), some early studies of myrmecophytes perceived ants as destructive rather than helpful to plants. Though this study aims to discuss the beneficial relation of the ants to the plants, it is worth mentioning that there may be some unintentional consequences that negatively affect the plants. Seeds and fruit may go un dispersed if seed dispersers are steered away from the plants due to protective ants (Thomas 1988; Hlldobler and Wilson 1990). Some ant protection may be ineffect ive as certain herbivores have defenses against the domatia residents and some ants do not defend against all herbivores (Beattie 1985). Wheeler (1942) found over 40 different ant species in domatia throughout Central and South America and saw no evidence to suggest the ants attacked any herbivores (Beattie 1985), though this runs
8 contrary to more recent studies (Micheangeli 2003) in which ant exclusion experiments showed that control plants (plants that were unaltered) had less herbivory than experimental plants (those that had ants removed) Izzo and Vasconcelos (2002) found that mature leaves of a myrmecophyte ( Hirtella myrmecophila ) drop domatia because its obligate partner ( Allomerus octoarticulatus ) may act as a parasite. In this relationship, the ants are both useful and harmful depending upon life stage of the plant. 5. ) Hypotheses: This study a ddresses the following hypotheses in the Lowland Amazon Rainforest: a. More leaves have some sign of herbivory as compar ed to those ent irely untouched. Casual observations indicated that most leaves have holes in them, although this was never quantified by previous research. b. 1 Myrmecophytes display fewer signs of herbivory than non myrmecophytes from the same genus ( Tococa ). Presumab ly, plants that have ant guards will experience less herbivory than those without. Work by Michelangeli (2003) and Dejean, et al. ( 2005 ) indicate s that ants and myrmecophytes form a symbiotic relationship where the plant provides a home and the ants may p rotect against herbivore s 2. Myrmecophytes have a greater proportion of older leaves as compared to non myrmecophytes. Myrmecophytes have ant guards defending leaf area so their leaves should live longer because they will be less fed upon. c Leaves have higher herbivory in the dry season as compared to the wet season. T here may be a seasonal difference in herbivory levels with more insects and more soft (new) leaves available during the dry season
9 d. Toughness Studies. 1. Toughness of leaves is negat iv ely correlated with percent herbivory. It seems as though leaves that are more difficult to chew through due to their tougher nature would experience less herbivory. Coley (1983) found a correlation between leaf toughness and levels of herbivory in a Panam anian forest. 2. Leaves are tougher near the plant base (and therefore will have less herbivory than at the tip). During my pilot study, I observed that leaves seem to be more eaten at the tip as opposed to the base. I propose this is for two reasons (t he former of whi ch is testable in this study): a Leaves are tougher near t he base and b The ants reside in domatia near the base of the plant and so may not have the speed or efficiency to fight herbivores on the far side of the leaf. e Insects are more prevalent in the canopy than the understory. It seems as though there would be more for insects (both herbivorous and predatory) to feed on in the canopy than the understory due to the high biodiversity in the canopy. f Temperature is higher in the canopy than the understory. The canopy is exposed to more sunlight and radiation, while the understory is shaded.
10 METHODS Location The research was conducted in the Peruvian Amazon at Ceiba Tops Lodge, Explorama Lodge, ExploraNapo Lodge and the ACTS canopy walkw ay (3.25 S, 72.91W) (Yanoviak, et al. 2008) The sites were 25 miles downriver from Iquitos (Ceiba Tops) to 98 miles from Iquitos (ExplorNapo) along the Amazon and Napo Rivers (Figure 3 ) (Explorama, Castner 2000). The lodges were adjacent the river with t rails through the jungle and were surrounded by forest (Figure 4 ). The Amazon Conservatory of Tropical Studies (ACTS) extends through the canopy affording access to much biodiversity as well as views of the area (Figure 5 ). Random surveys of herbivory and samples and observations of plants from the genus Tococa occurred along trails surrounding the lodges. Herbivory surveys were also conducted in the canopy along the ACTS canopy walkway Figure 3 Map showing lodges where research was conducted (Castner 2000).
11 Figure 4 Aerial view of ExplorNapo Lodge (Castner 2000). Figure 5 Aerial vi ew of ACTS Canopy Walkway (Castner 2000).
12 Broad Surveys of Herbivory Over 2000 leaves were assessed in the Peruvian Amazon Rainforest on the f orest floor, in the understory, and at the upper canopy. Leaves were examined in groups of 30 to calculate the proportions of leaves in each subsample affected by herbivory. A hand counter was used to count the number out of 30 leaves that displayed signs of herbivory. Alternatively, people worked in groups of two so that one could maintain a count for total leaves and another kept track of the herbivory. An effort was made to ensure the leaves Y oung and old leaves of dif ferent species and sizes at different heights survey to insure variability of samples from a middle school student to a university level psychology profes sor to other undergraduates. (Lowman personal communication) When examining leaves both sides from leaf tip to petiole were also checked for mining or bite marks. Leaves we re randomly sampled from the ground of the rainforest to the upper canopy at various locations, days, and species (June 26 July 2, 2007 and from January 25 31, 2008). Species Identification Studies Beating trays (1 meter squared cloth area supported by PVC pipes) were used to look at the number of insect herbivores active at different times of day to understand better the feeding patterns of herbivores (Figure 6 ). Beating tray samples were taken at different levels in the rainforest in the understory an d upper canopy during midday (1200h) and after dark (2100h). The beating tray was supported by one member of a two to four person team. Another person would shake the tree or bush branch over the tray for
13 10 seconds during which time insects and arachnids would fall into the beating tray. present and one member would record the data. Since we had to identify organisms quickly (before they flew, hopped, or walked away), we gen eralized based on family to determine which were herbivores. Temperature was recorded at different levels of the canopy to see if there was a correlation between the number of insects and temperature as well as to account for the differing microclimates wi thin the strata of the rainforest. Figure 6 View of Beating tray from underside (Ministry of Environment, Lands and Parks British Columbia 1998)
14 Myrmecophyte s vs. Non Myrmecophyte s Myrmecophyte s and non m yr mecophyte s, from the genus Tococa were observed simultaneously in order to determine which plants were actively protected against herbivore s. They were viewed at sunrise (0600 h), in the mid afternoon (1200 h), at sunset (1800 h), and in the eveni ng (2100 h). The fates of any insects that landed or crawled on any of the leaves of the two plants were rec orded. Samples of leaves at different ages were taken from both myrmecophytic and non myrmecophytic plants to compare age, levels of herbivory an d leaf toughness ( Figure 7 ). The f ield station had no electricity so in order to calculate the percentage of herbivory for each leaf, the leaves were traced onto graph paper to determine their size as well as the potential leaf area (how large the leaf wo uld be without any herbivory) and actual leaf area (area of the leaf as it currently exists) (Figure 8 ). Leaf mining was considered as herbivory because the author did not know how to categorize it another way and mining affects photosynthetic activity. Th en the percentage of herbivory was calculated by subtracting the actual leaf area from the potential leaf area, dividing by the potential leaf area and multiplying by 100. Data was recorded on a data sheet in the field before being entered into Excel (upon return to the United States). 100 X ((PLA ALA)/PLA) = Percent Herbivory
15 Figure 7 Sampled myrmecophyte leaves. Figure 8 Leaf on graph paper traced to calculate leaf area (mm 2 ) and percent herbivory.
16 Leaf Toughness Three leaves were sampled from each the myrmecophyte and the non myrmecophyte plants and were tested for toughness using a penetrometer. This tool determines the strength of a leaf is by measuring how much force it takes to penetrate the leaf, thereby simulating the efforts required by herbivores to consume leaf tissue. Each leaf was placed in between two plastic slides on a hole in the penetrometer (Figure 9). Then the weighted part of the penetrometer was placed over the upper slide. A small plastic Tupperware container, with a predete rmined weight, was set on the upper part of the apparatus at which point water would be poured into it until the leaf had been penetrated (Figure 10). Two persons were required to collect data: one could clearly watch for the point of penetration while the other poured the water slowly and evenly. The random sample of myrmecophytes and non myrmecophytes contained younger leaves and older leaves to compare strength of leaves and examine potential for herbivory throughout a life cycle. The leaves were all ha nd sampled from forest floor to about 2 meters.
17 Figure 9. Penetrometer side view and top view. Figure 10 Author using penetrometer with water weight to test leaf toughness.
18 RESULTS Proportion of leaves eaten (vs. leaves intact) by herbivor es M ore leaves showed some form of herbivory as compared to those entirely intact. Out of 84 groups of 30 24 of them had 0 untouched leaves (all leaves out of 750 showed some form of herbivory) (Figure 11). 92.5% (2331) of total leaves reviewed (252 0) showed some herbivory whereas only 7.5% (189) were entirely devoid of herbivory. The percent herbivory between myrmecophytes and non myrmecophytes was analyzed using the Wilcoxon rank nonparametric one way variance test through SAS/STAT software (Welk owitz et al. 2002) The Null Hypothesis was that the average rank of ant plants and non ant plants would be the same. Reject null hypothesis ; the ant plants had a statistically signific ant degree of reduced herbivory P<0.0001. Figure 11. Results of br oad survey of leaves. Most leaves displayed herbivory.
19 Myrmecophyte Data in Field i. Herbivory Comparison The average herbivory for the non m yrmecophyte s was 39.4% leaf area eaten, while the mean herbivory for the m yrmecophyte s was 20.9% leaf area eaten supporting the hypothesis that ants protect plants from herbivory. Figure 12 displays herbivory in myrmecophyte and non myrmecophyte plants. ii. Age of myrmecophyte leaves versus non myrmecophyte leaves Of all leaves sampled, (30 of each species), 29 were considered young, 16 were considered mature, and 15 were between the two age groups. Age was determined by looking at size and texture of leaves. Mature leaves were larger, darker, and had a rougher texture than younger leaves (L owman personal communication). nature. There were 19 young non myrmecophyte leaves and 3 that were considered mature whereas there were 11 young myrmecophyte leaves and 13 matur e leaves. Figure 12 Myrmecophyte ( Tococa ) displaying minor foliage loss and Non Myrmecophyte displaying significant foliage loss.
20 Temperature Data Temperature was recorded to examine aspects of microclimate in the day and evening and was consistent between trials. As predicted, the temperature was higher in the mid and upper canopy than the understory in the day; however, in the evening, the reverse was true (Table 2) Table 2 Temperature at ACTS 30 June 2007 C Day Unde rstory Mid Canopy Upper Canopy 1 26.9 1 30.9 1 30.9 2 26.5 2 29.4 2 29.8 3 28.9 3 29.2 3 29.6 4 27.6 4 29.0 4 30.1 5 27.6 5 29.6 5 30.4 6 26.6 6 29.3 6 30.2 7 26.7 7 29.6 7 29.6 8 28.0 8 29.6 8 31.0 9 27.6 9 29.6 9 30. 3 10 28.8 10 29.3 10 30.7 Night Understory Mid Canopy Upper Canopy 1 30.5 1 27.3 1 26.4 2 29.9 2 27.5 2 26.6 3 29.7 3 27.2 3 26.1 4 28.5 4 27.3 4 26.0 5 28.8 5 27.8 5 26.3 6 28.0 6 27.6 6 29.4 7 28.3 7 27.8 7 30 .8 8 28.0 8 28.2 8 29.2 9 27.6 9 28.2 9 29.3 10 27.7 10 28.1 10 27.6 Insect survey with beating tray An average of 3.2 herbivores per meter squared was found in the canopy and the understory with the beating tray An average of 9 insects was found in the understory, whereas an average of 13.3 insects was found in the canopy supporting the hypothesis
21 that more insects would be found in the canopy. Table 3 shows the percentage of herbivores found amongst all insects sampled with beating tray. Table 3 Snapshot survey of insects found using beating tray in understory and at various heights in canopy. Location Height (meters) Total Insects Percent Herbivores Trail 0 10 0 Trail 0 7 57 Trail 0 5 20 Trail 0 16 19 ACTS Walkway 27.5 12 50 ACTS Walkway 26.0 11 1 ACTS Walkway 25.5 13 46 ACTS Walkway 30.0 17 47 ACTS Walkway 35.0 18 0 ACTS Walkway 35.5 9 44 Leaf Toughness Studies a. Leaf toughness as compared to herbivory rate A Spearman Correlation Test was run using SAS to test percent herbivory versus leaf toughness where N=6 The r val ue for the correlation is 0.771. b Tip of leaf vs. base of leaf There was no statistical significance in means of toughness between the bas e, tip,
22 or mid portion of the leaves, though the mean was highest for base (22.25 grams ), second highest for middle (18.71 grams ), and least for tip (14.54 grams ). c Young leaves versus o ld leaves A Wilcoxon rank nonparametric one way variance test was used to compare toughness of young and old leaves. There was no significant difference in toughness between old and young leaves. Toughness values averaged 18.9 for young leaves and 18.11 for old leaves. Observations During the observations of the two spe cies of Tococa 2 7 organisms were found on the leaves at each time. Few of those ob served were herbivorous (Table 4 ). During an observation session, the domatia residing ants were seen attacking a hymenoptera species on the leaf area.
23 Table 4 Organisms observed 27 and 28 of January 2008 on adjacent Myrmecophyte and Non Myrmecophyte 10:35 17:15 21:25 7:57 Non Myrmecophyte 3 Diptera 2 Hymenoptera 1 live Arachnid 1 dead Arachnid 1 Coleoptera 1 Orthoptera 1 Phasma tidae 3 Diptera 2 Arachnid 1 Hymenoptera 1 Phasmatidae 4 Arachnid 1 Lepidoptera Myrmecophyte 1 Hymenoptera 2 Arachnid 1 Phasmatidae 5 Diptera 1 Coleoptera 2 Orthoptera 1 Arachnid 1 Lepidoptera 1 Diptera 1 Phasmatidae 1 Hymenoptera (being consumed by ants) 6 Arachnid 1 Lepidoptera 1 Diptera 1 Coleoptera 1 Arachnid 1 Orthoptera Temporal and Spatial Comparisons of Herbivory A Wilcoxon rank nonparametric one way variance test was also used to determine if there was a significant difference between herbivory of random samples during the wet and dry season and from the forest floor to the canopy. No statistical significance was shown however. Comparable levels of herbivory were observed during both seasons and at different height s in the rainforest.
24 DISCUSSION The broad survey of herbivory showed that a vast majority of leaves show some signs of herbivory, though relatively few plants were entirely defoliated (personal observation). Future research may examine why herbivores do not continue to feed on a leaf after initial damage. Some previous research has indicated that plants release chemical defenses after sustaining herbivore damage, which may make them less prone to future herbivores (Edwards and Wratten 1985 ; Heinrich and Collins 1983; Ha ssell and Southwood 1978; Fowler and Lawton 1985). Penetrometer studies comparing toughness to percent herbivory were not significant, though they did suggest a trend. The sample size for this test was 6 leaves. In future studies, I would like to more le aves. There was no significant difference in mean leaf toughness between bas e, mid, and tip portion. A negative trend was found for leaf toughness and herbivory rate but the sample size was too sm all to obtain significance. Though the temperature data showed that the Amazon has little difference in temperature between summer and winter (due to its proximity to the equator) there are noticeable differences of climatic and other factors (increased rain and mosquitoes in wet season). I found a NASA photo ( NASA website) showing the Amazon greener in the dry season as opp osed to the wet season (Figure 13 ) indicating that there are seasonal differences in vegetation
25 Figu re 13 Future R esearch As I looked over the data and reviewed the literature, I thought of several questions I would like to pursue in future research. Myrmecophytic leaves showed less herbivory and had older leaves, supporting the idea of ants as protectors with statis tically significant data but presents an important question: why have non m yrmecophyte s not been outcompeted e specially where m yrmecophyte s and non m yrmecophyte s were adjacent to one another ? One possible explanation is that herbivory might be beneficial because it may ward off future herbivores. An alternate possibility is that the cost to the plant in resources to maintain the ant species is about equal to the benefit of having the
26 ants. Little, if any prior resea rch has been done on this topic. Another question is how do non herbivorous organisms affect life on leaf surface? In beating tray samples and observations of plants, arachnids were consistently found. I roles) further. Perhaps the presence of arachnids explains the low number of herbivores. What role do miners play in overall herbivory? I wanted to test the hypothesis that l eaf mining leads to increased herbivory although I was unable due to limited tim e in the research area During my pilot study in the Amazon, I noticed leaves with mining and herbivory. H erbivory seemed to follow along the l ines of the mining (figure 14). It may be that the mining attracts other herbivores because it is easier to chew through the leaf (a layer has been taken away for them). Poorter et al. (2004) support this by suggesting that herbivores choose leaves based on digestibility. I suspect that the miners are protected from the domatia residing ants as they are unable to get into the leaf where the miners feed. Figure 14 Leaf displaying signs of mining, a form of herbivory. I would like to study the relationship between the absence of herbivores and
27 moonlight. An explanation for the similarities between daytime and nigh ttime beating tray samples was suggested by a Peruvian student who assisted with data collection. She proposed that fewer insects were present in the night than expected due to the full moon; the brightness of the moon enabled their predators, such as monk eys to see more clearly and so consume greater numbers of herbivorous insects than normal (Montero, personal communication) A study by Cooper (1990) examined the hunting patterns of hyaenas at various types of moonlight, and showed that their hunting succ ess was lower on partially moonlit nights, though was equal on full moon or no moon. Fernandez Duque (2003) found that phases of the moon affected hunting patters of owl monkeys not only during the night, but also the day after certain moon stages. Is th ere a best domatia size? I observed that plants with larger domatia seemed to have less herbivory. The size of domatia likely dictates the size of ants that can reside within it and larger ants may be more effective at reducing herbivores. Limitations It is difficult to assess accurately all variables related to complexity, height, and temporal changes wi thin a tropical rainforest. Another limitation of my data is that it is not a continuous long udies, which are considered less accurate (Lowman 1983). I attempt ed to compensate for this by having research from both the summer and winter (which is the dry and wet season in the Amazon) as well as sampling at different times of day; but the constrain ts of the undergraduate calendar were not conducive to more long term data collection at a remote site. Consequently, a substantial portion of my thesis included a literature review about tropical herbivory and m yrmecophyte s.
28 I was unable to ide ntify the species of ants that serve as ant guards for the plants. This was caused by a variety of factors many species of ants that resided in domatia, the lack of ant experts present during my field work, and I had no permits to collect and import them back to the United States. Future research would be to identify and study behavior of specific ant species, as well as further quantify the dynamics of herbivory on m yrmecophyte s.
29 APPENDIX Additional photographs of site. F igure 15. The author
30 Figure 16. Pathway near l odge (study area). Figure 17 Lowland tropical rainforest in wet season (Iquitos, Peru).
31 Figure 18 View from Upper Canopy at ACTS Canopy Walkway. Figure 19 Explora ma Lodge.
32 Figure 20. Sunset from upper canopy at ACTS.
33 REFERENCES Agrawal A.A. 1998. Induced responses to herbivory and increased plant performance. Science 279: 1201 1201. Amazon Greener in Dry Season than Wet. 2006. NASA Earth Observator y. 26 Apr. 2009. http://earthobservatory.nasa.gov/IOTD/view.php?id=6707. Beattie, A.J. 1985. The Evolutionary Ecology of Myrmecophyte Mutualisms. Cambridge: Cambridge University Press. Bray, J.R. 1961. Measuremnt of leaf utilization as an inde x of minimum level of primary consumption. Oikos 12(1): 70 74. Cabrera, M. and K. Jaffe. 1994. A trophic mutualism between the myrmecophytic Melastomataceae Tococa guinanensis Abulet and an Azteca ant species. Ecotropicos 7(2): 1 10. Castner, J.L. 2000. Gainesville: Feline Press. Cherrett, J.M. 1968. A simple penetrometer for measuring leaf toughness in insect feeding studies. J Econ Entomol 61: 1736 1738. Choong, M.F. 1996. What makes a leaf tough and how this affects the pattern of Castanopsis fissa leaf consumption by caterpillars. Functional Ecology 10 5:668 674. Coley, P.D. 1980. Effects of leaf age and plant life history patterns on herbivory. Nature 184: 545 546. Coley, P.D. 1983. He rbivory and defensive characteristics of tree species in a lowland tropical forest. Ecological Monographs 53(2): 209 229. Coley, P.D. 1998. Possible effects of climate change on plant/herbivore interactions in moist tropical forests. Climatic Change 39:45 5 472. Cooper, S.M. 1990. The hunting behaviour of spotted hyaenas (Crocuta crocuta) in a region containing both sedentary and migratory populations of herbivores. African Journal of Ecology 28: 131 141. Davidson, D.W., R.R. Snelling, J.T. Longino. 1989. Competition among ants for myrmecophytes and the significance of plant trichomes. Biotropica 21(1): 64 73. Davidson, D.W. and D. McKey. 1993. The evolutionary ecology of symbiotic
34 Myrmecophyte relationship. Journal of Hymenopterist Research 2:13 83. Dejean, A., P.J. Solano, J. Ayroles, B. Corbara, and J. Orivel. 2005. Insect behaviour: arboreal ants build traps to capture prey. Nature 434: 973. Edwards, P.J. and S.D. Wratten. 1983. Wound induced defenses in plants and their consequences for pat terns of insect grazing. Oecologia (Berl.) 59: 88 93. Edwards, P.J. and S.D. Wratten. 1985. Induced plant defences against insect grazing: fact or artifact? Oikos 44: 70 74. Explorama. Amazon Explorama Lodges. 05 Apr 2009. http://www.explorama.com/index.html Fernandez Duque, E. 2003. Influences of moonlight, ambient temperature, and food availability on the diurnal and nocturnal activity of owl monkeys (Aotus azarai). Behav Ecol Sociobiol 54: 431 440. Fiala, B., H. Grunsky, U. Maschwitz, and K.E. Linsenmair. 1994. Diversity of Myrmecophyte interactions: Protective efficacy in Macaranga species with different degrees of ant association. Oecologia 97: 186 192. Fonseca, C.R. 1999. Amazonian Myrmecophyte interactions and the nesting space limitation hypothesis. Journal of Tropical Ecology 15: 807 825. Fowler, S.V. and J.H. Lawton. 1985. Rapidly induced defenses and talking trees: the t 126(2): 181 195. De la Fuente, M.A.S. and R.J. Marquis. 1999. The role of ant tended extrafloral nectaries in the protection and benefit of a Neotropical rainforest tree. Oecologia 118: 191 202. Groff, G.W. and C.W. Howard. 1925. The cultured citrus of South China. Lingnan Science Journal(2):108 114. Grubbs, S.A. and K.W. Cummins. 1994. A leaf toughness method for directly measuring the processing of naturally entrained leaf detritus in streams. Journal of the North American Benthological Society 13 (1): 68 73. Hassell, M.P. and T.R.E. Southwood. 1978. Foraging strategies of insects. Annual Review of Ecological Systems 9:75 98. Heinrich, B. and S.L. Collins. 1983. Caterpillar leaf damage, and the game of hide and seek with birds. Ecology 64: 592 602.
35 Hlldobler, B., and E.O. Wilson. 1990. The Ants. Cambridge, MA: Harvard University Press. Horstmann, K. 1972. Investigations on the food consumption of red wood ants (Formica pollyctena Foerster) in an oak forest. 2. Effect of season and supp ly of food. Oecologia 15:187 204. Huxley, C. 1986. Evolution of benevolent Myrmecophyte relationships. In: B. Juniper a nd T.R.E. Southwood (Eds.). Insects and the plant surface, 257 282. Edward Arnold, London, England. Izzo, T.J. and H.L Vasconcelos. 2002. Cheating the cheater: domatia loss minimizes the effects of ant castration in an Amazonian Myrmecophyte Oecologia 133: 200 205. Janzen, G.H. 1966. Coevolution of mutualism between ants and acacias in Central America. Evolution 20:249 75. Jolivet, P. 1996. Ants and plants. Backhuys Publishers, Leiden, The Netherlands. Letourneau, D.K. 19183. Passive aggression: an alternative hypothesis for the Piper Pheidole association. Oecologia 60: 122 126. Lowman, M., R. Foster, and N. Erwin. 1994. Jason V: Ecology of the Rain Forest Canopy in Belize. Lowman, M.D. Insect herbivory in Australian rain forests is it higher than in the Neotropics? ESA Symposium Proceedings: Are Australian Ecosystems Different? Proc. Ecol. Soc. Aust. 14:109 119. Lowman, M.D. 1982. Seasonal variation in insect abundance among three Australian rain forests, with particular reference to phytophagous types. Australian Journal of Ecology 7, 353 361. Lowman, M.D. 1984. An assessment of techni ques for measuring herbivory: is rainforest defoliation more intense than we thought? Biotropica 16 (4): 264 268. Lowman, M.D. 1985. Temporal and spatial variability in insect grazing of the canopies of five Austrlalian rain forest tree species. Australi an Journal of Ecology 10: 7 24. Lowman, M.D. 1992. Leaf growth dynamics and herbivory in five species of Australian rain forest canopy tress. Journal of Ecology. 80, 433 447. Lowman, M.D. and H.B. Rinker. 2004. Forest Canopies Second Edition. Academic
36 Press: San Diego. Madigosky S.R. in Lowman, M.D. and H.B. Rinker. 2004. Forest Canopies Second Edition. Academic Press: San Diego. McCook, H.C. 1882. Ants as beneficial insecticides. Proceedings of the Academy of Natural Sciences of Philadelphia: 263 271. McNaughton, S.J. 1983. Compensatory plant growth as a response to herbivory. Oikos 40(3): 329 336. Michelangeli, F.A. 2000. A cladistic analysis of the genus Tococa (Melastomataceae) based on morphological data. Systematic Botany 25(2): 211 234. Michelangeli, F.A. 2003. Ant protection against herbivory in three species of Tococa (Melastomateaceae) occupying different environments. Biotropica 35(2): 181 188. Ministry of Environment, Lands and Parks Resources Inventory Branch for the Terrestrial Ecosystems Task Force Resources Inventory Committee. 1998. Morawetz, W., M. Henzl, and B. Wallnfer. 1992. Tree killing by herbicide producing ants for the establishment of pure Tococa oc cidentalis populations in the Peruvian Amazon. Biodiversity and Conservation 1: 19 33. Poorter, L., M. van de Plassche, S. Willems, and R.G.A. Boot. 2004. Leaf traits and herbivory rates of tropical tree species differing in successional status. Plant Biol 6:746 754. Reichle, D.E., R.A. Goldstein, R.I. Van Hook Jr., and G.J. Dodson. 1973. Analysis of insect consumption in a forest canopy. Ecology 54: 1076 1084. Ren ner, S.S. and R.E. Ricklefs. 1998. Herbicidal activity of domatia inhabiting ants in patche s of Tococa guianensis and Clidemia heterophylla. Biotropica 30(2): 324 327. Rickson, F.R. 1975. The ultrastructure of Acacia cornigera L. beltian body tissue. American Journal of Botany 62(9): 913 922. Rico Gray, V. and P.S. Oliveira. 2007. The Ecology and Evolution of Myrmecophyte Interactions. Chicago: The University of Chicago Press. Roth, I. 1976. Internal structure of leaf dolmatia in Tococa,l Melastomaceae. Acta Biologica Venezolana 9: 221 258.
37 Strong, D.R., J.H. Lawton, T.R.E. Southwood 1984. Insects on Plants: Community Patterns and Mechanisms. Blackwell Scientific, Oxford, GB. Suckling D.M., G.M. Burnip, A.R. Gibb, F.J.L. Stavely, and S.D. Wratten. 1996. Comparison of suction and beating tray sampling for apple pests and other natur al enemies. Proc. 49 th N.Z. Plant Protection Conf: 41 47. Svoma, E. and W. Morawetz. 1992. Glandular trichomes, emergences and leaf domatia of the myrmecophyte Tococa occidentalis (Melastomatacea). Botanische Jahrbuecher fuer Systematik Pflanzengeschichte und Pflanzengeographie 114: 185 200. Taylor, T.H.C. 1937. The Biological Control of Insect in Fiji. London: Imperial Institute of Entomology. Thomas, D.W. 1988. The influence of aggressive ants on fruit removal in the tropical tree, Ficus capensis (Mora ceae). Biotropica, 20(1): 49 53. Wellington, W.G. 1957. The synoptic approach to studies of insects and climate. Annual Reviews 2: 143 162. Welkowitz, J., R.B. Ewen, and J. Cohen. 2002. Introductory Statistics for the Behavioral Sciences Fifth Edition. H oboken: John Wiley & Sons, Inc. Wettstein, R.R. von 1889. Pflanzen und Ameisen. Vereine zur Verbreit. Naturwissenschaft. Kenntnisse Wien: Schrfiten des Vareines 29: 309 327. Wheeler, W.M. 1942. Studies of neo tropical Myrmecophyte s and their a nts. Bulletin of the Museum of Comparative Zoology 90: 1 262. Wilson, E.O. (1987). The arboreal ant fauna of Peruvian Amazon forests: a first assessment. Biotropica 19 (3): 245 251. Wurdack, J.J. 1973. Melastomataceae 1 819, in Flora de Venezuela Vol. VI II, ed. T. Lasser. Instituto Botanico, Caracas. Yanoviak, S.P., M. Kaspari, R. Dudley, and G. Poinar Jr. 2008. Parsite induced fruit mimicry in a tropical canopy ant. The American Naturalist 171 (4): 536 544.