Eastern Michigan University DigitalCommons@EMU Master's Theses, and Doctoral Dissertations, and Master's Theses and Doctoral Dissertations Graduate Capstone Projects

2011 Diet of three mormoopid bats (mormoops blainvillei, pteronotus quadridens, and pteronotus portoricensis ) on Puerto Rico Ashley K. Rolfe

Follow this and additional works at: http://commons.emich.edu/theses Part of the Biology Commons

Recommended Citation Rolfe, Ashley K., "Diet of three mormoopid bats (mormoops blainvillei, pteronotus quadridens, and pteronotus portoricensis ) on Puerto Rico" (2011). Master's Theses and Doctoral Dissertations. 349. http://commons.emich.edu/theses/349

This Open Access Thesis is brought to you for free and open access by the Master's Theses, and Doctoral Dissertations, and Graduate Capstone Projects at DigitalCommons@EMU. It has been accepted for inclusion in Master's Theses and Doctoral Dissertations by an authorized administrator of DigitalCommons@EMU. For more information, please contact [email protected]. Diet of Three Mormoopid Bats (Mormoops blainvillei, Pteronotus quadridens, and

Pteronotus portoricensis) on Puerto Rico

by

Ashley Kay Rolfe

Thesis

Submitted to the Department of Biology

Eastern Michigan University

in partial fulfillment of the requirements

for the degree of

MASTER OF SCIENCE

in

Biology

Thesis Committee:

Allen Kurta, Ph.D. Chair

Daniel Clemans, Ph.D.

Jamin Eisenbach, P.h.D.

Armando Rodríguez-Durán, Ph.D

Submitted October 10, 2011

DEDICATION

This thesis is dedicated to my family, who have not only tolerated my life-long weirdness, but embraced it by educating others about bats and my research! Thank you for all your love, support, and encouragement through the hard times.

i

ACKNOWLEDGMENTS

I thank everyone who helped me toward a successful project. To A. Kurta, thank you for giving me the opportunity to conduct research with you and, more importantly, taking a risk in letting me do my master's work in Puerto Rico. I learned more through my experiences at Eastern Michigan University and through working with you than I ever thought possible, and for that I am eternally grateful. I thank J. Eisenbach for teaching me that A = B + C and for being my go-to guy, as well as D. Clemans for taking the time to work with me on my project, teaching me about molecular biology, and trusting an ecologist enough to work in his lab. To Tio (A. Rodríguez-Durán), gracias por cuidarme mientras yo estaba en Puerto Rico y por dejarme usar Mata de Plátano para mi investigación. A special thanks to E. Clare for timely, helpful, and thorough responses to my emails containing numerous questions on the techniques of PCR; and to J. O. Whitaker, Jr., for teaching me how to conduct dietary analysis of guano. Others in the Department of Biology to whom I give thanks are M. Hanes, for encouraging conversations; K. Greenwald, for phylogenetic assistance; S. Francoeur, for statistical advice and a backup copy of CANOCO; J.

VandenBosch and G. Walker, for laughs and beers at the Corner that kept me going strong; K.

Mitchell and J. Kirk, for friendship and lab expertise; and M. Laporte, for providing funding that allowed me to give professional presentations on my research. I also thank J. Sandoval for being one of the greatest friends I’ll ever have, and for being the world’s best bat-catching partner! Additional thanks to S. LaZorick, for assisting me in the field; L. Stackhouse, for being the best friend I could ever have; and the Acme Seminar Series (formally known as the

DSS), for moral support. Last, I thank the Department of Natural Resources of Puerto Rico for permission to catch bats on the island and Ciudadanos del Karso and Interamerican

ii

University at Bayamón for allowing me to stay and work at Mata de Plátano. Thank you all for making my experience at EMU and in Puerto Rico nothing but awesome. I received two

Meta Hellwig Research Fellowships from the Department of Biology, Eastern Michigan

University, and a grant from the Theodore Roosevelt Memorial Fund of the American

Museum of Natural History to help finance my research; additional funding was provided through two Faculty Research Fellowships awarded to A. Kurta by Eastern Michigan

University.

iii

ABSTRACT

This study used visual analysis to determine the percent volume and percent frequency of orders of insects in the guano of the Antillean ghost-faced bat (Mormoops blainvillei), sooty mustached bat (Pteronotus quadridens), and Puerto Rican mustached bat (Pteronotus portoricensis). The most common orders for all three species were Coleoptera,

Hymenoptera, and Lepidoptera, although the relative proportions of these orders differed among species. Variation in diet was primarily due to species, but season, habitat, and sex also affected the composition of insects in the feces. In addition, species-level identification of prey was achieved through molecular techniques that examined mitochondrial DNA isolated from fragments of arthropods contained within the guano of M. blainvillei and P. quadridens. A total of 18 and 6 species of insect were identified to species for M. blainvillei and P. quadridens, respectively, and nine of these species were either pests of agricultural crops or vectors of human disease.

iv

TABLE OF CONTENTS

Dedication...... i

Acknowledgments...... ii–iii

Abstract...... iv

List of Tables...... vi

List of Figures...... vii

Chapter 1: Standard Fecal Analysis...... 1

Introduction...... 2

Methods...... 12

Results...... 17

Discussion...... 20

Literature Cited...... 30

Chapter 2: Molecular Analysis of Feces...... 64

Introduction...... 65

Methods...... 69

Results...... 73

Discussion...... 75

Literature Cited...... 83

v

LIST OF TABLES

1.1. Number of pellets used for standard fecal analysis...... 38

1.2. Number of individuals used for standard fecal analysis...... 39

1.3. The percent frequency of arthropods consumed by each species……...... 40

1.4. The percent volume of insects consumed by each species of bat………………...... ….. 41

1.5. Ranking of independent variables for the interspecific CCA………...……...…....….... 42

1.6. Ranking of independent variables for the intraspecific CCA...... 43

1.7. Comparison of overall Simpson’s indices...... 44

1.8. Comparison of dietary diversity between moist and xeric habitats...... 45

1.9. Comparison of dietary diversity between seasons……...... 46

1.10. Comparison of dietary overlap between north and south……...... 47

1.11. Comparison of dietary overlap between seasons…………...... 48

1.12. Percent volume between habitats and between seasons……...... 49

2.1. Number of samples of DNA extracted and amplified……...... 89

2.2. Insects found in the diet of M. blainvillei……...... ……...... 90

2.3. Insects found in the diet of P. quadridens……...... ……...... 92

2.4. Amplification of samples for different methods of preservation ………………..….… 93

2.5. Comparison of the methods of preservation………………………………….……..… 94

2.6. Amplification of samples for different primers……………..……..………………..… 95

2.7. Amount of amplification among studies……………….…………………..………….. 96

2.8. Comparison of the amount of amplification among studies..………………..….…..... 97

vi

LIST OF FIGURES

1.1. Physiographic regions of Puerto Rico...... 50

1.2. Average annual precipitation in Puerto Rico...... 51

1.3. Life zones of Puerto Rico...... 52

1.4. Subtropical moist forest habitat...... 53

1.5. Subtropical dry forest habitat...... 54

1.6. Hot cave and the segregation of mormoopids...... 55

1.7. Portrait of the Antillean ghost-faced bat...... 56

1.8. Portrait of the sooty mustached bat...... 57

1.9. Portrait of the Puerto Rican mustached bat...... 58

1.10. Location of cave sites...... 59

1.11. CCA dietary centroids for all mormoopids...... 60

1.12. Percent dietary overlap among the Puerto Rican mormoopids...... 61

1.13. Percent dietary overlap between seasons...... 62

1.14. Number of males and females captured for each species……………………………. 63

2.1. Areas of coffee farming on Puerto Rico...... 98

vii

CHAPTER 1 STANDARD ANALYSIS OF FECES

1

INTRODUCTION

Coexistence of species with similar body mass, dietary requirements, life histories, and behavior has been a long-standing topic of interest to biologists. Consequently, many studies have looked at the importance of niche partitioning (Dueser and Shugart, 1979; Heske et al., 1994; Kalcounis-Rüppell and Millar, 2002; Knuth and Barrett, 1984; Morris, 1984;

Wolff and Dueser, 1986) and competition (Abramsky et al., 1979; Dueser and Hallett, 1980;

Morris, 1999; Seagle, 1985) in structuring assemblages of small mammals (Christopher and

Barret, 2006). Syntopic species that are closely related, in particular, are valuable in studies of niche partitioning because these species are most likely current or past competitors

(Kalcounis-Rüppell and Millar, 2002).

Syntopic species depend on mechanisms that will ultimately reduce the amount of interspecific competition and enhance individual fitness (Kinahan and Pilley, 2008).

Ecological niches of syntopic species can be separated spatially, behaviorally, temporally, or through dietary differentiation (Christopher and Barret, 2006; Kalcounis-Rüppell and Millar,

2002; Kinahan and Pilley, 2008). Niche partitioning maximizes availability of habitat, reduces competitive exclusion, and allows for the coexistence of species (Kozlowski et al.,

2008).

Bats are the second largest order of mammals, with over 1,100 species that have diverse diets, including blood, fish, pollen, fruit, nectar, and especially insects (Schnitzler and

Kalko, 2001). Previous studies on bats (Aldridge and Rautenbach, 1987; Findley and Black,

1983) have shown that species with similar morphology tend to have similar foraging behavior (Patterson et al., 2003) and, therefore, similar diets (Rakotoarivelo et al., 2007).

Because habitat and food are potential limiting factors that affect the assemblages of animals,

2 including bats (Hickey et al., 1996), dietary partitioning can be a key mechanism for the coexistence of syntopic species.

Dietary partitioning occurs when each species in an assemblage alters its diet so that each utilizes different sources of food (Kinahan and Pilley, 2008). These alterations in diet can be due to changes in the behavior or morphology of one or more of the species in an assemblage over time (Aldridge and Rautenbach, 1987). For dietary differences to be a mechanism for coexistence of syntopic species (Husar, 1976), each species must be more efficient at locating, handling, and digesting its particular food than its competitors (Kinahan and Pilley, 2008). However, there is conflicting evidence as to whether syntopic species of bats actually exploit different foods (Findley, 1993; Carter et al., 2004; Whitaker, 2004;

Hickey et al., 2006). For example, Aldridge and Rautenbach (1987) found that bats foraging in the same area consumed the same types of insects (Findley, 1993; Hickey et al., 2006), whereas Black (1974) was able to group syntopic species of bats into moth-specialists or beetle-specialists. Husar (1976) reported similar diets for long-eared myotis (Myotis evotis) and southwestern myotis (Myotis auriculus) when in allopatry; however, when the species were syntopic the diet of M. evotis was altered, presumably to reduce competition. In addition, assemblages of bats that use the same foraging strategy can partition food by using different foraging habitats (Aldridge and Rautenbach, 1987; Saunders and Barclay, 1992;

Whitaker, 1994) or by exploiting the same foraging habitat at different times throughout the night (Kunz, 1973; Reith, 1980).

The amount of competition between syntopic species over a source of food may be reduced in intensity in areas with harsh environmental conditions (Chesson and Huntly,

1997). Nevertheless, the severity of the habitat has direct negative effects on the growth of

3 populations, and competitive exclusion can occur in harsh environments, even though competition may be less severe than in more moderate settings (Chesson and Huntly, 1997).

Habitats that receive little precipitation and are sparse in vegetation and other resources can be considered a harsh environment, especially relative to areas where rainfall is abundant and foliage is diverse. Therefore, when comparing diet among wide-ranging sympatric species, one would expect the degree of dietary overlap to be less in arid environments where the availability of resources is more limited, thus reducing competition and allowing coexistence of ecologically similar species.

Dietary variation within species.—Diets can differ between species, but also within species. This intraspecific variation in diet results from intrinsic factors, such as increased energetic and nutritional needs for hibernation, migration, growth, and especially reproduction (Murray and Kurta, 2002). Reproduction is energetically expensive, and the cost of reproduction as a proportion of the daily energy budget appears to be higher in smaller animals, such as bats, than in larger mammals (Hanwell and Peaker, 1977).

Moreover, energetic demands by insectivorous bats are highest during lactation and can be

50% greater than during pregnancy (Kurta et al., 1989, 1990). Nutritional demands (e.g., a need for calcium) also vary depending on reproductive condition (Barclay, 1994) and could lead to differences in the composition of the diet of insectivorous bats (Anthony and Kunz,

1977). Intraspecific differences in diet also have been found between juvenile and adult bats

(Anthony and Kunz, 1977; Hamilton and Barclay, 1998; Rolseth et al., 1994), which presumably results from juveniles being less experienced than adults at handling prey and at flying (Hamilton and Barclay, 1998).

The diet of an insectivorous bat is necessarily dependent on the types of insects that

4 are available, and because the abundance of insects is highly variable in space and time

(Aldridge and Rautenbach, 1987; Whitaker, 1994), one might expect the diet of an insectivorous species to vary with the seasons and even throughout the night (Murray and

Kurta, 2002). Tropical areas typically are characterized as having two distinct seasons: a wet season, which usually occurs from June to October, and a dry season, which often lasts from

November to May. The amount of rainfall received in areas of the tropics is directly and negatively correlated with the length and intensity of the dry season, which also greatly affect the diversity, abundance, and composition of insects (Janzen and Schoener, 1968). Thus, it is reasonable to expect that the diet of a wide-ranging insectivorous bat will differ with geographic location, especially if the annual amount of precipitation varies greatly among sites.

Bats in the Caribbean.—Oceanic islands are typically rich in endemic species, many of which are bats, but islands are also poorer in overall species richness than assemblages on the mainland (Groom, 2006). Insular populations also are usually small and subject to local catastrophes, such as hurricanes, that affect the availability of habitat and food, thus contributing to high rates of extinction (MacArthur and Wilson, 1967). For example, biologists have documented 69 ―extinction events‖ involving bats on islands in the

Caribbean, with an extinction event defined as any one species becoming extinct on any one island (Soto-Centeno and Kurta, 2006). Knowledge of diet is basic information needed for proper management of any species, and such information is critical for maintaining populations of endemic or endangered species, like those on oceanic islands (Soto-Centeno and Kurta, 2006).

Bats and Puerto Rico.—With an area of ca. 8,900 km2, Puerto Rico is the smallest of

5 the Greater Antilles, which also include Cuba, Hispaniola, and Jamaica (Gannon et al., 2005).

The island of Puerto Rico contains three broad physiographic regions: coastal plains, a northern karst region, and a mountainous interior (Fig. 1.1). Prevailing winds from the northeast and the interior mountains combine to produce a rain-shadow effect, resulting in striking differences in precipitation between the northern (moist) and southern (xeric) portions of the island (Gannon et al., 2005; Fig. 1.2), and these differences in moisture result in conspicuous differences in vegetative composition.

Biologists divide the island of Puerto Rico into six major life zones, following the system of Holdridge (1967; Gannon et al., 2005), although only two of these life zones are common in coastal areas (Fig. 1.3). Northern Puerto Rico is in the zone of subtropical moist forest, and most of southern Puerto Rico is classified as subtropical dry forest (Fig. 1.3).

Subtropical moist forest contains semi-deciduous trees that reach heights of 20 m or more and have rounded crowns (Fig. 1.4). In contrast, subtropical dry forest contains trees that are generally less than 15 m in height, with broad, flattened crowns. Twigs and trunks in the dry forest often are armored, and the leaves are typically tough and leathery or small and succulent (Fig.1.5). Annual precipitation in the moist and dry subtropical forest is 100–220 and 60–110 cm, respectively (Gannon et al., 2005).

On Puerto Rico, bats are the only native terrestrial mammals, although many introduced species, such as cats, dogs, goats, and pigs, inhabit the island in a feral state

(Gannon et al., 2005). There are currently 13 species of bats on Puerto Rico, representing five families: Molossidae (two species), Mormoopidae (three), Noctilionidae (one),

Phyllostomidae (five), and Vespertilionidae (twoGannon et al., 2005). Diet of the three species of Mormoopidae will be the focus of this study.

6

Mormoopids on Puerto Rico.—The Mormoopidae is a neotropical family, the members of which are characterized by flap-like outgrowths of skin below the lower lip and funnel-shaped ears (Dávalos, 2006). The nine extant species of mormoopid bats are classified into only two genera, Mormoops and Pteronotus (Dávalos, 2006). Within the

Puerto Rican mormoopids, two of the three species—the Antillean ghost-faced bat

(Mormoops blainvillei) and the sooty mustached bat (Pteronotus quadridens)—are endemic to the Greater Antilles (Gannon et al., 2005). The third species was long considered a subspecies of Parnell's mustached bat (Pteronotus parnellii), a widespread neotropical taxon; however, ongoing molecular studies suggest that the population of P. parnellii endemic to

Puerto Rico should receive full specific status as the Puerto Rican mustached bat (Pteronotus portoricensisDávalos, 2006), and I will refer to it as such throughout this study.

Mormoopid bats typically roost in caves where thousands and often hundreds of thousands of individuals from multiple species reside, although the different species often remain segregated within the caves (Silva-Taboada, 1979; Fig. 1.6). Members of the

Mormoopidae favor so-called ―hot caves,‖ where ambient temperatures typically exceed

30oC (Silva-Taboada, 1979). The high internal temperature results from production of heat by a large number of living bats and masses of decomposing guano, which warm the surrounding air. The warm air rises and is trapped inside, because the entrance to a hot cave is small and located lower than the internal chambers, thus minimizing exchange of air and heat with the outside environment (Gannon et al., 2005; Fig. 1.6).

Mormoops blainvillei weighs 8–11g (Lancaster and Kalko, 1996) and is characterized by a short snout containing a labionasal plate with lateral fleshy outgrowths around the nasal openings (Gannon et al., 2005). The external ears of this species are very distinct; they are

7 short and broad, with the lower edge continuous with the lower lip, thus giving the appearance of a funnel leading to the opening of the ear (Fig. 1.7). This species dwells in hot caves and emerges from roosts later than the other mormoopid species of Puerto Rico

(Gannon et al., 2005), exiting the cave during the second hour of bat activity (Rodríguez-

Durán and Lewis, 1987).

Pteronotus quadridens is the smallest species of bat on Puerto Rico, weighing only 4–

7 g. Its lower lip and nostrils have cutaneous flaps with wart-like tubercles, and this species is distinguished by small, tooth-like projections along the leading edge of the ear (Rodríguez-

Durán and Kunz, 1992; Fig. 1.8). This tiny bat also roosts in hot caves and is the first species to emerge at dusk for foraging, making it more prone to predation by diurnal birds (Gannon et al., 2005).

Pteronotus portoricensis is the largest of the three mormoopids on Puerto

Rico, weighing 10–18 g, and is characterized by slightly up-turned nostrils and obvious tufts of hair protruding from its snout (Gannon et al., 2005; Fig. 1.9). Similar to the other two mormoopids, P. portoricensis roosts in warm and hot caves and emerges shortly after sunset, after the sooty mustached bat but before the Antillean ghost-faced bat (Herd, 1983).

Previous dietary studies of mormoopids on Puerto Rico.—No published information exists on the diet of P. portoricensis, and only limited data are available concerning foods eaten by P. quadridens and M. blainvillei on Puerto Rico. For example, Rodríguez-Durán and Lewis (1987) collected fecal pellets of M. blainvillei and P. quadridens on only 5 days at

Cucaracha Cave, in Aguadilla, Puerto Rico, and determined what proportion (percent frequency) of the pellets contained various orders of insects. M. blainvillei consumed

8

Lepidoptera most frequently (58%), as well as smaller amounts of Coleoptera, Diptera,

Hemiptera, and Homoptera. For P. quadridens, Lepidoptera and Coleoptera were detected in equal frequency (47 and 48%, respectively), with Diptera, Hemiptera, and Homoptera found in 20–28% of the pellets.

Although use of percent frequency provides some measure of how often bats consume certain types of insects, the total volume of a pellet that is comprised of a specific taxon (percent volume) is more reflective of the overall importance of a particular item of prey to the diet of a bat (Sparks and Valdez, 2003; Whitaker, 1988). To date, no study has estimated percent volume of the different insects consumed for any mormoopid on Puerto

Rico or any of the Greater Antilles. Furthermore, diet of all three Puerto Rican mormoopids has never been compared between species, sexes, ages (juvenile vs. adult), time of year (wet versus dry season), or geographic distributions (wet versus dry habitats), or among reproductive conditions (pregnant, lactating, and post-lactating).

Ecological studies and canonical correspondence analysis.—The overall structure of assemblages is influenced jointly by both biotic and abiotic factors (Danielson, 1992;

Dunson and Travis, 1991; Hart, 1992; Magnan et al., 1994; Schoener, 1982), and this pluralist concept is widely accepted among ecologists. However, to simplify statistical approaches, assemblages are often analyzed in a single-factored approach, in which biologists examine the effect of only one variable at a time and naively assume that the interaction between environmental factors has a neutral effect on the structure of the assemblage (Magnan et al., 1994). Because both biotic and abiotic variables are critical in constructing assemblages, it is important to evaluate which factors account for most variation in the ecological structure, to understand the underlying mechanisms that influence the

9 structure of the assemblage (Persson and Diehl, 1990; Magnan et al., 1994; Schoener, 1986).

Similarly, because ecological patterns also vary spatially, both biotic and abiotic factors can change depending on location; therefore, it is also necessary to examine the role of spatial components in structuring assemblages (Magnan et al., 1994).

Assemblage-based studies that contain numerous environmental variables often have data that can be difficult to interpret, which can make drawing conclusions about the underlying factors influencing the structure of the assemblage complicated. This difficulty is partly due to an inability to graph and visualize the area of overlap of niche hyperspaces when more than three explanatory variables (dimensions) are used (McGarigal et al., 2000).

However, through the use of multivariate methods, such as ordination, it is possible to analyze systematic variation in the data (ter Braak and Verdonschot, 1995) and combine highly correlated explanatory variables. Combining independent variables reduces the number of dimensions that explain variation within the data, allowing one to graph and visualize better the overall niche hyperspace of a particular species (McGarigal et al., 2000).

Ecological data differ from other types of multivariate data in two critical ways: most species in an assemblage are only in a subset of the data, and the relationship between abundance of a species and environmental variables typically is nonlinear (ter Braak and

Verdonschot, 1995). Therefore, traditional, linear-based multivariate analyses, such as principle components analysis (PCA), are unsuitable for analyzing niches and the structure of assemblages (ter Braak and Verdonschot, 1995). These complications in analyzing the nonlinear data of ecological assemblages can be overcome through the use of canonical correspondence analysis (ter Braak, 1986; ter Braak and Verdonschot, 1995).

Canonical correspondence analysis (CCA) is a multivariate approach that uses

10 weighted averages, while simultaneously analyzing the relationship among species abundance and numerous environmental variables, and builds on typical ordination techniques through the use of regression (ter Braak and Verdonschot, 1995). Due to these characteristics, CCA provides a general framework for estimating and statistically testing the effects of different variables on biological assemblages, even if some effects are hidden by other large sources of variation (ter Braak and Verdonschot, 1995). CCA is frequently used to group independent variables, which may have interactive effects and overlap in explanatory power, to create gradients that illustrate which factors are the most important variables in determining the composition of the assemblage (ter Braak and Verdonschot,

1995).

The techniques used in CCA can help interpret how a multitude of species, such as the assemblage of insects found within the guano of a bat, simultaneously change with respect to a suite of abiotic and biotic factors, such as location of the population of bats, as well as the season, species, sex, age, and reproductive condition of the individuals. In addition, CCA is also widely used in long-term studies, because this method treats seasonal variation as a covariate and assesses to what extent the temporal variation can be explained by the associated environmental factors used in the study (ter Braak and Verdonschot, 1995).

Purpose of this study.—The primary goal of this study was to provide information on the diet of each of the three Puerto Rican mormoopids, by determining the percent volume, as well as percent frequency, of various taxa of arthropods found within the guano. In addition, this study investigated differences in diet among the three species, as well as potential interspecific dietary variation related to age, sex, reproductive condition, time of year (wet versus dry season), and geographic location (moist versus xeric portions of Puerto

11

Rico). Finally, this study introduces the use of CCA to dietary studies of insectivorous bats, as a way of determining which independent variables correlated most strongly with the composition (percent volume) of insects found within the feces.

METHODS

Areas of study.—Bats were obtained at three different sites on Puerto Rico: two on the northwestern and one on the southwestern portion of the island. Sampling in the north took place at Culebrones Cave, located ca. 7 km SW of Arecibo (18o44’N, 66o70’W), and at

Cucaracha Cave, which is found in Barrio Caimital Bajo (18°43' N, 67°15' W), near

Aguadilla (Fig. 1.10). Sampling in the south took place at Murciélagos Cave, also known as

Lago de Guano Cave, which is found within the Guánica State Forest (17°57'N, 65°52'W), along the southwestern coast of the municipality of Guánica (Fig. 1.10).

Culebrones Cave, a hot cave with relative humidity near 100% and ambient temperatures up to 40°C, is located on the property of Mata de Plátano, a field station managed by Interamerican University and the Puerto Rican organization Citizens of the

Karst. This cave houses more than 300,000 bats of six different species, of which 61 and

20% are M. blainvillei and P. quadridens, respectively (Rodríguez-Durán, 2009). Pteronotus portoricensis also roosts in Culebrones Cave but in limited numbers compared with the other two mormoopids. The entrance to Culebrones Cave is ca. 2 m in diameter and is located along the wall of a limestone sinkhole.

Cucaracha Cave reaches temperatures of 35°C and shelters ca. 700,000 bats, with P. quadridens and M. blainvillei making up 19 and 6% of this population, respectively (Placer,

1998). The opening to Cucaracha Cave has dimensions of ca. 1 m by 1.5 m and is located at

12 the bottom of a sinkhole, at the edge of a pasture that is surrounded by limestone hills. The habitat near both Cucaracha Cave and Culebrones Cave consists of subtropical moist forest dominated by semi-deciduous trees.

Murciélagos Cave has seven entrances, including one large opening that is ca. 40 m in diameter (Conde-Costas and González, 1990). A large body of stagnant water, infused with guano from thousands of bats, lies at the bottom of the major entrance and extends 168 m into the cave (Conde-Costas and González, 1990). At least five species of bats reside in

Murciélagos Cave, including all three mormoopids. Vegetation surrounding this site consists of subtropical dry forest comprised of cacti and scrub forest on limestone outcroppings

(Murphy and Lugo, 1986).

Field techniques.—Bats were captured during both the wet and dry seasons, during six different trips to the island: 27–28 February 2009, 15 May–6 August 2009, 23–26

November 2009, 25–29 December 2009, 3–5 March 2010, and 26–29 June 2010. To catch bats, a portable harp trap (Kunz and Kurta, 1988; Palmeirim and Rodrigues, 1993) was placed at the opening of Culebrones Cave and Cucaracha Cave. Because of the large size of the primary entrance to Murciélagos Cave, as well as the pool of fetid water, use of a harp trap was impractical, so instead, a mist net (20-m wide, and 2.5-m high) was set near the main entrance. Sampling at all sites occurred between 0100 and 0600 h.

All mormoopids that were captured were identified to species, sexed, and aged, and reproductive condition of adult females was determined. Bats were categorized as either juvenile or adult based on degree of ossification of the metacarpal-phalangeal joints

(Anthony, 1988). A female was considered pregnant if a fetus was felt by gently palpating the abdomen, lactating if milk could be expressed from the nipples, and post-lactating if there

13 was obvious re-growth of hair around the nipples and milk could not be expressed (Racey,

1988). An adult female was considered non-reproductive if she was not palpably pregnant, lactating, or post-lactating.

After initial processing, each bat was placed in a separate cloth holding bag until defecation occurred, which was typically less than 1 h after capture. The bats were then released unharmed, and all guano was transferred to individually labeled plastic bags (Ziploc,

S.C. Johnson, Racine, Wisconsin). Fecal pellets were dried in an oven later that day and stored until examination.

Dietary analysis.—Composition of the diets of M. blainvillei, P. quadridens, and P. portoricensis was ascertained to the level of order through standard fecal analysis (Whitaker,

1988). I considered a sample to be all guano collected from an individual bat. In the laboratory, fecal analysis was limited for any date of collection to 10 samples from males and

10 from females of each species, for a maximum of 60 samples per date. If more than 10 samples were collected from either sex of any species, a random-number generator was used

(Microsoft Excel, Redmond, Washington) to determine which 10 bags were selected for examination. If more than one reproductive condition occurred among females for a species on a given night, the 10 samples were split as evenly as possible among the different reproductive conditions. In addition, if both juveniles and adults of a sex of a particular species were captured within a night, the 10 samples for that sex were divided between the ages as evenly as possible. If there were fewer than 10 samples for a species or for a sex of any given species for a particular night, then all available samples were analyzed.

Dissection of five fecal pellets is sufficient to determine the diet of an individual bat

(Whitaker et al., 1996), so five pellets were randomly selected from each sample for dietary

14 analysis. Individual pellets were placed on a grid of numbered squares, and a random- number generator was used to determine which five pellets were selected. If a sample consisted of fewer than five pellets, all pellets were examined. Dried pellets were soaked in

70% ethanol and teased apart under a dissecting microscope (Murray and Kurta, 2002). The percent volume of every order of arthropods within each pellet was determined visually, and the average percent volume of each order was calculated per individual bat (Whitaker and

Clem, 1992). The percent frequency of occurrence for each order was determined by calculating the proportion of individuals of each species of bat that consumed the order in question, regardless of its total contribution to the overall volume for each bat. In this study, insects from the previously recognized order of Homoptera were included in the Hemiptera

(Triplehorn and Johnson, 2005). Presence of other items within the pellets, such as hair, seeds, vegetation, or dirt, was recorded but not included in any statistical analyses.

Statistical analyses.—A canonical correspondence analysis (CCA) was performed on the dietary data (percent volume) using CANOCO, a program provided by C. J. F. ter Braak

(Agricultural Mathematics Group, The Netherlands Organization for Applied Scientific

Research, Box 100, NL-6700 AC Wageningen, The Netherlands). This program allowed me to examine differences in percent volumes with respect to species, sexes, ages, and reproductive conditions, as well as between the wet season (June–October) and dry season

(November–May) and between types of habitat (moist versus xeric), all of which were treated as independent variables for the analysis. A preliminary analysis indicated no significant difference in the percent volume of insects found within feces collected at the two northern sites (F7, 464 = 1.84, P = 0.09), so data from these locales were combined before examining differences in diet between the northern (moist) and southern (xeric) sites.

15

Ordination scores recovered from the CCA were used to create a regression biplot that simultaneously showed the dietary items (orders) and all the independent variables; the biplot was constructed in CanoDraw (another program provided by C.J.F. ter Braak). This same process was repeated to assess intraspecific differences in diet; however, the independent variable ―species,‖ as well as any variable that did not significantly correlate with composition of the insects in the interspecific CCA, was removed before performing any species-level CCA. The best independent variables used as predictors for the types of insects found within the guano were selected in CANOCO through Monte-Carlo tests (1,000 permutations and an alpha of 0.05—Magnan et al., 1994) that compared the observed data with random samples generated by CANOCO (Hope, 1968).

To characterize dietary diversity, Simpson’s Index (Brower and Zar, 1984) was calculated,

∑푛i (푛i – 1) Ds = 1– , 푁(푁 – 1) where ni represents the abundance of each order of prey that was consumed and N is the total number of insects found in the guano of each species of bat. Simpson's Index was chosen because it is less sensitive to variations in sample size and gives less weight to rare taxa

(Banna and Gardner, 1996). In addition, compared with other diversity indices, such as the

Shannon Index, Simpson's Index is not only unbiased but also has the smallest standard deviation (Lande, 1996). Pairwise comparisons of diversity indexes of different groups of bats were made using a modified t-test:

(퐷 ) ‒ (퐷 ) t∞ = 2 2 , √푠 +푠 with the variance defined as:

16

3 2 2 4 [ ∑푝i – (∑푝i ) s2 = , 푁 and the proportion of each taxon found in the diet indicated by:

푛i pi = . 푁

The percentage of interspecific dietary overlap or similarity was quantified using

Schoener's Index:

I = 1 – 0.5(Σ|pxi – pyi |),

where pxi is the proportion of an order of arthropods found within the guano of the first

species of bat and pyi is the proportion of that same order in the guano of the second species of bat (Haroon and Pittman, 1998). Pairwise comparisons of dietary overlap were made using a modified t-test, similar to the one performed on the diversity indexes (Banna and

Gardner, 1996; Ruszczyk and Araujo, 1992). For comparison of diversity and overlap between all possible combinations of mormoopids, alpha was adjusted using Bonferroni procedures (Rosner, 2000).

RESULTS

Dietary composition.—Dietary analyses were performed on 2,200 pellets (Table 1.1) from 567 bats (Table 1.2). Eight orders of arthropods occurred in the diet of mormoopid bats on Puerto Rico: Araneae, Coleoptera, Diptera, Hemiptera, Hymenoptera, Lepidoptera,

Odonata, and Orthoptera (Table 1.3). In this study, 99.7, 98.7, and 82.9% of all M. blainvillei consumed members of the orders Lepidoptera, Coleoptera, and Hymenoptera, respectively

(Table 1.3). Other insects that frequently were eaten by M. blainvillei were Hemiptera

(24.2%) and Diptera (7.8%). Similar to M. blainvillei, most P. quadridens ate Coleoptera

(98.7%), Lepidoptera (89.8%), and Hymenoptera (87.7%), but Hemiptera and Diptera also

17 occurred in guano that was recovered from 23.4 and 14.9% of the bats, respectively. P. portoricensis appeared more eclectic in terms of its diet, with all seven orders of insects well- represented. The orders that were consumed most frequently included Coleoptera (100%),

Lepidoptera (89.5%), Hymenoptera (73.7%), and Hemiptera (42.1%), although Diptera,

Orthoptera, and Odonata also were found in 13–21% of the pellets.

In addition to percent frequency, the percent volume of each order of arthropods in the diet of each species of bat was examined (Table 1.4). The volume of the diet of M. blainvillei was composed mostly of lepidopterans (67%), followed by coleopterans (16%) and hymenopterans (12%), with three other orders each contributing less than 1% to the overall volume. For P. quadridens, Coleoptera (42%) dominated the diet, along with

Hymenoptera (27%) and Lepidoptera (24%), whereas four orders each represented less than

1% of total volume. Although P. portoricensis did not eat Araneae, these bats did consume all seven orders of insects found in the diet of the other two species of Puerto Rican mormoopids. In addition, all orders found in the guano of P. portoricensis, except

Orthoptera, contributed at least 1% to the overall diet for this bat. Coleopterans (40%) formed the largest proportion of the volume of foods eaten by this bat, although lepidopterans (27%) and hymenopterans (18%) also formed a substantial part of the diet.

Canonical correspondence analysis.—The eight orders of arthropods found within the diet of the mormoopids were reduced to only two dimensions via CCA, and the centroid for each mormoopid species was plotted (Fig. 1.11). The regression biplot of the centroids for each of the mormoopids showed obvious segregation of diet at the ordinal level, with

Lepidoptera associated mostly with M. blainvillei (F7, 559 = 206.73, P < 0.001), and

Coleoptera (F7, 559 = 3.14, P = 0.048) and Hymenoptera (F7, 559 = 3.90, P = 0.43) correlating

18 mostly with P. quadridens. Hemiptera (F7, 559 = 29.53, P < 0.001), Odonata (F7, 559 = 22.45,

P < 0.001), and Orthoptera (F7, 559 = 39.78, P < 0.001) were associated with P. portoricensis.

Species of bat explained ca. 16% of the variation in the composition of arthropods; in addition, sex accounted for ca. 3% of the variation, and season and habitat explained ca. 1% of the variation (Table 1.5). There was no significant association between reproductive condition or age with the composition of arthropods in the overall diet of the mormoopid bats

(Table 1.5); therefore, these variables, in addition to the variable species, were eliminated prior to performing intraspecific CCA.

For M. blainvillei, only season and habitat correlated with dietary composition, and each variable explained ca. 1% of the variation (Table 1.6). The makeup of arthropods in the diet of P. quadridens also correlated significantly with season and habitat, explaining ca. 3 and 2% of the variation, respectively (Table 1.6). For P. portoricensis, habitat accounted for

10% of variation in the diet, but unlike the other mormoopids, sex explained 15% of the variation, which indicated significant partitioning between males and females of this species

(Table 1.6).

Dietary diversity and overlap.—The diet of P. portoricensis, as indicated by

Simpson’s Index, was the most diverse (0.70), followed closely by that of P. quadridens

(0.66) and ultimately M. blainvillei (0.47). Diversity of prey differed significantly between

M. blainvillei and. P. quadridens and between M. blainvillei and P. portoricensis, but not between the two species of Pteronotus (Table 1.7). Within each species, the effects of habitat and season on diversity at the ordinal level generally were not significant (Tables 1.8–1.9).

The only exception was that diversity was greater for P. quadridens during the wet season than the dry season (Table 1.9).

19

Dietary overlap among all three species, as shown by Schoener's Index, was 58.3%, whereas pairwise analyses indicated a similarity of 44.8–89.6% (Tables 1.10–1.11). The effects of habitat and season on dietary similarity at the ordinal level generally were significant, although the amount of overlap between the northern (moist) and southern (xeric) populations of P. quadridens and P. portoricensis was not (Table. 1.10). Dietary overlap among all three species was less in the southern population (48.9%) than the northern population (55.5%; Table 1.10; Fig. 1.15). Between the wet season and dry season, dietary similarity between pairs of species ranged from 44.8 to 85%, and overlap among all three species was significantly greater during the wet season (55.7%) than the dry season (46.5%;

Table 1.11; Fig. 16).

DISCUSSION

Percent frequency of arthropods in the diet of mormoopid bats.—Eight orders of arthropods were found in the diet of mormoopids on Puerto Rico, with Coleoptera,

Hymenoptera, and Lepidoptera occurring most frequently. There were many differences, however, between my results and the previous reports of Rodríguez-Durán and Lewis (1987), who analyzed fecal pellets from Puerto Rico, and of Silva-Taboada (1979), who examined the contents of stomachs from Cuba (Table 1.4). For example, I found that M. blainvillei consumed the order Coleoptera 2–19 times as frequently as reported by Rodríguez-Durán and

Lewis (1987) and Silva-Taboada (1979), whereas P. quadridens in my study ate dipterans only about one third as often as indicated by Rodríguez-Durán and Lewis (1987) and Silva-

Taboada (1979). The most consistent discrepancy among studies concerned Hymenoptera.

My analysis showed that over 80% of the M. blainvillei and P. quadridens had eaten hymenopterans, but these insects were not detected in the diet of M. blainvillei and were

20 taken by only a few of the P. quadridens examined by Rodríguez-Durán and Lewis (1987) and Silva-Taboada (1979). I also found that ca. 74% of P. portoricensis had consumed hymenopterans, but Silva-Taboada (1979) reported their presence in only 7% of the stomachs taken from the closely related P. parnellii on Cuba.

Some of these differences in percent frequency of dietary items probably are related to the number of animals examined and other sampling considerations. For instance, I obtained pellets from 296 M. blainvillei, whereas Rodríguez-Durán and Lewis (1987) and

Silva-Taboada (1979) examined material provided by only 49–65 animals of this species.

Rodríguez-Durán and Lewis (1987:353) collected guano on only ―five dates during summer

1983,‖ whereas my data were collected on 30 dates, during both the wet and dry seasons of 2 calendar years. Furthermore, Rodríguez-Durán and Lewis (1987) obtained feces from a single location (Cucaracha Cave), whereas my sampling included locations in both mesic and xeric habitats.

Even though differences in sample size, time of year, and habitat may explain some of the variation in diet reported by different studies, it also is possible that the assemblage of insects found on Puerto Rico has changed since 1983, when guano was collected by

Rodríguez-Durán and Lewis (1987). Changes in habitat may increase resources for some species of insects, while decreasing the resources utilized by other species (Barberena-Arias and Aide, 2002). Hurricanes, in particular, can significantly alter habitats by producing landslides, floods, and large amounts of fallen wood (Barberena-Arias and Aide, 2002).

Since the study of Rodríguez-Durán and Lewis (1987), Puerto Rico has suffered severely from Hurricane Hugo (1989), Hurricane Georges (1998—Gannon et al., 2005), and

Hurricane Jeanne (2004—Franklin et al., 2006), as well as numerous tropical storms, and

21 these events have undoubtedly modified the vegetation in parts of the island. Following a major hurricane, reproduction of mature trees is often suppressed for months or even years, which allows for seedlings of pioneer species of plants to be favored in these areas (Brokaw and Grear, 1991; Fernández and Fletcher, 1991; and Gannon et al., 2005), altering the makeup of the habitat and perhaps, ultimately, the assemblage of insects in the vicinity.

Populations of many species of Lepidoptera, for example, can increase dramatically after a hurricane, due to the higher abundance of early successional vegetation (Barberena-Arias and

Aide, 2002; Torres, 1992). Similarly, the amount of wooded area has drastically changed; in

1991, ca. 28.2% of Puerto Rico was forested, whereas 39.7% of the island was composed of forest in 2000 (Parés-Ramos et al., 2008). Such differences in the land cover of Puerto Rico could lead to differences in the abundance and diversity of arthropods, ultimately altering the percent frequency of occurrence of the various orders of insects found within the diet of M. blainvillei and P. quadridens.

Although percent frequency is useful in understanding the diet of an animal, this measurement can be somewhat misleading, especially when few individuals are included in the study. In a sample of 50 bats, for instance, detection of a particular order of prey in just one additional bat will increase the overall frequency of occurrence of that order by 2%, even if only a trace of this order was found in the guano. Moreover, percent frequency is not necessarily indicative of the importance of that food to the overall energy budget of a bat.

For example, almost every M. blainvillei (98.7%) ate at least one coleopteran during my study, but this order made up only ca. 16% of the volume of the diet for this bat (Tables 1.3–

1.4). Similarly, 23.4% of all P. quadridens consumed Hemiptera, even though these insects amounted to less than 1% of the dietary volume (Tables 1.3–1.4). A thorough study of diet

22 should report both percent frequency and percent volume of the items of prey eaten by a predator.

Percent volume of arthropods in the diet of mormoopid bats.—Eight orders of arthropods were eaten by the mormoopids, with each species consuming 6–7 orders (Table

1.3). Coleoptera, Hymenoptera, and Lepidoptera formed the bulk of the diet, contributing to

84–93% of total volume for each mormoopid. Lepidopterans, however, dominated the foods of M. blainvillei, forming two thirds of the fecal volume, whereas coleopterans were the major dietary constituent for P. quadridens and P. portoricensis, accounting for ca. 41% of the arthropods that were eaten by both species. Only Coleoptera, Hymenoptera, and

Lepidoptera individually contributed more than 1% of total volume to the diets of M. blainvillei and P. quadridens, whereas six different orders provided at least 1% of the diet of

P. portoricensis.

As hypothesized, the overall composition of insects within the guano of

Puerto Rican mormoopids correlated mostly with the independent variable species, as demonstrated by CCA, thus providing statistical evidence for dietary partitioning among the three species. There are many potential reasons for these interspecific differences in diet. Freeman (1979, 1981), for instance, proposed that the structure of the jaws and associated muscles would determine whether bats could consistently forage on hard-bodied insects, such as coleopterans, or whether they would depend on soft-bodied prey, such as lepidopterans. Based on dietary information from the current study (Table 1.3), one would predict that the jaw of P. quadridens and P. portoricensis should be equally robust, to facilitate consumption of a large number of coleopterans, whereas the jaw of M. blainvillei likely is less robust, thus explaining

23 the low volume of hard coleopterans and the high volume of soft lepidopterans in its diet. However, based on a multivariate analysis of 18 characters of the cranium, dentary, and teeth, Rodríguez-Durán (1993) found that the structure of the skull among these three bats was opposite that expected from the observed diet. Although the two Pteronotus overlapped in the robustness of their skulls, the jaws and teeth of

M. blainvillei actually were more suited to hard foods than either of the Pteronotus.

This contradiction suggests that other factors affect the diet of an insectivorous bat, in addition to morphology of the skull.

Another factor that may account for some differences in diet is overall size of the body. A large animal should be less restricted in the size of prey it can consume, leading to a larger range of food sizes and a greater dietary diversity (Ashmole, 1968; Schoener, 1971), and thus, body size may help explain some of the differences between P. portoricensis, the largest of the three species, and the other two mormoopids. More orders of insects comprised at least 1% of the overall diet for P. portoricensis, and its diet contained six well-represented orders of insects compared to only three orders for the other two species. Moreover, dietary diversity (Simpson’s Index) of P. portoricensis was numerically the greatest of the three mormoopids (0.70), although significantly higher than that of only M. blainvillei (Table 1.7).

Also, the larger size of P. portoricensis may partly explain why large-bodied insects, like odonates (Corbet, 1980) and orthopterans (Whitman, 2008), formed a small, but consistent portion of their diet.

Variation in the morphology of the wings among species of bat can yield differences in flight and hunting behavior (Fenton et al., 1998; Jennings et al., 2004; Norberg and

Rayner, 1987) and ultimately diet. Based on aspect ratio, wing loading, and the wing-tip

24 index (a measure of the pointedness of the wings), Jennings et al. (2004) classified the three mormoopids as slow-flying aerial hawkers. However, M. blainvillei had the lowest tip index, indicating the least ability to hover or maneuver in highly cluttered spaces (Jennings et al.,

2004), and these bats are believed to forage around the crowns of trees and along woodland edges (Gannon et al., 2005). The shape of the wing is very similar in the two species of

Pteronotus, allowing these bats to forage in more highly cluttered areas than M. blainvillei

(Jennings et al., 2004), which may help explain the greater diversity in diet of the two

Pteronotus, compared with M. blainvillei. P. parnellii, which is closely related to P. portoricensis (Dávalos, 2006), often flies low to the ground near vegetation (Bateman and

Vaughan, 1974), and if P. portoricensis does the same, then this bat would have greater opportunities than M. blainvillei and P. quadridens to capture Odonata and Orthoptera, which are often low-flying (Tripplehorn and Johnson, 2005), grass-dwelling insects.

In addition, the time at which a bat emerges from its dayroost can influence the types of insects that are available to it (Rodríguez-Durán 1984; Fenton et al., 1998; Rydell et al.,

1996). P. portoricensis and P. quadridens both emerge from their caves near sunset (Gannon et al., 2005; Rodríguez-Durán and Lewis, 1987), giving these species the opportunity to consume both diurnal insects, which are still active near dusk, as well as nocturnal insects that will be in flight later in the evening. Emergence behavior by the bats may be an additional factor that accounts for the presence of Odonata and Orthoptera in the diet of

Pteronotus but not M. blainvillei (Table 1.4). Furthermore, moths typically are more active well after sunset (Rydell et al., 1996), and these insects form a greater proportion of dietary volume for the late-emerging M. blainvillei compared with the Pteronotus.

Effects of season and habitat.—In addition to species, the interspecific CCA showed

25 that season and habitat explained 1–3% of the overall variation in the volume of arthropods in the guano (Table 1.5). The explanatory power of each variable is low, but such a pattern

(< 10% of variation explained) is typical for ecological data, due to the presence-absence nature of the data and the resulting abundance of zeros that appear in such analyses (ter

Braak and Verdonschot, 1995). In my study, for instance, a sample of guano from an individual bat generally contained the remains of only one or a few orders of arthropods, with most of the eight orders not present.

The variables season and habitat also were generally significant in the intraspecific

CCA (Table 1.6), although there was no consistent trend in differences in the percent volume of insects consumed by each species between the wet and dry season or between moist and xeric habitats (Table 1.12). For example, the amount of Coleoptera was approximately the same between moist and xeric sites for M. blainvillei, moderately increased for P. quadridens, and strongly decreased for P. portoricensis. Similarly, the volume of

Lepidoptera between the wet and dry seasons was about the same for M. blainvillei, declined by 70% for P. quadridens, but increased by 100% for P. portoricensis. These results suggest that each mormoopid is taking advantage of changes in the abundance of different groups of appropriate insects (i.e., particular size, hardness, etc.) that occur within the foraging habitats of each species.

The amount of dietary differences in an assemblage of closely related species may differ between habitats if the environmental stressors, such as limiting resources, differ between the habitats, causing an increase in competition. The degree of dietary differences should be higher (i.e., overlap in diet should be less) in xeric areas, such as southwestern

Puerto Rico, where resources presumably are more limiting (Chesson and Huntly, 1997),

26 relative to other areas of the island that are more lush and diverse in vegetation, such as the northwestern coast. As predicted, the amount of dietary overlap between and among the mormoopids on Puerto Rico typically was greater in the moist than the xeric habitat (Table

1.10; Fig. 1.12). Similarly, dietary overlap was significantly higher during the wet season than the dry season (Table 1.11; Fig. 1.13).

Effects of age, reproductive condition, and sex.—The CCA indicated that the variables age and reproductive condition did not explain any of the variation in the composition of insects eaten by the Puerto Rican mormoopids, and that the variable sex only accounted for intraspecific variation in the diet of P. portoricensis (Tables 1.5–1.6).

Juveniles are inexperienced with the mechanics of flight and echolocation, and, consequently, they often fly in areas that are less cluttered with vegetation than do the adults (Adams, 1996,

1997; Anthony and Kunz, 1977). Because juveniles are more restricted in the types of habitats in which they can forage, dietary differences can be expected between ages. Such differences have been documented in a number of insectivorous species, including the little brown bat (Myotis lucifugus—Adams, 1996, 1997), big brown bat (Eptesicus fuscus—

Hamilton and Barclay, 1998), and hoary bat (Lasiurus cinereus—Rolseth et al., 1994). Even though no differences were found between ages for the mormoopids, this conclusion is tentative because of the small sample of juveniles, only 32 for M. blainvillei and 6 for P. quadridens (Table 1.2).

Reproductive condition did not affect the diet of the Puerto Rican mormoopids, even though such differences have been reported in a number of other insectivorous species (e.g.,

Anthony and Kunz, 1977; Kunz, 1974). Nevertheless, Swift et al. (1985) also documented a lack of dietary differentiation among reproductive conditions for the common pipistrelle

27

(Pipistrellus pipistrellus) in Scotland, as did Kunz et al. (1995) for the Brazilian free-tailed bat (Tadarida brasiliensis) in Texas. Similar to the Mormoopidae on Puerto Rico, the orders

Coleoptera, Hymenoptera, and Lepidoptera composed a significant amount of the diet for T. brasiliensis (Kunz et al., 1995). Kunz et al. (1995) claimed that the bodies of these insects are typically rich in energy and that the lack of differences among reproductive conditions was understandable, because all reproductive females should have their energetic demands met without having to switch to alternative prey.

The variable sex explained 15% of the variation in the composition of insects in the diet of P. portoricensis but not for the other two bats. Because male and female P. portoricensis appear equally abundant at my three study sites (Fig. 1.14), there should be increased pressure for these groups of individuals to differentiate their diets. Conversely, the skewed proportions of male and female M. blainvillei and P. quadridens at my three locations (Fig. 1.14) suggest that these species may segregate spatially to a certain degree, alleviating the pressure for dietary differences between sexes.

Closing remarks.—My study provides evidence for the partitioning of resources among the three mormoopid bats on Puerto Rico through means of dietary differentiation.

However, diets of each species quantitatively differed from those reported by earlier studies, between the wet season and the dry season, and between moist and xeric habitats. These differences suggest that mormoopid bats are somewhat adaptable and are not locked in to a specific suite of prey. Although certain types of arthropods are commonly eaten, the exact frequencies at which they are consumed and the proportion of the volume of the diet that they represent for each bat apparently varies, presumably with availability in the specific foraging sites frequented by each species. Insectivorous birds and bats are typically less impacted by

28 catastrophes, like hurricanes, than are frugivorous species that are directly dependent on mature trees and shrubs for their food (Jones et al., 2001; Waide, 1991), and the ability of the mormoopids to modify the composition of their diet should further aid their long-term survival in the Caribbean Basin.

Through CCA, the variables species, sex, habitat, and season explained ca. 20% of dietary variation found in this study. However, the remaining 80% of unexplained variation in the diet of the Puerto Rican Mormoopidae could be attributed to variables that were either not previously considered or were infeasible to analyze during this project. For example, it is likely that foraging habitat and behavior contribute to differences in the diet of these species of bat; however, due to rolling hills throughout the coastal regions of Puerto Rico (Gannon et al., 2005), studying the foraging behavior of these bats is extremely difficult, if not impossible.

Determining ordinal percent volume and dietary diversity through standard fecal analysis is a good start at indicating the range of prey that a species of bat can consume, but it does not allow biologists to understand fully the dietary requirements of an insectivorous bat.

For instance, if the percent volume or dietary diversity of two or more insectivorous bats is similar at the level of order, the diet of these bats is not necessarily identical at the level of family, genus, or species. Unfortunately, determining the identity of insects to such low taxonomic levels through visual fecal analysis is difficult and usually impossible, which ultimately limits the conclusions biologists can draw in dietary studies.

29

LITERATURE CITED

Abramsky, Z., M. I. Dyer, and P. D. Harrison. 1979. Competition among small mammals in experimentally perturbed areas of the shortgrass prairie. Ecology 60:530–536.

Adams, R. A. 1996. Size-specific resource use in juvenile little brown bats, Myotis lucifugus (Chiroptera: Vespertilionidae): is there an ontogenetic shift? Canadian Journal of Zoology, 74:1204–1210.

Adams, R. A. 1997. Onest of volancy and foraging patterns of juvenile little brown bats, Myotis lucifugus. Journal of Mammalogy, 78:239–246.

Aldridge, H., and I. L. Rautenbach. 1987. Morphology, echolocation and resource partitioning in insectivorous bats. Journal of Animal Ecology 56:763–778.

Anthony, E. L. P. 1988. Age determination in bats. Pp.47–58, in Ecological and behavioral methods for the study of bats (T.H. Kunz, ed.). Smithsonian Institution Press, Washington, D.C.

Anthony, E. L. P., and T. H. Kunz. 1977. Feeding strategies of the little brown bat, Myotis lucifugus, in southern New Hampshire. Ecology, 58:775–786.

Ashmole, N. P. 1968. Body size, prey size, and ecological segregation in five sympatric tropical terns (Aves: Laridae). Systematic Zoology, 17: 292-304.

Banna, A. L., and S. L. Gardner. 1996. Nematode diversity of native species of Vitis in California. Canadian Journal of Zoology, 74: 971–982.

Barberena-Arias, M. F., and T. M. Aide. 2002. Variation in species and trophic composition of insect communities in Puerto Rico. Biotopica, 34:357–367.

Barclay, R. M. R. 1994. Constraints on reproduction by flying vertebrates: energy and calcium. American Naturalist, 144:1021–1031.

Bateman, G. C., and T. A. Vaughan. 1974. Nightly activities of Mormoopid bats. Journal of Mammalogy, 55: 45–65.

Black, H. L. 1974. A north temperate bat community: structure and prey populations. Journal of Mammalogy, 55:138–157.

Brokaw, N. V. L., and J. S. Grear. 1991. Forest structure before and after Hurricane Hugo at three elevations in the Luquillo Mountains, Puerto Rico. Biotropica, 23:386–392.

Brower, J. E., and J. H. Zar. 1984. Field and laboratory methods for general ecology. Second ed. Wm. C. Brown Publishers, Dubuque, Iowa.

30

Carter, T. C., M. A. Menzel, B. R. Chapman, and K. V. Miller. 2004. Partitioning of food resources by syntopic eastern red (Lasiurus borealis), Seminole (L. seminolus) and evening (Nycticeius humeralis) bats. American Midland Naturalist, 151:186–191.

Chesson, P., and N. Huntly. 1997. The roles of harsh and fluctuating conditions in the dynamics of ecological communities. American Naturalist, 150: 519–553.

Christopher, C. C., and G. W. Barrett. 2006. Coexistence of white-footed mice (Peromyscus leucopus) and golden mice (Ochrotomys nuttalli) in a southeastern forest. Journal of Mammalogy, 87:102–107.

Conde-Costas, C., and C. González. 1990. Las cuevas y cavernas en el bosque xerofítico de Guánica. Acta Científica, (1–3): 113–126.

Corbet, P. S. 1980. Biology of Odonata. Annual Review of Entomology, 25:189–217.

Danielson, B. J. 1991. Communities in a landscape: the influence of habitat heterogeneity on the interactions between species. American Naturalist, 138:1105–1120.

Dávalos, L. M. 2006. The geography of diversification in the mormoopids (Chiroptera: Mormoopidae). Biological Journal of the Linnean Society, 88:101–118.

Dueser, R. D., and J. G. Hallett. 1980. Competition and habitat selection in a forest-floor small mammal fauna. Oikos, 35:293–297.

Dueser, R. D., and H. H. Shugart. 1979. Niche pattern in a forest floor small-mammal fauna. Ecology, 60:108–118.

Dunson, W. A., and J. Travis. 1991. The role of abiotic factors in community organization. American Naturalist: 138:1067–1091.

Fenton, M. B., I. L. Rautenbach, J. Rydell, H. T. Arita, J. Ortega, S. Bouchard, M. D. Hovorka, B. Lim, E. Odgren, C. V. Portfors, W. M. Scully, D. M. Syme, and M. J. Vonhof. 1998. Emergence, echolocation, diet and foraging behavior of Molossus ater (Chiroptera: Molossidae). Biotropica, 30:314–320.

Fernández, D. S., and N. Fletcher. 1991. Changes in light availability following Hurricane Hugo in a subtropical montane forest in Puerto Rico. Biotropica, 23:393–399.

Findley, J. S. 1993. Bats: a community perspective. Cambridge University Press, New York.

Findley, J. S., and H. Black. 1983. Morphological and dietary structuring of a Zambian insectivorous bat community. Ecology, 64:625–630.

31

Franklin, J. L., R. J. Pasch, L. A. Avila, J. L. Beven, II, M. B. Lawrence, S. R. Stewart, and E. S. Blake. 2006. Atlantic hurricane season of 2004. Monthly Weather Review, 134: 981–1025.

Freeman, P. W. 1979. Beetle-eating and moth-eating molossid bats. Journal of Mammaology, 60:467–479.

Freeman, P. W. 1981. Correspondence of food habits and morphology in insectivorous bats. Journal of Mammalogy, 62:166–173.

Gannon, M. R., A. Kurta, A. Rodríguez-Durán and M. R. Willig. 2005. Bats of Puerto Rico: an island focus and a Caribbean perspective. Texas Tech University Press, Lubbock, Texas.

Gonzalez, R. L. 2009. Regiones o provincias geomórficas. . Accessed 30 June, 2011.

Groom, M. J. 2006. Threats to biodiversity. Pp. 61–109, in Principles of conservation biology (M.J. Groom, G.K. Meffe, and C.R. Carroll, eds.). Third ed. Sinauer Associates, Sunderland, Massachusetts.

Hamilton, I. M., and R. M. R. Barclay. 1998. Diets of juvenile, yearling, and adult big brown bats (Eptiscus fuscus) in southeastern Alberta. Journal of Mammalogy, 79:764–771.

Hanwell, A., and M. Peaker. 1977. Physiological effects of lactation on the mother. Symposia of the Zoological Society of London, 41:297–312.

Haroon, A. K. Y, and K. A. Pittman. 1998. Intraspecific dietary breadth, overlap indices and feeding strategies of Puntius gonionotus Bleeker and Oreochromis spp. in a shallow pond from Bangladesh. Asian Fisheries Science, 10:303–316.

Hart, D .D. 1992. Community organization in streams: the importance of species interactions, physical factors, and chance. Oecologia, 91:220–228.

Herd, R. M. 1983. Pteronotus parnellii. Mammalian Species, 209:1–5.

Heske, E. J., J. H. Brown, and S. Mistry. 1994. Long-term experimental study of a Chihuahuan Desert rodent community: 13 years of competition. Ecology, 75:438– 445.

Hickey, M. B. C., L. Acharya, and S. Pennington. 1996. Resource partitioning by two species of vespertilionid bats (Lasiurus cinereus and Lasiurus borealis) feeding around street lights. Journal of Mammalogy, 77:325–334.

Holdridge, L. R. 1967. Life zone ecology. Tropical Science Center. San Jose, Costa Rica.

32

Hope, A. C. A. 1968. A simplified Monte Carlo significance test procedure. Journal of Royal Statistical Society, Series B (Methodological), 30:582–589.

Husar, S. L. 1976. Behavioral character displacement: evidence of food partitioning in insectivorous bats. Journal of Mammalogy 57:331–338.

Janzen, D. H., and T. W. Schoener. 1968. Differences in insect abundance and diversity between wetter and drier sites during a tropical dry season. Ecology, 49:96–110.

Jennings, N. V., S. Parsons, K .E. Barlow, and M. R. Gannon. 2004. Echolocation calls and wing morphology of bats from the West Indies. Acta Chiropterologica, 6: 75–90.

Jones, K. E., K. E. Barlow, N. Vaughn, A. Rodríguez-Durán, and M. R. Gannon. 2001. Short-term impacts of extreme environmental disturbance on the bats of Puerto Rico. Animal Conservation, 4:59–66.

Kalcounis-Rüppell, M. C., and J. S. Millar. 2002. Partitioning of space, food, and time by syntopic Peromyscus boylii and P. californicus. Journal of Mammalogy, 83:614–625.

Kinahan, A. A., and N. Pillay. 2008. Does differential exploitation of folivory promote coexistence in an African savanna granivorous rodent community? Journal of Mammalogy, 89:132–137.

Knuth, B. A., and G. W. Barrett. 1984. A comparative study of resource partitioning between Ochrotomys nuttalli and Peromyscus leucopus. Journal of Mammalogy, 65:576–583.

Kozlowski, A. J., E. M. Gese, and W. M. Arjo. 2008. Niche overlap and resource partitioning between sympatric kit foxes and coyotes in the Great Basin Desert of western Utah. American Midland Naturalist, 160:191–208.

Kunz, T. H. 1973. Resource utilization: temporal and spatial components of bat activity in central Iowa. Journal of Mammalogy, 54:14–32.

Kunz T. H. 1974. Feeding ecology of a temperate insectivorous bat (Myotis velifer). Ecology, 55:693–711.

Kunz, T. H., and A. Kurta. 1988. Capture methods and holding devices. Pp. 1–29, in Ecological and behavioral methods for the study of bats (T.H. Kunz, ed.). Smithsonian Institution Press, Washington, D.C..

Kunz, T. H., J. O. Whitaker, Jr., M. D. Wadanoli. 1995. Dietary energetics of the Mexican free-tailed bat (Tadarida brasiliensis) during pregnancy and lactation. Oecologia, 101:406–415.

Kurta, A., T. H. Kunz, and K. A. Nagy. 1990. Energetics and water flux of free-ranging big brown bats (Eptesicus fuscus) during pregnancy and lactation. Journal of Mammalogy, 71:59–65.

33

Kurta, A., G. P. Bell, K. N. Nagy, and T. H. Kunz. 1989. Energetics of pregnancy and lactation in free-ranging little brown bats (Myotis lucifugus). Physiological Zoology, 62:804–818.

Lancaster, W. C., and E. K. V. Kalko. 1996. Mormoops blainvillii. Mammalian Species, 544:1–5.

Lande, R. 1996. Statistics and partitioning of species diversity, and similarity among multiple communities. Oikos, 76:5–13.

MacArthur, R. H., and E. O. Wilson. 1967. The theory of island biogeography. Princeton University Press, Princeton, New Jersey.

Magnan, P., M. A. Rodríguez, P. Legendre, and S. Lacasse. 1994. Dietary variation in a freshwater fish species: relative contributions of biotic interations, abiotic factors, and spatial structure. Canadian Journal of Fisheries and Aquatic Sciences, 51:2856–2865.

McGarigal, K., S. Cushman, and A. Stafford. 2000. Multivariate statistics for wildlife and ecology research. Springer-Verlag, New York.

Morris, D. W. 1984. Microhabitat separation and coexistence of two temperate-zone rodents. Canadian Field-Naturalist, 98:215–218.

Morris, D. W. 1999. A haunting legacy for isoclines: mammal coexistence and the ghost of competition. Journal of Mammalogy, 80:375–385.

Murphy, P. G., and A. E. Lugo. 1986. Structure and biomass of a subtropical dry forest in Puerto Rico. Biotropica, 18:89–96.

Murray, S., and A. Kurta. 2002. Spatial and temporal variation in diet. Pp. 182– 92, in The Indiana Bat: Biology and Management of an Endangered Species (A. Kurta, J. Kennedy, eds.). Austin, Texas: Bat Conservation International.

Norberg, U. M., and J. M. V. Rayner. 1987. Ecological morphology and flight in bats (Mammalia: Chiroptera): wing adaptations, flight performance, foraging strategy and echolocation. Philosophical Transactions of the Royal Society of London. Series B, Biological Sciences, 316:335–427.

Palmeirim, J. M., and L. Rodrigues. 1993. The 2-minute harp trap for bats. Bat Research News, 34:60–64.

Parés-Ramos, I. K., W. A. Gould, and T. M. Aide. 2008. Agricultural abandonment, suburban growth, and forest expansion in Puerto Rico between 1991 and 2000. Ecology and Society, 13:1–19.

Placer, J. J. 1998. The bats of Puerto Rico. Bats, 16:12–15.

34

Patterson, B. D., M. R. Willig, and R. D. Stevens. 2003. Trophic strategies, niche partitioning, and patterns of ecological organization. Pp. 536–579, in Bat ecology (T.H. Kunz and M.B. Fenton, eds.). University of Chicago Press, Chicago, Illinois.

Persson, L., and S. Diehl. 1990. Mechanistic individual-based approaches in the population/community ecology of fishes. Annales Zoologici Fennici, 27:165–182.

Racey, P. A. 1988. Reproductive assessment in bats. Pp. 31–46, in Ecological and behavioral methods for the study of bats (T.H. Kunz, ed.). Smithsonian Institution Press, Washington, D.C..

Rakotoarivelo, A. A., N. Ranaivoson, O. R. Ramilijaona, A. F. Kofoky, P. A. Racey, and R. K. B Jenkins. 2007. Seasonal food habits of five sympatric forest microchiropterans in western Madagascar. Journal of Mammalogy, 88:959–966.

Reith, C. C. 1980. Shifts in times of activity of Lasionycteris noctivagans. Journal of Mammalogy, 61:104–108.

Rodríguez-Durán, A. 1984. Community structure of a bat colony at Cueva Cucaracha. M.Sc. thesis. University of Puerto Rico, Mayaguez, Puerto Rico.

Rodríguez-Durán, A. 1996. Foraging ecology of the Puerto Rican boa (Epicrates inornatus): bat predation, carrion feeding, and piracy. Journal of Herpetology, 30:533–536.

Rodríguez-Durán, A. 2009. Bat assemblages in the West Indies: the role of caves. Pp. 265– 279, in Island bats: evolution, ecology, and conservation. (T.H. Fleming and P.A. Racey, eds.). University of Chicago Press, Chicago, Illinois..

Rodríguez-Durán, A., and T. H. Kunz. 1992. Pteronotus quadridens. Mammalian Species, 395:1–4.

Rodríguez-Durán, A., and A. R. Lewis. 1987. Patterns of population size, diet, and activity time for a multispecies assemblage of bats at a cave in Puerto Rico. Caribbean Journal of Science, 23:352–360.

Rodríguez-Durán, A, A. R. Lewis, and Y. Montes. 1993. Skull morphology and the diet of Antillean bat species. Caribbean Journal of Science, 29:258–261.

Rolseth, S. L., C. E. Koehler, and R. M. R. Barclay. 1994. Differences in the diets of juvenile and adult hoary bats, Lasiurus cinereus. Journal of Mammalogy, 75:394–398.

Ruszczyk, A., and A. M. D. Araujo. 1992. Gradients in butterfly species diversity in an urban area in Brazil. Journal of the Lepidopterists' Society, 46:255–264.

Rydell, J., A. Entwistle, and P. A. Racey. 1996. Timing of foraging flights of three species of bats in relation to insect activity and predation risk. Oikos, 76:243–252.

35

Saunders, M. B., and R. M. R. Barclay. 1992. Ecomorphology of insectivorous bats: a test of predictions using two morphologically similar species. Ecology, 73:1335–1345.

Schoener, T. W. 1971. Theory of feeding strategies. Annual Review of Ecology, Evolution, and Systematics, 11:369–404.

Schoener, T. W. 1982. The controversy over interspecific competition. American Scientist, 70:586–595.

Schoener, T. W. 1986. Mechanistic approaches to community ecology: a new reductionism? American Zoologist, 26:81–106

Schnitzler, H., and E. K. V., Kalko. 2001. Echolocation by insect-eating bats. BioScience, 51:557–569.

Seagle, S. W. 1985. Competition and coexistence of small mammals in an east Tennessee pine plantation. American Midland Naturalist, 114:272–282.

Silva-Taboada, G. 1979. Los murciélagos de Cuba. Academia de Ciencies de Cuba, Havana, Cuba.

Soto-Centeno, J. A., and A. Kurta. 2006. Diet of two nectarivorous bats, Erophylla sezekorni and Monophyllus redmani (Phyllostomidae), on Puerto Rico. Journal of Mammalogy, 87:19–26.

Spatial Climate Analysis Service. 2002. . Accessed: 1 August, 2011.

Sparks, D. W., and E. W. Valdez. 2003. Food habits of Nyctinomops macrotis at a maternity roost in New Mexico, as indicated by analysis of guano. Southwestern Naturalist, 48:132–135.

Swift, S. M., P. A. Racey, and M.I. Avery. 1985. Feeding ecology of Pipistrellus pipistrellus (Chiroptera: Vespertilionidae) during pregnancy and lactation. II. Diet. Journal of Animal Ecology, 54:217–225. ter Braak, C. J. F. 1986. Canonical correspondence analysis: a new eigenvector technique for multivariate direct gradient analysis. Ecology, 67:1167–1179. ter Braak, C. J. F., and P. F. M. Verdonschot. 1995. Canonical correspondence analysis and related multivariate methods in aquatic ecology. Aquatic Sciences, 57:255–289.

Torres, J. A. 1992. Lepidoptera outbreaks in response to successional changes after the passage of Hurricane Hugo in Puerto Rico. Journal of Tropical Ecology, 8:525–535.

36

Triplehorn, C. A., and N. F. Johnson. 2005. Borror and DeLong's introduction to the study of insects. Seventh ed. Thomas Brooks/Cole Publishing, Pacific Grove, California.

Waide, R. B. 1991. The effect of Hurricane Hugo on bird populations in the Luquillo Experimental Forest, Puerto Rico. Biotropica, 23:475–480.

Whitaker, J. O., Jr. 1988. Food habits analysis of insectivorous bats. Pp. 171–189, in Ecological and behavioral methods for the study of bats (T.H. Kunz, ed.). Smithsonian Institution Press, Washington, D.C..

Whitaker, J. O., Jr. 1994. Food availability and opportunistic versus selective feeding in insectivorous bats. Bat Research News, 35:75–77.

Whitaker, J. O., Jr. 2004. Prey selection in a temperate zone insectivorous bat community. Journal of Mammalogy, 85: 460–469.

Whitaker, J. O., Jr., and P. Clem. 1992. Food of the evening bat Nycticeius humeralis from Indiana. American Midland Naturalist, 127:211–214.

Whitaker, J. O., Jr., C. Neefus, and T. H. Kunz. 1996. Dietary variation in the Mexican free- tailed bat (Tadarida brasiliensis). Journal of Mammalogy, 77:716–724.

Whitman, D. W. 2008. The significance of body size in the Orthoptera: a review. Journal of Orthoptera Research, 17:117–134.

Wolff, J. O., and R. D. Dueser. 1986. Noncompetitive coexistence between Peromyscus spp. and Clethrionomys gapperi. Canadian Field-Naturalist, 100:186–191.

37

TABLE 1.1.—Total number pellets that were used for standard fecal analysis, broken down by species, age, sex, and reproductive condition.

Male Male Female Post- Non- Total Total Species Species Pregnant Lactating (juvenile) (adult) (juvenile) lactating reproductive male female total M. blainvillei 41 733 60 5 50 10 415 774 540 1,314 P. quadridens 3 248 3 101 278 13 122 251 517 768 P. portoricensis 31 38 49 31 87 118

38

38

TABLE 1.2.—Total number of individuals from which 2,200 pellets were obtained for standard fecal analysis, according to by species, age, sex, and reproductive condition.

Male Male Female Post- Non- Total Total Species Species (juvenile) (adult) (juvenile) Pregnant Lactating lactating reproductive male female total M. blainvillei 12 160 20 2 14 2 86 172 124 296 P. quadridens 3 82 3 37 67 8 35 85 150 235 P. portoricensis 10 9 17 10 26 36

39

39

TABLE 1.3.—Percent frequency (%) of occurrence of arthropods in the diet of M. blainvillei, P. quadridens, and P. portoricensis, as well as P. parnellii, a closely related congener of P. portoricensis. Sources of information are: 1, this study (fecal samples); 2, Rodríguez-Durán and Lewis, (1987—fecal samples from Puerto Rico); and 3, Silva-Taboada, (1979—stomach contents from Cuba).

__ M. blainvillei ____ P. quadridens __ P. portoricensis_ P. parnellii _ Order 1 (n = 296) 2 (n = 65) 3 (n =49 ) 1 (n = 235) 2 (n =158 ) 3 (n = 195) 1 (n = 36) 3 (n = 47) Coleoptera 98.7 45 5.3 98.7 83 75 100 14.3 Dermaptera 1

Diptera 7.7 14 14.9 45 41.2 21.1 21.4

Dictyoptera 1

Hemipteraa 24.2 34 10.5 23.4 82 14.2 42.1

Hymenoptera 82.9 87.7 16 1 73.7 7.1

Lepidoptera 99.7 90 100 89.8 84 17.6 89.5 93 Odonata 1.3 15.8 14.3

40 Orthoptera 1.3 0.4 16.2 13.4 14.3

Araneae 1 0.8 a Contains members of the previously recognized order Homoptera.

40

TABLE 1.4.—Average percent volume (%) of arthropods consumed by each species of mormoopid.

Order M. blainvillei P. quadridens P. portoricensis Coleoptera 16.1 41.8 40.4 Diptera 0.2 0.2 1.6 Hemiptera 0.7 0.8 4.2 Hymenoptera 11.7 27.0 17.9 Lepidoptera 66.7 24.3 27.4 Odonata 1.8

Orthoptera 0.1 0.7

Araneae 0.1 0.1

Unknown 4.5 5.5 4.8 Hair, rocks, dirt etc. 0.2 1.2

41

TABLE 1.5.—Ranking of the independent variables in importance by their effect on the composition of insects in the guano of all three mormoopids.

Variable Percent variation explained by variable P Age 0.0 0.06 Reproductive condition 0.0 0.06 Habitat 1.0 0.001 Season 1.0 0.001 Sex 3.0 0.002 Species 16.0 0.001

42

TABLE 1.6.—Ranking of the independent variables in importance by their effect on the composition of insects in the guano of M. blainvillei.

Percent variation Species Variable P explained by variable M. blainvillei Sex 0.0 0.17 Habitat 1.0 0.002

Season 1.0 0.001

P. quadridens Sex 1.0 0.19 Habitat 2.0 0.001

Season 3.0 0.001

P. portoricensis Season 6.0 0.40 Habitat 10.0 0.05

Sex 15.0 0.001

43

TABLE 1.7.—Use of a t-test with infinite degrees of freedom to compare dietary diversity (Simpson's Indices) between pairs of mormoopid species.

Comparison t Pa M. blainvillei and P. portoricensis -3.8 <0.001 M. blainvillei and P. quadridens -3.4 <0.001 P. portoricensis and P. quadridens 1.1 0.13 a Alpha was set to 0.016, using a Bonferroni correction.

44

TABLE 1.8.—Use of a t-test with infinite degrees of freedom to compare dietary diversity (Simpson's Indices) of each mormoopid species between moist and xeric parts of Puerto Rico.

Species Moist habitat Xeric habitat t P M. blainvillei 0.47 0.56 -1.3 0.09 P. quadridens 0.65 0.67 -0.4 0.33 P. portoricensis 0.70 0.64 0.09 0.46

45

TABLE 1.9. — Use of a t-test with infinite degrees of freedom to compare dietary diversity (Simpson's Indices) between the wet season and dry season for each species of mormoopid.

Species Wet season Dry season t P M. blainvillei 0.48 0.52 -0.5 0.3 P. quadridens 0.67 0.55 2.5 0.006 P. portoricensis 0.67 0.69 -0.4 0.35

46

TABLE 1.10.— Use of a t-test with infinite degrees of freedom to compare the percent dietary overlap (Schoener's Index) between northern and southern populations for each pair of mormoopids and among the three species.

Moist Xeric Comparison t Pa habitat habitat M. blainvillei and P. portoricensis 71.9 47.7 4.3 <0.001 M. blainvillei and P. quadridens 55.5 66.6 -3.2 <0.001 P. quadridens and P. portoricensis 77.3 75.1 0.6 0.3 M. blainvillei, P. quadridens, and P. portoricensis 55.5 48.9 7.4 <0.001 a Alpha was set to 0.013, using a Bonferroni correction.

47

TABLE 1.11.— Use of a t-test with infinite degrees of freedom to compare the dietary overlap (Schoener's Index) between the wet season and the dry season for each pair of mormoopids and among the three species.

Wet Dry Comparison t Pa season season M. blainvillei and P. portoricensis 58.2 66.9 -1.6 0.06 M. blainvillei and P. quadridens 63.5 44.8 5.2 <0.001 P. quadridens and P. portoricensis 85.0 72.0 2.7 0.004 M. blainvillei, P. quadridens, and P. portoricensis 55.7 46.5 3.8 <0.001 a Alpha was set to 0.013, using a Bonferroni correction.

48

TABLE 1.12.—Percent volume (%) of the orders of insects consumed by each of the Puerto Rican mormoopids with respect to habitat as well as season.

Species Order of insect Moist habitat Xeric habitat Wet season Dry season M. blainvillei (267, 29, 181,115)a Coleoptera 16.1 14.8 15.5 16.7 Lepidoptera 67.1 59.3 68.3 63.3 Hymenoptera 11.3 14.6 9.5 15.1 P. quadridens (186, 49, 183, 52) Coleoptera 43.6 35.3 37.4 57.7 Lepidoptera 21.2 36 26.7 15.4 Hymenoptera 28.3 22.1 29.8 17.1 P. portoricensis (19, 17, 23, 13) Coleoptera 29.9 52.3 46.3 30 Lepidoptera 36.9 16.5 20.1 40

49 Hymenoptera 18.6 16.8 19 15.5

Diptera 0.4 2.8 2 0.6 Hemiptera 4.4 3.9 4.2 3.9 Odonata 2.6 1.5 0.9 3.4

Orthoptera 0.9 1.1 a Numbers in parentheses indicate number of bats from which feces was obtained in the moist habitat, xeric habitat, wet season, and dry season, respectively.

49

FIG. 1.1.—Three major physiographic regions of Puerto Rico (Gonzales, 2009; adjusted by A. Rolfe).

50

FIG. 1.2.—Average annual precipitation in Puerto Rico (Spatial Climate Analysis Service, 2002; adjusted by A. Rolfe).

51

FIG. 1.3.—Distribution of life zones in Puerto Rico (Gannon et al., 2005; adjusted by A. Rolfe).

52

FIG. 1.4.—Subtropical moist forest, characteristic of northern Puerto Rico. Photo by A. Rolfe.

53

FIG. 1.5.—Subtropical dry forest typical of the southern Puerto Rico. Photo by A. Rolfe.

54

FIG. 1.6.—Segregation of the three mormoopid bats within an idealized hot cave (Rodríguez-Durán, 2009; adjusted by A. Rolfe).

55

FIG. 1.7.—Portrait of the Antillean ghost-faced bat (Mormoops blainvillei). Photo by A. Rolfe.

56

FIG. 1.8.—Portrait of the sooty mustached bat (Pteronotus quadridens). Photo by A. Rolfe.

57

FIG. 1.9.—Portrait of the Puerto Rican mustached bat (Pteronotus portoricensis). Photo by A. Rolfe.

58

FIG. 1.10.—Approximate location of the three caves where mormoopids were captured. A) Culebrones Cave, B) Cucaracha Cave, and C) Murciélagos Cave. Image obtained from Google Maps.

59

FIG. 1.11.—Centroids for the overall diet of each mormoopid species. The direction in which the arrows point are indicative of the maximum change in the value associated with that variable. Therefore, the diet of M. blainvillei is heavily influenced by Lepidoptera and the diet of P. quadridens is heavily influenced by Coleoptera.

60

100%

90%

80%

70%

60%

50% All individuals 40% NorthMoist habitat SouthhabitatXeric habitat

30% Percentage dietary overlap of Percentage

20%

61

10%

0% M. blainvillei and M. blainvillei and P. quadridens and M. blainvillei, P. quadridens P. portoricensis P. portoricensis P. quadridens, and P. portoricensis

FIG. 1.12.—Percent of dietary overlap between and among all the Puerto Rican mormoopids, as well as the populations found in the moist habitat and xeric habitat.

61

90%

80%

70%

60%

50%

40% Wet Season Dry Season

30%

62

Percentage of dietary overlap dietary Percentage of 20%

10%

0% M. blainvillei and M. blainvillei and P. quadridens and M. blainvillei, P. quadridens P. portoricensis P. portoricensis P. quadridens and P. portoricensis

FIG. 1.13.—Percent of dietary overlap between and among all the Puerto Rican mormoopids, between the wet season and the dry season.

62

500

450

400

350

300

250 Male Female 200

150 Number of Individuals Captured Individuals Number of 100

50

0 M. blainvillei P. quadridens P. portoricensis

FIG. 1.14.—Number of males and females captured during this study for each of the Puerto Rican mormoopids.

63

CHAPTER 2 MOLECULAR ANALYSIS OF FECES

64

INTRODUCTION

Knowledge of diet is basic information needed for proper management of any species, and such information is particularly critical for maintaining populations of endemic or endangered species (Soto-Centeno and Kurta, 2006). Many oceanic islands, such as

Puerto Rico, are rich in endemic species, but these islands are also poorer in overall species richness than assemblages on the mainland (Groom, 2006). Populations on islands are usually small and subject to local catastrophes, such as tropical storms, that affect the availability of shelter and food, thus contributing to high rates of extinction on islands

(MacArthur and Wilson, 1967).

Bats represent ca. 20–25% of all species of mammals on earth (Wilson and Reeder,

2005), and they often are the only terrestrial mammals that are resident on oceanic islands

(Groom, 2006). Although the diet of bats is highly varied and can include fruit, nectar, fish, blood, and other bats, most species are insectivorous (Schnitzler and Kalko, 2001).

Insectivorous bats thoroughly chew their food, and it generally is impossible to identify fragments of prey found in feces beyond the level of order or occasionally family (Clare et al., 2009). Consequently, studies investigating dietary partitioning among species of bats through standard fecal analysis are potentially misleading, because even though two species of bat may consume equal quantities of an order, they may not be eating insects in the same family, genus, or species. Similarly, it is impossible for wildlife managers to comprehend completely the dietary requirements of any species without knowing the exact prey that an animal consumes in the wild.

Feces and isolating the mtDNA of prey.—Newly developed molecular approaches allow biologists to examine predator-prey relationships in more detail by targeting trace

65 amounts of deoxyribonucleic acid (DNA) from prey that are found within feces gathered in the field (Clare et al., 2009). Although DNA tends to degrade rapidly during digestion (Zaidi et al., 1999), small, fast-evolving, multi-copy gene regions found within mitochondrial DNA

(mtDNA—Bensasson et al., 2001; Symondson, 2002) regularly survive, and techniques based on polymerase chain reactions (PCR) can be used to amplify these genes with low levels of contamination (Clare et al., 2009).

Isolating and amplifying mtDNA of organisms is more successful than if nuclear

DNA is used, because more mtDNA is available in a cell compared with nuclear DNA. In addition, mtDNA has a faster mutation rate than nuclear DNA, due to the greater turnover of mitochondria within cells (Brown, 1979). The rapid rate of evolution, haploid mode of inheritance, and lack of individual variation in mtDNA permit biologists to decipher whether individuals or groups of individuals are closely related or sufficiently distinct that they should be classified as separate species (Stone et al., 2001). Similarly, the quick rate of mutation in mtDNA also allows biologists to assign a sample of mtDNA to a particular species.

An extremely rapid mutation rate in the mtDNA of individuals, however, could lead to designating two organisms as different species, even though they are just distantly related conspecifics. To combat this potential error, many studies focus on isolating and sequencing the cytochrome c oxidase subunit I (COI) gene of the mtDNA, because it appears to be the most conservative protein-coding gene of the mitochondrial genome (Clare et al., 2009,

2011; Folmer et al., 1994; Hebert et al., 2003; Zeale et al., 2010). This process of sequencing a standard genetic region, such as COI, to aid in identification of species is commonly referred to as ―DNA barcoding‖ (Hebert et al., 2003; Hebert et al., 2004), and this

66 process can have many applications, including analyzing the diet of animals (Clare et al.,

2009, 2011; Farrell et al., 2000; King et al., 2008; Symondson, 2002; Zeale et al., 2010).

Despite the thorough mastication of insects eaten by bats, isolating mtDNA from fragments of insects within the guano of a bat and amplifying it via PCR make it possible to obtain a species-level identification of the prey. Although use of PCR and sequencing of mtDNA are not novel to dietary analysis (Farrell et al., 2000; King et al., 2008; Symondson,

2002), few published reports have applied such approaches to bats (Clare et al., 2009, 2011;

Zeale et al., 2010), and no previous studies have used these molecular procedures to examine the diet of any neotropical species of bat.

Bats as potential biological control agents.—Through the process of barcoding mtDNA, it is possible to assess whether bats consume insects of economic importance.

Various insects pose a threat to humans by competing for food, fiber, and timber; however, with an effective method of control, such as natural predation, densities of many pests can be kept at low levels (Cleveland et al., 2006). A decline in the natural pest-control services provided by insect-eating organisms could have serious economic, environmental, and human-health implications (Cleveland et al., 2006) by increasing the need for pesticides.

Because herbivory by insects constrains reproduction by plants and also impacts their diversity and distribution (Kalka et al., 2008), bats may have a positive effect on the growth and economic success of agricultural crops by reducing the abundance of pest species. Kalka et al. (2008) and Williams-Guillén et al. (2008) found an increased number of herbivorous arthropods in areas where bats were excluded, thus demonstrating the importance of bats as biological-control agents for economically important pests (Boyles et al., 2011; Kunz et al.,

2011), and Boyles et al. (2011) recently estimated that the value of pest-control services

67 provided by bats to agriculture in the United States was $23 billion annually. Efforts at conservation of animals typically concentrate on charismatic megafauna, such as bald eagles

(Haliaeetus leucocephalus) or timber wolves (Canis lupus), but if a species of bat can be shown to be a biological control agent for an agricultural pest or a vector for a human disease, then this service provided by the bat would be a powerful economic force promoting the conservation of an otherwise noncharismatic species.

Description of study species and their diet.—This study examined the diet of two species of insectivorous bat in the family Mormoopidae. The Mormoopidae is a small, neotropical family, the members of which are characterized by flap-like outgrowths of skin below the lower lip and funnel-shaped ears (Dávalos, 2006). Of the eight species in this family, three species are found on Puerto Rico: the Antillean ghost-faced bat (Mormoops blainvillei), sooty mustached bat (Pteronotus quadridens), and Puerto Rican mustached bat

(Pteronotus portoricensis—Gannon et al., 2005). M. blainvillei and P. quadridens are endemic to the Greater Antilles (Gannon et al., 2005), whereas P. portoricensis is endemic to

Puerto Rico (Dávalos, 2006). Like most insectivorous bats, mormoopids are small, with P. portoricensis being the largest member of the family on Puerto Rico at 10–18 g, followed by

M. blainvillei at 8–11 g, and P. quadridens at 4–7 g (Gannon et al., 2005). Mormoopid bats typically roost in caves, where thousands and often hundreds of thousands of individuals from multiple species reside (Silva-Taboada, 1979). This study focused on M. blainvillei and

P. quadridens, two of the most abundant species of insectivorous bat on Puerto Rico and throughout the Greater Antilles.

Purpose of this study.—The primary goals of this study were to provide species-level information on prey consumed by M. blainvillei and P. quadridens through the use of

68 molecular techniques and to determine whether these bats are potential biological-control agents for insects that are important to the Puerto Rican economy. Secondary goals of this investigation were to examine the efficacy of different field and laboratory methods on the ability to extract and amplify DNA contained in the feces of insectivorous bats. Specifically, this study compared the ability to amplify the mtDNA of insects after three methods of storing guano in the field and following two specific protocols for PCR in the laboratory.

METHODS

Area of study.—Bats were captured at Culebrones Cave, which is located ca. 7 km

SW of Arecibo, Puerto Rico, on the grounds of Mata de Plátano, a field station managed by

Interamerican University and the Puerto Rican organization Citizens of the Karst. The cave is a so-called ―hot cave,‖ with relative humidity near 100% and ambient temperatures reaching 40°C (Soto-Centeno and Kurta et al., 2006; CHAPTER 1). This cave houses more than 300,000 bats of six different species, but most individuals are M. blainvillei (61%) or P. quadridens (20%—Rodríguez-Durán, 1996, 2009). Mata de Plátano occupies 0.5 km2 of subtropical, moist, deciduous forest and has a mean ambient temperature of 25°C and average annual rainfall of 154 cm (Soto-Centeno and Kurta et al., 2006).

Trapping bats and collecting guano.—Guano for molecular analysis was collected from M. blainvillei and P. quadridens on 23–25 November 2009, 25–29 December 2009, 3–5

March 2010, and 26–29 June 2010. To capture bats, a harp trap (Kunz and Kurta, 1988) was placed at the opening of Culebrones Cave, and each member of M. blainvillei and P. quadridens that was caught was placed in an individual cloth holding bag for ca. 1 h, or until defecation occurred, and then released unharmed. All guano collected in the field was stored

69 in one of three ways—in an individual plastic bag (Ziploc, S.C. Johnson, Racine, Wisconsin), a vial containing Longmire buffer (Longmire et al., 1988), or a vial containing 95% ethanol.

All samples were then stored in a freezer at -20°C (Clare et al., 2009) within 2 h of collection and were kept frozen until DNA was extracted, regardless of the method of preservation.

Extraction of mtDNA of prey from feces.—Extraction of mtDNA followed the procedures of Clare et al. (2009). Fecal pellets collected for molecular analysis were soaked in 95% molecular-grade ethanol and teased apart under a dissecting microscope. Up to four large fragments of prey, such as legs, wings, antennae, eyes, or pieces of exoskeleton, were removed and placed into an individual well of a 96-well PCR plate, containing 45 μL of lysis buffer and 5 μL of proteinase K (Clare et al., 2009). The PCR plate was then centrifuged, sealed, and incubated in a water bath at 56°C overnight. After incubation, all DNA was isolated using an automated glass-fiber technique (Ivanova et al., 2006). The extracted DNA was suspended in 50 μL of nuclease-free water, and 5 μL of this solution were used in each

PCR, allowing up to ten reactions for every fragment of insect found within the guano.

Original PCR and gel electrophoresis.—The PCR was performed initially using a polymerase master mix (GoTaq Colorless Master Mix, Promega, Madison, Wisconsin), a forward primer (LepF1: 5-ATTCAACCAATCATAAA GATAT-3), and a reverse primer

(LepR1: 5-TAAACTTCTGGATGTCCAAAAA-3), to amplify a 648-bp target region of the

COI gene (Hebert et al., 2004) that have successfully amplified the COI gene in previous studies examining the diet of insectivorous bats (Clare et al., 2009, 2011). If these primers failed to amplify the mtDNA, an alternate reverse primer (AltRev: 5-CTTATATTATTTATTC

GTGGGAAAGC-3) was used instead of LepR1 to generate a 350-bp product (Hebert et al.,

2004). Thermocycling conditions consisted of one cycle of 1 min at 94°C, six cycles of 1

70 min at 94°C, 1.5 min at 45°C, and 1.25 min at 72°C; these steps were followed by 36 cycles of 1 min at 94°C, 1.5 min at 51°C, and 1.25 min at 72°C, with a final step of 5 min at 72°C

(Hebert et al., 2004).

Electrophoresis was then performed on the products of the PCR in a 1% agarose

(GibcoBRL, Carlsbad, California) gel, made with tris/borate/ethylenediaminetetraacetic acid buffer (TBE buffer) at concentration of 0.5x (Sambrook et al., 1989). In addition, 6 μL of nucleic-acid stain (GelRed, Biotium, Hayward, California) were added to every 100 ml of gel solution. The power supply for electrophoresis was set to ca.100 V. Upon completion of electrophoresis, gels were photographed using a gel-documentation system (Bio-Rad,

Hercules, California), and products that clearly showed amplification of 648 bp were purified using a clean-up reagent (ExoSAP-IT PCR, USB Corporation, Cleveland, Ohio) and sent to the University of Michigan DNA Sequencing Core (http://seqcore.brcf.med.umich.edu/) for sequencing of COI. This protocol for the amplification of mtDNA was used on all samples of isolated DNA.

Modified PCR and gel electrophoresis.—Amplification of COI following the original protocol for PCR was successful for only one of ten 96-well plates containing isolated DNA, so a modified protocol was adopted (Zeale et al., 2010). This procedure used a different forward primer (ZBJ-F1: 5-AGATATTGGAACWTTATATTTTATTTTTGG-3) and reverse primer (ZBJ-R2: 5-WACTAATCAATTWCCAAATCCTCC-3), which amplify a 157-bp region of the COI gene. Thermocycling conditions for this particular PCR consisted of one cycle of 3 min at 94°C; followed by 16 cycles of 0.5 min at 94°C, 0.5 min at 61°C (decreased by 0.5°C/cycle), and 0.5 min at 72°C; followed by 24 cycles of 0.5 min at 94°C, 0.5 min at

53°C, and 0.5 min at 72°C; and a final incubation for 10 min at 72°C. Electrophoresis was

71 then performed on the products of the PCR in a 1.5% agarose gel, and all PCR products of

157 bp were cleaned and sent for sequencing using the same procedures as in the original protocol. This second protocol for amplification of mtDNA was adopted late in the study and used on only a subset of the samples of isolated DNA.

Identifying sequences of mtDNA.—All sequences of COI obtained from chitinous fragments in the feces were compared with ca. 1,360,000 reference sequences (accessed

September, 2011) of species of arthropods that were present in the Barcode of Life Data

System (BoLD—Clare et al., 2009, 2011; Hebert et al., 2004; Ratnasingham and Hebert,

2007). For this study, species-level identifications of mtDNA were based on a match of at least 98% to a reference sequence, with no equivalent similarity to any other reference (Clare et al., 2009). When a sequence of COI matched numerous members of a genus but could not be unequivocally identified to the level of species, only a generic identification was made

(Clare et al., 2009). This same process was used to default to other, higher taxonomic levels when necessary.

Statistical analyses.—Differences in the success of amplification of mtDNA among preservation methods (frozen only, in ethanol, or in Longmire buffer) were determined by a z-test for independent proportions, and a Bonferroni correction was implemented to counteract the problem of increased Type I errors caused by multiple comparisons (Rosner,

2000). Because the primers from Clare et al. (2009) were used on every sample of isolated

DNA and the primers from Zeale et al. (2010) were used on a subset of these same samples, differences in the frequency of amplification between these primers were determined by

McNemar’s test for correlated proportions (Rosner, 2000). In addition, differences in success of obtaining matches in BoLD and the frequency of amplification of unwanted DNA (fungal

72 or bacterial DNA), with respect to the different primers, also was examined using

McNemar’s test. Differences in the frequency of amplification and success of recovering matches in BoLD between this study and those of Clare et al. (2009, 2011) and Zeale et al.

(2010) were determined by a z-test for independent proportions. For all tests, alpha was set to 0.05, unless a Bonferroni correction was necessary. All statistical tests were performed using VassarStats (Lowry, 2008).

RESULTS

Extraction of DNA and identification of prey.—Deoxyribonucleic acid was extracted from 620 and 340 fragments of insects that were removed from the feces of 135 M. blainvillei and 86 P. quadridens, respectively. From these samples of extracted DNA, 34 and

54% were amplified successfully for COI from 64 M. blainvillei and 61 P. quadridens, respectively (Table 2.1). Of these amplified samples, 79 and 74% showed strong bands and were sent in for sequencing for COI for M. blainvillei and P. quadridens, respectively (Table

2.1). From these submitted samples, 92 and 90% of the samples from the two species, respectively, were successfully sequenced for COI (Table 2.1), and 83 (54%) and 19 (16%) of these sequences for M. blainvillei and P. quadridens matched an arthropod listed in BoLD

(Tables 2.2–2.3), yielding 50 different taxa. The remaining 152 sequences did not correspond to any arthropod in BoLD, matched a reference in BoLD at less than 98% similarity, or was a sequence of non-arthropodal DNA.

Prey from 4 orders, 13 families, 31 genera, and 18 species of insects were identified within the diet of M. blainvillei (Table 2.2). In the guano of P. quadridens, sequences of mtDNA were identified from five orders, nine families, eight genera, and six species of

73 arthropods (Table 2.3). Three orders (Diptera, Hymenoptera, and Lepidoptera), three families (Crambidae, Geometridae, and Lecithoceridae), two genera (Biston and

Lecithocera), and one species (Biston betularia) were identified in the feces of both M. blainvillei and P. quadridens (Tables 2.2–2.3). Of the 50 different taxa of arthropods that were identified through BoLD, nine indicated a genus or species of insect that is either a vector of human disease or a pest of agricultural crops (Tables 2.2–2.3).

Differences in the preservation of guano and success of primers.—Differences occurred in the frequency of amplification among methods of preservation of guano. DNA extracted from guano that was simply frozen, placed in ethanol, or stored in Longmire buffer was amplified from 57, 65, and 4% of the samples, respectively (Table 2.4). Amplification of mtDNA was significantly higher when guano was placed in plastic bags and frozen than when it was stored in Longmire buffer (Table 2.5). Similarly, DNA extracted from guano that was preserved in 95% ethanol amplified statistically more often than mtDNA extracted from guano placed in Longmire buffer (Table 2.5). Moreover, samples that were stored in

95% ethanol amplified mtDNA more often than guano placed in plastic bags and frozen

(Table 2.5).

The primers from Clare et al. (2009) were used to attempt amplification of all 960 samples of isolated DNA, whereas the primers from Zeale et al. (2010) were applied to 624 of the same samples (Table 2.6). The frequency of amplification differed between primers,

2 with those from Zeale et al. (2010) more often amplifying mtDNA (χ 2= 1.57; P < 0.0001;

Table 2.6) than primers from Clare et al. (2009). The number of sequences of COI that matched an arthropod referenced in BoLD also differed, with primers from Zeale et al.

2 (2010) amplifying more recognized sequences (χ 2 = 1.76; P <0.0001; Table 2.6) than those

74 of Clare et al. (2009). The amount of unwanted DNA from non-arthropods that was

2 amplified also varied (χ 2 = ∞; P <0.0001; Table 2.6), with primers from Clare et al. (2009) amplifying significantly more unwanted DNA than primers used by Zeale et al. (2010).

The frequency of amplification of COI in the current project was 58, 55, and 17% lower than in studies by Clare et al. (2009), Zeale et al. (2010), and Clare et al. (2011), respectively (Table 2.7). Moreover, the investigations by Clare et al. (2009), Zeale et al.

(2010), and Clare et al. (2011) resulted in 61, 55, and 48% more matches in BoLD, respectively, than this study (Table 2.7). Both the frequency of amplification of bands suitable for sequencing, as well as the proportion of sequences of COI that matched a taxon of arthropods referenced in BoLD, were significantly lower in the present study than in the other three reports (Table 2.8).

DISCUSSION

Methods of preservation.—Many factors contribute to the ability to extract and amplify DNA from feces, including temperature and humidity. Lucchini et al. (2002) report that samples of scat collected during winter yield better results than samples collected during summer. Similarly, extracting DNA from fecal samples frozen immediately after defecation is more successful than isolating DNA from feces preserved at temperatures above 0°C

(Prugh et al., 2005). Humidity also affects the rate of extraction and amplification of DNA, with feces collected in dry habitats yielding more amplified DNA than fecal samples from moist environments (Piggott, 2004). Although temperature and humidity play a role in the success of molecular studies, a critical factor affecting the ability to extract and amplify DNA apparently is the manner in which fecal samples are stored. Some biologists preserve

75 samples of scat in ethanol (Oehm et al., 2011; Zeale et al., 2010), but most workers examining diet through molecular techniques simply freeze their samples until analysis

(Clare et al., 2009, 2011; Cnops and Esbroeck, 2010; Wasser et al., 1997).

In this study, success of amplification and the classification of arthropods in BoLD varied significantly among methods of preservation. Mitochondrial DNA was amplified most frequently (65%) when guano was preserved in vials of 95% ethanol and least often

(4%) from guano stored in Longmire buffer. Although Longmire buffer is an effective preservative for blood and tissues (Longmire et al., 1997), data from this study (Tables 2.4–

2.5) indicate that this solution is inappropriate for storing fecal samples from insectivorous bats for later analysis of mtDNA . Panasci et al. (2011) recently found similar results using fecal samples from carnivores. They noted that samples preserved in ethanol or a solution made from dithiothreitol, ethylenediaminetetraacetic acid, and tris-hydrochloride (i.e., DET buffer) were equally efficient at preserving DNA found in feces, and both preservatives were superior to Longmire buffer in frequency of amplification.

Differences in primers.—The primers used to amplify COI varied in success of amplification and species-level classification in BoLD, as well as the amount of unwanted

DNA that they amplified (Table 2.6). Although some DNA survived digestion and was amplified through PCR, this genetic material is undoubtedly degraded to some extent, and according to Deagle et al. (2006), it is difficult to amplify segments of DNA that are 300 bp or larger. This problem of degraded DNA helps explain why amplification of COI was almost three times more successful with the procedures of Zeale et al. (2010), which target a

157-bp segment of mtDNA, than primers from Clare et al. (2009), which target a 648-bp region (Table 2.6). Primers from Zeale et al. (2010) also provided species-level

76 identifications of arthropods in BoLD over four times more frequently (Table 2.6) than the techniques of Clare et al. (2009). This difference in successful identification is most likely due to the primers from Clare et al. (2009) having a higher affinity for the DNA of non- arthropods, such as bacteria or fungi, than primers from Zeale et al. (2010). In my study, primers from Clare et al. (2009) amplified 23 sequences of non-arthropodal DNA, whereas the procedure of Zeale et al. (2010) amplified DNA from only arthropods.

Improvements for future studies analyzing mtDNA.—In the current study, the proportion of sequences amplified for COI that matched an insect referenced in BoLD was significantly lower than in previous reports (Tables 2.7–2.8), and this discrepancy in the success of identifying sequences using BoLD is probably related to the location of the various projects. The present study is the first to apply molecular techniques to neotropical bats and, therefore, neotropical insects, whereas Clare et al. (2009, 2011) and Zeale et al.

(2010) worked in southwestern Ontario and southern England, respectively. It seems likely that fewer samples of mtDNA from Puerto Rican arthropods have been isolated, sequenced, and made available for public access compared with arthropods from Canada and England.

Hence, biologists considering the application of molecular techniques to dietary analysis of neotropical bats should consider sampling local insects that are potential prey and adding the appropriate sequences of COI to BoLD, thus increasing the likelihood of receiving a match when performing DNA-barcoding procedures.

The diet of mormoopids.—Thorough mastication by bats makes identifying fragments of prey in the guano to the level of species by visual means nearly impossible; however, mitochondrial DNA from arthropods was isolated and amplified successfully from both M. blainvillei and P. quadridens, thus providing more evidence that the DNA of arthropods is

77 capable of surviving the digestive system of a bat (Clare et al., 2009; Zeale et al., 2010).

Through molecular techniques, 10 families, 28 genera, and 16 species of Lepidoptera were identified in the diet of M. blainvillei, which was impossible to determine through standard fecal analysis (Table 2.2; CHAPTER 1). The order Ephemeroptera also was found within the diet of M. blainvillei after analyzing mtDNA, but this order went undetected during simple visual analysis of feces (CHAPTER 1). Soft-bodied insects, such as ephemeropterans, typically are digested more completely by bats than other groups of insects (Belwood and

Fenton, 1976; Warner 1985), making ephemeropterans difficult, if not impossible, to find during visual analysis (Clare et al., 2009). Although dipterans and hymenopterans were discovered in the diet of M. blainvillei during standard fecal analysis (CHAPTER 1), identification was impossible beyond the level of order; through PCR, though, insects of these orders were classified to at least the level of genus (Table 2.2). Molecular techniques also showed that P. quadridens consumed representatives of nine families, eight genera, and six species of insects (Table 2.3), all of which were not identified beyond the level of order during standard fecal analysis (CHAPTER 1).

Economically important insects found within the guano.—Insects posing threats to agricultural crops and to health of humans were found within the diet of both M. blainvillei and P. quadridens (Tables 2.2–2.3). The noctuid moths Mocis latipes and M. disseverans, the larvae of which are major threats to sugarcane (Landoldt, 1995; Ogunwolu and Habeck,

1975), corn, and rice (Cave, 1992), were discovered in the guano of M. blainvillei. In addition, Spodoptera frugiperda, another noctuid, whose larvae have been called the most important pest of corn in Puerto Rico (Becker and Miller, 2002; Storer et al., 2010), also was eaten by M. blainvillei.

78

In addition, M. blainvillei consumed Musca domestica, the common housefly. This omnipresent insect is a detriment to farmers of egg-producing poultry, because the adult fly causes egg-spotting, thus reducing the marketable quality of the eggs (Barson et al. 1994).

Moreover, M. domestica often is associated with disease-causing pathogens of humans and animals, such as Salmonella, anthrax, and bacteria that cause bovine mastitis (Barson et al.,

1994). Although the housefly is diurnal (Eesa and Cutkomp, 1995), M. blainvillei apparently encounters the fly when foraging near dusk or dawn.

In addition to insects that were identified to species, a few sequences obtained from feces of M. blainvillei were traced to specific genera containing some members that are economically important. Larvae in the geometrid genus Oxydia, for which there are 59 species (Scolde et al., 1999), feed on leaves of avocado (Peña, 2003) and coffee (Borkhataria et al., 2006; Cotte, 1989). Coffee is one of the most important crops on Puerto Rico, with gross revenues of $36 million in 2001 alone (Inglés et al., 2002), and most coffee is grown in humid and hilly areas of the island (Inglés et al., 2002), not far from Culebrones Cave, where collection of guano took place (Fig. 2.1).

Mormoops blainvillei also consumed noctuid moths from the neotropical genus

Eulepidotis, of which there are 105 species (Pogue and Aiello, 1999), although only three species (E. hebe, E. striaepuncta, and E. superior) are found on Puerto Rico (Poole, 1989).

The larvae of members of this genus are major defoliators of ceiba (Ceiba pentandra), a common tree of Puerto Rico that flowers only once every 5–10 years (Kricher, 1999). The water-resistant wood from this tree is used to make plywood, paper, and canoes, and the pulp of its fruit yields silk cotton, which is a major constituent of mattresses and pillows and used as a substitute for down (Chinea-Rivera, 1990; Kricher, 1999). Finally, DNA associated with

79

Characoma, a pantropical group of noctuids, was isolated from the feces of M. blainvillei.

The larvae of members of this genus feed on leaves and pods of cacao (Akotoye and Kumar,

1976) and foliage of ornamental black olive (Bucida buceras), a tree that is common in

Puerto Rico (Duryea et al., 2007).

For P. quadridens, eight sequences of COI were identified to at least the level of genus, and two of these sequences matched species of insects of agricultural or human-health concern (Table 2.3). Melanotus depressus, an elaterid beetle, the larvae of which are major pests of corn (Brown and Keaster, 1986), was found within the diet of P. quadridens. In addition, the mosquito Aedes vexans is a vector of canine heartworm (Dirofilaria immitis—

Bemrick and Sandholm, 1966; O’Malley, 1990), West Nile virus (Turell et al., 2001), and eastern equine encephalitis virus (O’Malley, 1990), and this culicid fly also was discovered in the guano of P. quadridens.

Use of pesticides and the importance of bats to agriculture.—The United States spends approximately $10 billion annually on about 500 million kg of pesticides (Pimentel and Greiner, 1997), and each dollar spent on pesticides returns $4 in protected crops

(Pimentel, 2005). Despite the application of pesticides, insects still destroy ca. 13% of crops each year (Pimentel, 2005). The use of these chemicals has led to cases of human pesticide poisoning, with ca. 300,000 cases in the United States and 220,000 fatalities worldwide each year (Hart and Pimentel, 2002; Pimentel, 2005). Moreover, ca. 18% of all insecticides contain known carcinogenic chemicals, and use of pesticides correlates with increased rates of cancer in humans (National Research Council, 1987; Pimentel, 2005). Despite these dangers, ca. 73% of fruits and vegetables sold in stores contain residues of pesticides

(Pimentel, 2005).

80

Many pest species of insects are becoming resistant to pesticides, which creates a need for several additional applications of the chemicals to maintain yield of the crops

(Pimentel, 2005), which cost farmers in the United States ca. $1.4 billion each year (Hart and

Pimentel, 2002). This increase in application of pesticides not only intensifies the problems associated with human health but also acts as a selective force on the insects, resulting in further genetic resistance to the chemicals (Pimentel, 2005). However, natural predators, such as insectivorous bats, should have major effects on populations of crop-damaging arthropods (Williams-Guillén et al., 2008), due to their consumption of many herbivorous insects (Kalka et al., 2008).

The Brazilian free-tailed bat (Tadarida brasiliensis), for example, consumes adults of the corn earworm (Helicoverpa zea), the larva of which is a major pest of corn (Cleveland et al., 2006). Cleveland et al. (2006) concluded that this bat reduced damage to crops in Texas, eliminated at least one annual application of pesticide, and even delayed the time when pesticide was first used on the crop, all of which have positive economic, environmental, and human-health benefits. The increased yield of crops, as well as the amount of money saved on pesticides, is evidence that T. brasiliensis specifically and insectivorous bats generally can act as agents of biological control for species of pests (Kalka et al., 2008; Williams-Guillén et al., 2008).

Without use of molecular-based techniques, it would have been impossible to determine that M. blainvillei and P. quadridens consume insects that pose threats to various crops on Puerto Rico or jeopardize the health of humans and their pets. Molecular analysis allows biologists to gather more specific data on the diet of insectivorous bats and helps identify taxa of arthropods within the guano that may otherwise be overlooked through visual

81 fecal analysis. Although my results indicate nine taxa of concern within the diet of M. blainvillei and P. quadridens, a molecular examination of fecal contents from these bats throughout the island likely will result in the discovery of other arthropods of economic importance. Simple identification, however, of a pest in the diet of a few animals is only a first step. Ultimately, it will be necessary to determine how frequently such species are found within the diet of these bats, and, therefore, deduce the extent to which M. blainvillei and P. quadridens may or may not actually control these pests.

82

LITERATURE CITED

Akotoye, N. A. K., and R. Kumar. 1976. Population dynamics of Characoma strictigrapta HMPS. (Lepidoptera: Noctuidae), on cocoa in Ghana. Journal of Applied Ecology, 13:753–773.

Barson, G., N. Renn, and A. F. Bywater. 1994. Laboratory evaluation of six species of entomopathogenic fungi for the control of the house fly (Musca domestica L.), a pest of intensive animal units. Journal of Invertebrate Pathology, 64:107–113.

Becker, V. O., and S. E. Miller. 2002. The large moths of Guana Island, British Virgin Islands: a survey of efficient colonizers (Sphingidae, Notodontidae, Arctiidae, Geometridae, Hyblaeidae, Cossidae). Journal of the Lepidopterists’ Society, 56:9–44.

Belwood, J. J., and M. B. Fenton. 1976. Variation in the diet of Myotis lucifugus (Chiroptera: Vespertilionidae). Canadian Journal of Zoology, 54:1674–1678.

Bemrick, W. J., and H. A. Sandholm. 1966. Aedes vexans and other potential mosquito vectors of Dirofilaria immities in Minnesota. Journal of Parasitology, 52:762–767.

Bensasson, D., D. X. Zhang, D. L. Hartl, and G. M. Hewitt. 2001. Mitochondrial pseudogenes: evolution's misplaced witnesses. Trends in Ecology and Evolution, 16:314–321.

Borkhataria, R.R., J.A. Collazo, and M.J. Groom. 2006. Additive effects of vertebrate predators on insects in a Puerto Rican coffee plantation. Ecological Applications, 16:696–703.

Boyles, J. G., P. M. Cryan, G. F. McCracken, and T. H. Kunz. 2011. Economic importance of bats in agriculture. Science, 332:41–42.

Brown, E. A., and A. J. Keaster. 1986. Activity and dispersal of adult Melanotus depressus and Melanotus verberans (Coleoptera: Elateridae) in a Missouri cornfield. Journal of the Kansas Entomological Society, 59:127–132.

Brown, W. M., M. George, Jr., and A. C. Wilson. 1979. Rapid evolution of animal mitochondrial DNA. Proceedings of the National Academy of Sciences, 76:1967– 1971.

Cave, R. D. 1992. Inventory of parasitic organisms of the striped grass looper, Mocis latipes (Lepidoptera: Noctuidae), in Honduras. Florida Entomologist, 75:592–598.

Chinea-Rivera, J. D. 1990. Ceiba pentandra (L.) Gaertn. Ceiba, Kapok, silk cotton tree. Bombacaceae. Bombax family. USDA Forest Service, Southern Forest Experiment Station, Institute of Tropical Forestry, SO-ITF-SM, 29:1–4.

83

Clare, E. L., B. R. Barber, B. W. Sweeney, P. D. N. Hebert, and M. B. Fenton. 2011. Eating local: influences of habit on the diet of little brown bats (Myotis lucifugus). Molecular Ecology, 20: 1772–1780.

Clare, E. L., E. E. Fraser, H. E. Braid, M. B. Fenton, and P. D. N. Hebert. 2009. Species on the menu of a generalist predator, the eastern red bat (Lasiurus borealis): using a molecular approach to detect arthropod prey. Molecular Ecology, 18:2532–2542.

Cleveland, C. J., M. Betke, P. Frederico, J. D. Frank, T. G. Hallam, J. Horn, J. D. López, Jr., G. F. McCracken, R. A. Medellín, A. Moreno-Valdez, C. G. Sasone, J. K. Westbrook, and T. H. Kunz. 2006. Economic value of the pest control service provided by Brazilian free-tailed bats in south-central Texas. Frontiers in Ecology and the Environment, 5:238–243.

Cnops, L., and M. V. Esbroeck. 2010. Freezing of stool samples improves real-time PCR detection of Entamoena dispar and Entamoeba histolytica. Journal of Microbiological Methods, 80:310–312.

Cotte, O. 1989. Oxydia vesulia a geometrid caterpillar of coffee in Puerto Rico. Journal of Agriculture, University of Puerto Rico, 73: 159–160.

Dávalos, L. M. 2006. The geography of diversification in the mormoopids (Chiroptera: Mormoopidae). Biological Journal of the Linnean Society, 88:101–118.

Deagle, B. E., J. P Eveson, and S. N. Jarman. 2006. Quantification of damage in DNA recovered from highly degraded samples: a case study on DNA in faeces. Frontiers in Zoology, 3:11.

Duryea, M. L., E. Kampf, R. C. Littell, and C. D. Rodríguez-Pedraza. 2007. Hurricanes and the urban forest: II. Effects on tropical and subtropical tree species. Arboriculture and Urban Forestry, 33:98–112.

Eesa, N. M., and L. K. Cutkomp. 1995. Pesticide chronotoxicity to insects and mites: an overview. Journal of Islamic Academy of Science, 8: 21-28.

Farrell, L. E., J. Roman, and M. E. Sunquist. 2000. Dietary separation of sympatric carnivores identified by molecular analysis of scats. Molecular Ecology, 9:1583– 1590.

Folmer, O., M. Black, W. Hoeh, R. Lutz, and R. Vrijenhoek. 1994. DNA primers for amplification of mitochondrial cytochrome coxidase subunit I from diverse metazoan invertebrates. Molecular Marine Biology and Biotechnology, 3:294–299.

Gannon, M. R., A. Kurta, A. Rodríguez-Durán and M. R. Willig. 2005. Bats of Puerto Rico: an island focus and a Caribbean perspective. Texas Tech University Press, Lubbock, Texas.

84

Groom, M. J. 2006. Threats to biodiversity. Pp. 61–109, in Principles of conservation biology (M.J. Groom, G.K. Meffe, and C.R. Carroll, eds.). Third ed. Sinauer Associates, Sunderland, Massachusetts.

Hart, K. A., and D. Pimentel. 2002. Environmental and economic costs of pesticide use. Pp. 237–249, in Encyclopedia of pest management, Vol. 1 (D. Pimentel, ed.). Marcel Dekker, New York, New York.

Hebert, P. D. N., A. Cywinska, S. L. Ball, and J. R. DeWaard. 2003. Biological identifications through DNA barcodes. Proceedings of the Royal Society B: Biological Sciences, 270:313–321.

Hebert, P. D. N., E. H. Penton, J. M. Burns, D. H. Janzen, and W. Hallwachs. 2004. Ten species in one: DNA barcoding reveals cryptic species in the neotropical skipper butterfly Astraptes fulgerator. Proceedings of the National Academy of Sciences, 101:14812–12817.

Inglés, M. M., W. Almodóvar, H. O’Farrill, and A. N. Alvarado-Ortiz. 2002. Crop profile for coffee in Puerto Rico. . Accessed 21 August 2011.

Ivanova, N. V., J. R., deWaard, and P. D. N. Hebert. 2006. An inexpensive automation- friendly protocol for recovering high-quality DNA. Molecular Ecology Notes, 6:998– 1002.

Kalka, M. B., A. R. Smith, and E. K. V. Kalko. 2008. Bats limit arthropods and herbivory in a tropical forest. Science, 320:71.

King, R. R., D. S. Read, M. Traugott, W. O. C. Symondson. 2008. Molecular analysis of predation: a review of best practice for DNA-based approaches. Molecular Ecology, 17:947–963.

Kricher, J. C. 1999. A neotropical companion: an introduction to the animals, plants, and ecosystems of the new world tropics. Princeton University Press, Princeton, New Jersey.

Kunz, T. H., and A. Kurta. 1988. Capture methods and holding devices. Pp. 1–29, in Ecological and behavioral methods for the study of bats (T.H. Kunz, ed.). Smithsonian Institution Press, Washington, D.C.

Kunz, T. H., E. Braun de Torrez, D. Bauer, T. Lobova, and T. H. Fleming. 2011. Ecosystem services provided by bats. Annals of the New York Academy of Science, 1223:1–36.

Landolt, P. J. 1995. Attraction of Mocis latipes (Lepidoptera: Noctuidae) to sweet baits in traps. Florida Entomologist, 78:523–530.

85

Longmire, J. L., M. Maltbie, and R. J. Baker. 1997. Use of ―lysis buffer‖ in DNA isolation and its implication for museum collections. Texas Tech University Museum, Occasional Papers, 163:1–4.

Longmire, J. L., A. K. Lewis, N. C. Brown, J. M. Buckingham, L. M. Clark, M. D. Jones, L. J. Meincke, J. Meyne, R. L. Ratliff, F. A. Ray, R. P. Wagner, and R. K. Moyzis. 1988. Isolation and molecular characterization of a highly polymorphic centromeric tandem repeat in the family Falconidae. Genomics, 2: 14–24.

Lowry, R. 2008. VassarStats: website for statistical computation. Vassar College, Poughkeepsie, New York. < http://faculty.vassar.edu/lowry/VassarStats.html>. Accessed August 2011.

Lucchini, V., E. Fabbri, F. Marucco, S. Ricci, L. Boitani, and E. Randi. 2002. Noninvasive molecular tracking of colonizing wolf (Canis lupus) packs in the western Italian Alps. Molecular Ecology, 11:857–868.

MacArthur, R. H., and E. O. Wilson. 1967. The theory of island biogeography. Princeton University Press, Princeton, New Jersey.

National Research Council. 1987. Regulating pesticides in food. The Delaney Paradox. National Academy Press, Washington, D.C.

Oehm, J., A. Juen, K. Nagiller, S. Neuhauser, and M. Traugott. 2011. Molecular scatology: how to improve prey DNA detection success in avian faeces? Molecular Ecology Resources, 11:620–628.

Ogunwolu, E. O., and D. H. Habeck. 1975. Comparative life-histories of three Mocis spp. in Florida (Lepidoptera: Noctuidae). Florida Entomologist, 58:97–103.

O'Malley C. M. 1990. Aedes vexans (Meigen): an old foe. Proceedings of the New Jersey Mosquito Control Association, 77:90–95.

Panasci, M., W. B. Ballard, S. Breck, D. Rodriguez, L. D. Densmore, III, D. B. Webster, and R. J. Baker. 2011. Evaluation of fecal DNA preservation techniques and effects of sample age and diet on genotyping success. Journal of Wildlife Management, 75: 1616–1624.

Peña, J. E. 2003. Pests of avocado in Florida. Proceedings V World Avocado Congress: 487– 494.

Piggott, M. P. 2004. Effect of sample age and season of collection on the reliability of microsatellite genotyping of faecal DNA. Wildlife Research, 31:485–493.

86

Pimentel, D. 2005. Environmental and economic costs of the application of pesticides primarily in the United States. Environment, Development, and Sustainability, 7:229– 252.

Pimentel, D., and A. Greiner. 1997. Environmental and socio-economic costs of pesticide use. Pp. 51–78, in Techniques for reducing pesticide use: environmental and economic benefits (D. Pimentel, ed.). John Wiley and Sons, Chichester, United Kingdom.

Poole, R. W. 1989. Lepidopterorum catalogus. Noctuidae. Fasc. 118. E.J Brill Publications, Leyde, New York.

Pogue, M. G., and A. Aiello. 1999. Description of the immature stages of three species of Eulepidotis guenee (Lepidoptera: Noctuidae) with notes on their natural history. Proceedings of the Entomological Society of Washington, 101: 300–311.

Prugh, L. R., C. E. Ritland, S. M. Arthur, and C. J. Krebs. 2005. Monitoring coyote population dynamics by genotyping faeces. Molecular Ecology, 14:1585–1596.

Ratnasingham, S., and P. D. N. Hebert. 2007. BoLD: the Barcode of Life Data System (www.barcodinglife.org). Molecular Ecology Notes, 7:355–364.

Rodríguez-Durán, A. 1996. Foraging ecology of the Puerto Rican boa (Epicrates inornatus): bat predation, carrion feeding, and piracy. Journal of Herpetology, 30:533–536.

Rodríguez-Durán, A. 2009. Bat assemblages in the West Indies: the role of caves. Pp. 265– 279, in Island bats: evolution, ecology, and conservation. (T.H. Fleming and P.A. Racey, eds.). University of Chicago Press, Chicago, Illinois.

Rosner, B. 2000. Fundamentals of biostatistics. Fifth ed. Duxbury, Pacific Grove, California.

Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecular cloning: a laboratory manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, New York.

Schnitzler, H., and E. K. V. Kalko. 2001. Echolocation by insect-eating bats. BioScience, 51:557–569.

Scolde, M. J., L. M. Pitkin, M. S. Parsons, and M. R. Honey. 1999. Geometrid moths of the world: a catalogue. CSIRO Publishing, Collingwood, Victoria, Australia.

Silva-Taboada, G. 1979. Los murciélagos de Cuba. Academia de Ciencies de Cuba, Havana, Cuba.

Soto-Centeno, J. A., and A. Kurta. 2006. Diet of two nectarivorous bats, Erophylla sezekorni and Monophyllus redmani (Phyllostomidae), on Puerto Rico. Journal of Mammalogy, 87:19–26.

87

Stone, A. C., J. E. S. Starrs, L. L. M., and M. Stoneking. 2001. Mitochondrial DNA analysis of the presumptive remains of Jesse James. Journal of Forensic Science, 46:173–176.

Storer, N. P., J. M Babcock, M. Sclenz, T. Meade, G. D. Thompson, J. W. Bing, and R. M. Huckaba. 2010. Discovery and characterization of field resistance to Bt maize: Spodoptera frugiperda (Lepidoptera: Noctuidae) in Puerto Rico. Journal of Economic Entomology, 103:1031–1038.

Symondson, W. O. C. 2002. Molecular identification of prey in predator diets. Molecular Ecology, 11:627–641.

Turell, M. J., M. L. O’Guinn, D. J. Dohm, and J. W. Jones. 2001. Vector competence of North American mosquitoes (Diptera: Culicidae) for West Nile virus. Journal of Medical Entomology, 38:130–134.

Warner, R. M. 1985. Interspecific and temporal dietary variation in an Arizona bat community. Journal of Mammalogy, 66:45–51.

Wasser, S. K., C. S. Houston, G. M. Koehler, G. G. Cadd, and S. R. Fain. 1997. Techniques for application of faecal DNA methods to field studies of ursids. Molecular Ecology, 6:1091–1097.

Williams-Guillén K., I. Perfecto, and J. Vandermeer. 2008. Bats limit insects in a neotropical agroforestry system. Science, 320:70.

Wilson, D. E., and D. M. Reeder. 2005. Mammal species of the world: a taxonomic and geographic reference. Third ed. John Hopkins University Press, Baltimore, Maryland.

Zaidi, R. H., Z. Jaal, N. J. Hemingway, and W. O. C. Symondson. 1999. Can multiple-copy sequences of prey DNA be detected amongst the gut contents of invertebrate predators? Molecular Ecology, 8:2081–2087.

Zeale, M. R. K., R. K. Butlin, G. L. A. Barker, D. C. Lees, and G. Jones. 2010. Taxon- specific PCR of DNA barcoding arthropod prey in bat faeces. Molecular Ecology Resources, 11:236–244.

88

TABLE 2.1.—Number of samples of DNA that were extracted, successfully amplified, and sequenced for both M. blainvillei and P. quadridens.

Number of Number of Number of samples Number of samples Number of matches to Species samples of DNA amplified successfully sent in for sequencing arthropods in BoLD isolated samples sequenced M. blainvillei 620 210 165 153 83 P. quadridens 340 185 136 122 19

89

89

TABLE 2.2.—Insects identified in the diet of M. blainvillei.

Similarity Number of matches Number of bats Order Family Genus Specific epithet (%) in BoLD that ate insect Diptera Muscidae Musca domestica a 98.7 1 1 Ephemeroptera Isonychiidae Isonychia arida 98.0 1 1 Hymenoptera Ichneumonidae Probles 98.0 1 1 Lepidoptera Crambidae Notarcha 98.9 1 1 Sparagmia gigantalis 99.3 1 1 Geometridae Biston betularia 100.0 1 1 Enypia venata 100.0 1 1 Oxydia a 98.1 1 1 Patalene hamulata 98.7 1 1 Ptychamalia 98.0 1 1 Sphacelodes vulneraria 100.0 4 3

90 Tephronia 100.0 2 2 Lecithoceridae Lecithocera 98.9 12 7 Lycaenidae Callophrys augustinus 99.0 1 1 Noctuidae Anomis erosa 98.6 2 2 Characoma a 100.0 1 1 Eulepidotis a 98.0 1 1 Mocis disseverans a 99.2 1 1 latipes a 100.0 5 3 Ophisma tropicalis 98.5 1 1 Parachabora 100.0 1 1 Physula albipunctilla 99.3 13 10 Continued

90

TABLE 2.2.—Continued.

Similarity Number of matches Number of bats that Order Family Genus Specific epithet (%) in BoLD ate insect Lepidoptera Noctuidae Rejectaria 98.5 1 1 Schinia 100.0 1 1 Spodoptera frugiperda a 99.7 2 2 Nolidae 100.0 2 2 Notodontidae 100.0 1 1 Nymphalidae Heliconius hecale 99.0 2 2 Oeneis 98.1 1 1 Saturniidae Lemaireia 98.0 1 1 Torticidae Ancylis spiraeifoliana 99.3 1 1 Clepsis 98.1 3 3

91

Homona trachyptera 99.0 3 1 Olethreutes inorantana 99.1 1 1 a Insect of agricultural or human-health concern.

91

TABLE 2.3.— Arthropods identified in the diet of P. quadridens.

Number of Number of bats that Order Family Genus Specific epithet Similarity (%) matches in BoLD ate insect Aranae 100.0 1 1 Coleoptera Carabidae Harpalis pennsylvanicus 99.2 1 1 Dytiscidae Bidessus grossepunctatus 98.8 1 1 Elateridae Melanotus depressus a 100.0 2 2 Diptera Culicidae Aedes vexans a 100.0 1 1 Hymenoptera Formicidae 100.0 1 1 Lepidoptera Bombycidae Phiditia 100.0 1 1 Crambidae Oenobotys vinotinctalis 98.5 1 1 Geometridae Biston betularia 98.7 2 1

92 Lecithoceridae Lecithocera 98.9 11 5 a Insect of agricultural or human-health concern.

92

TABLE 2.4.—Number of samples successfully amplified over the total number of samples of DNA isolated from fragments of arthropods, followed by the percent in parentheses, according to type of primer and method of preservation. Amplification with primers from Clare et al. (2009) was attempted with every sample (960) of isolated DNA, but primers from Zeale et al. (2010) were used to amplify only a subset (624 samples) of the isolated DNA.

Primer Frozen Ethanol Longmire buffer Total for primer Clare et al. (2009) 122/364 (34) 26/276 (9) 0/320 (0) 148/960 (15) Zeale et al. (2010) 79/112 (71) 153/268 (57) 15/244 (6) 247/624 (40) Total for method of preservation 208/364 (57) 180/276 (65) 15/320 (4)

93

93

TABLE 2.5.—Comparisons of the proportion of amplified samples, with respect to primers and methods of preservation.

Primer Comparison z Pa Clare et al. (2009) Frozen and ethanol 7.16 <0.0001 Frozen and Longmire buffer 10.72 <0.0001 Ethanol and Longmire buffer 4.31 <0.0001

Zeale et al. (2010) Frozen and ethanol 2.45 0.008 Frozen and Longmire buffer 12.79 <0.0001 Ethanol and Longmire buffer 12.26 <0.0001

a Alpha was set to 0.017, using a Bonferroni correction.

94

94

TABLE 2.6.—Number of samples of DNA that were extracted and successfully amplified, using primers from Clare et al. (2009) and Zeale et al. (2010).

Number of Number of Number of Number of Number of Number of samples matches to samples of Primers samples of amplified samples sent in successfully arthropods in nonarthropodal DNA isolated samples for sequencing sequenced BoLD DNA Clare et al. (2009) 960 148 109 101 6 23 Zeale et al. (2011) 624 247 192 174 44 0

95

95

TABLE 2.7.—Differences in the number of samples submitted for sequencing for COI, sequence matches in BoLD, and rate of unwanted DNA between current and previous studies.

Number of Number of Number of samples Number of samples Percent of matches to Yield Study samples of sent in for of nonarthropodal amplified arthropods in (%) DNA isolated sequencing DNA samples BoLD Clare et al. (2009) 896 797 621 77.9 176 22.1 Clare et al. (2011) 2,016 971 631 64.9 340 35.0 Zeale et al. (2010) 240 207 150 72.0 0 0.0

Current study 960 301 102 33.8 23 7.6

96

96

TABLE 2.8.—Proportion of isolated samples of DNA that amplified, as well as the proportion of amplified samples that yielded matches in BoLD for this study, compared with Clare et al. (2009, 2011) and Zeale et al. (2010). Process Study z P

Isolation of DNA to amplification Clare et al. (2009) -25.23 <0.0001

Clare et al. (2011) -8.67 <0.0001

Zeale et al. (2010) -15.39 <0.0001

Amplification to matches in BoLD Clare et al. (2009) -13.72 <0.0001

Clare et al. (2011) -9.54 <0.0001

Zeale et al. (2010) -8.55 <0.0001

97

FIG. 2.1.—Municipalities on Puerto Rico where coffee is farmed. The star indicates the location of Culebrones Cave, where guano was collected for this study.

98