ISSN-0034-7744

Volume 62 (3) • September, 2014 • Volumen 62 (3) • Setiembre, 2014 San José, Costa Rica • www.biologiatropical.ucr.ac.cr

Associated Colleges of the Midwest (ACM) 50 years / 50 años

In 2014 the Associated Colleges of the Midwest, Costa Rica Program, celebrates 50 years. Our journal dedicates this issue to the Program and includes a dossier of representative short articles produced by ACM students based on their field research.

En el 2014 el Programa de Costa Rica de la Associated Colleges of the Midwest, celebra sus 50 años. Nuestra revista le dedica este número al Programa e incluye un dossier de artículos cortos representativos que produjeron estudiantes de la ACM basados en sus investigaciones de campo. Telefax (506) 2511-5550 • (506) 2511-8982 [email protected] • www.ots.ac.cr/tropiweb • www.biologia.ucr.ac.cr/rbt/ Revista de Biología Tropical • Universidad de Costa Rica • 11501-2060 San José, Costa Rica

Graphic Editor / Editor Gráfico: Sergio Aguilar Mora

Development Editor / Editor de Desarrollo: Alonso Prendas Vega

Cover photograph / Fotografía de portada: Some of the many students who did fieldwork in Costa Rica. Background photo: (Upper) ChristopherVaughan; (Lower) Field Research in the Environment, Social Sciences, & Humanities (Staff and Students of Spring 2013). 1. Michael McCoy. 2. Amanda Cormier. 3. Lauren Carlson. 4. Nicole Nemeth. 5. Jonathan Henn. 6. Kathleen Shea. 7. Shasta Webb. 8. Peggy Barlett. 9. Victoria Egedus. 10. Judith Magnan. 11. Mark Myers. 12. Emma Cornwell. 13. John Seng. 14. Ken Salazar. 15. Jeffrey Nadel. Photographs courtesy of ACM.

Algunos de los muchos estudiantes que han realizado trabajo de campo en Costa Rica. Foto de fondo: (Superior) Christo- pher Vaughan; (Inferior) Investigaciones de Campo en el Ambiente, las Ciencias Sociales & las Humanidades (personal y estudiantes de la primavera de 2013). 1. Michael McCoy. 2. Amanda Cormier. 3. Lauren Carlson. 4. Nicole Nemeth. 5. Jonathan Henn. 6. Kathleen Shea. 7. Shasta Webb. 8. Peggy Barlett. 9. Victoria Egedus. 10. Judith Magnan. 11. Mark My- ers. 12. Emma Cornwell. 13. John Seng. 14. Ken Salazar. 15. Jeffrey Nadel. Fotografías cortesía de ACM.

574.05 R Revista de Biología Tropical / Universidad de Costa Rica. ––Vol. 1 (1953)– . –– San José, C. R. : Editorial Universidad de Costa Rica, 1953– v.

ISSN–0034–7744

1. Biología – Publicaciones periódicas, 2. Publicaciones periódicas costarricenses.

BUCR

Attribution (BY - NC - ND) • (BY) You must attribute the work in the manner specified by the author or licensor (but not in any way that suggests that they endorse you or your use of the work). • (NC) You may not use this work for commercial purposes. • (ND) You may not alter, transform, or build upon this work. Open Reconocimiento (BY - NC - ND) • (BY) Debe atribuir el trabajo de la manera especificada por el autor o el licenciador access (pero no de manera que sugiera que tiene la aprobación para el uso del trabajo). • (NC) No puede utilizar esta obra para fines comerciales. • (ND) Usted no puede alterar, transformar, o ampliar este trabajo.

Edición aprobada por la Comisión Editorial de la Universidad de Costa Rica

© Editorial Universidad de Costa Rica, Ciudad Universitaria Rodrigo Facio, Costa Rica. Apdo. 11501-2060 • Tel.: 2511-5310 • Fax: 2511-5257 E-mail: [email protected] • Página web: www.editorial.ucr.ac.cr

Prohibida la reproducción total o parcial salvo normas Open Access. Todos los derechos reservados. Hecho el depósito de ley. Over the past fifty years, the ACM Costa Rica program has influenced the lives of many students. Due to space limitations, we cite only one example:

“In 1975, I participated in the Associated Colleges of the Midwest Pro- gram in Costa Rica. The experience shaped much of my future including a deep interest in Latin America. Many times in the United States Senate and at the White House, I have spoken with Presidents and Senators about the imperative of strong mutually supportive relationships with Central and South America. I am rooted in this view in part because of the ACM program and my study of President Kennedy’s Alliance for Progress. I believe the future of the United States, and Latin America, are inextricably tied together because of our shared history and shared demographics.”

Ken Salazar, JD

Partner in the law firm WilmerHale. Secretary of the Interior of the United States, 2009- 2013. U.S. Senator for Colorado, 2005-2009. Attorney General of Colorado, 1999-2005. Executive Director of the Colorado Department of Natural Resources, 1990-1994. Chief legal counsel to the Governor of Colorado, 1986-1990. preface

Fifty years of undergraduate scientific field research in the Associated Colleges of the Midwest (ACM) Costa Rica program

Christopher Vaughan1,2, Judith Magnan1 & Michael B. McCoy1 1. Associated Colleges of the Midwest, San Pedro de Montes de Oca, Costa Rica; [email protected] 2. Department of Forest and Wildlife Ecology, University of Wisconsin, Madison, Wisconsin; [email protected]

The importance of preparing students for Olaf Colleges and Lawrence University, was global citizenship and the development of joined by Macalester and Colorado Colleges intercultural competency in an increasingly in 1967, Lake Forest College in 1974, the Rica program Costa Special section (ACM) interdependent world has long been recog- College of the University of Chicago between nized by many higher education institutions 1988 and 2008, and Luther College in 2009 in the United States. Experiential learning (ACM 2013a). ACM currently has 13 member through study abroad is one means of achiev- colleges and one university from five states ing this goal (Brewer & Cunningham, 2010). and administers 16 study abroad programs For decades, Costa Rica has offered a stable, in 11 countries. welcoming and culturally rich environment for Initial consortial goals for joint action study abroad programs. focused on the educational effectiveness, effi- For undergraduate students interested ciency of administrative and cultural opera- in pursuing further studies and academic or tions, and development of additional sources of research careers, the opportunity to develop revenue of the member colleges and university. field research skills is a valuable tool. To carry In 2013, ACM´s mission is to strengthen its out professionally guided research with the members as leaders in liberal arts education cultural and linguistic preparation afforded by through joint efforts to improve the profes- study abroad not only provides students with sional effectiveness of faculty and administra- academic and cultural development, but it may tive leaders, to provide exemplary liberal arts enable them to make a significant contribution learning through off-campus studies, and to to the host country. promote excellence in teaching and learning through collaboration (ACM 2013b). HISTORY OF THE ACM COSTA RICA PROGRAM ACM Costa Rica Program

ACM Consortium In 1963, in response to a growing inter- est in curricular development through foreign In 1958, 10 liberal arts colleges of the study and research, and after considering other Midwestern United States formed the Associat- potential Central American sites, ACM created ed Colleges of the Midwest (ACM) to promote the Central American Field Studies Program institutional collaboration and the achievement in San José, Costa Rica. The program was of common objectives. The founding group, designed to provide participants with experi- comprised of Beloit, Carleton, Coe, Cornell, ence in scientific field studies in preparation Grinnell, Knox, Monmouth, Ripon, and St. for careers related to the tropical environment

Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 843-858, September 2014 843 and to allow ACM faculty to develop research When external program funding ended interests in Latin American studies. The result- in 1967, the relationship with the TSC ended, ing field research projects would contribute to although Hunter continued to serve as director. the growing knowledge of the tropical environ- Language was incorporated into the curriculum ment and to teaching materials in anthropol- and students were housed with Costa Rican ogy, biology, economics, geography, geology, families, facilitating greater cultural integration. and sociology. The program did not expand to other ACM designated the non-profit research Central American countries as was initially and teaching organization, the Tropical Sci- planned, and in 1970, was renamed the Costa ence Center (TSC), founded in 1962 by tropi- Rican Development Studies Program. In 1974, cal ecologist Leslie R. Holdridge, geographer under the direction of anthropologist Ridgeway Satterthwaite, the focus of the fall semester Joseph A. Tosi, and tropical biologist J. Robert changed to “Studies in Latin American Culture Hunter to administer the program. Dr. Hunter and Society,” leaving field research exclusively was named program director. The initial plans for the spring. In 1977, the field study semester for the program were broadened to include became the Tropical Field Research Program implementing independent study and research (ACM 2013c; Dennis, 1989). throughout Central America, and develop- In the late 1970´s, ACM developed a ing a relationship with the Consejo Supe- relationship with the University of Costa Rica rior Universitario Centroamericano (CSUCA). (UCR) that led to the signing of a legal agree- The first group of students began the pro- ment between the two institutions. There are gram in July of 1964 under Hunter´s direction three essential components of this relation- (P. Dennis, 1989). ship: a) the ACM scholarship program through which approximately 200 UCR students have Costa Rica Program development received full, one-year scholarships to attend Over the past fifty years, Costa Rica has ACM and other participating colleges and universities; b) a one-year visiting professor provided the ideal environment for student aca- position for a recent ACM graduate to teach demic and cultural growth. During this time, the English at UCR; and c) visiting student status program has maintained a strong field research for ACM students at UCR. component in the spring semester while the fall In 2009, the Latin American cultural focus semester has evolved in response to changing of the fall program was redefined with a contemporary issues in Latin America in gen- more local emphasis as Costa Rica: Language, eral, and Costa Rica in particular. Since 1964, Society, & the Environment. That year, the approximately 2 050 students have participated breadth of research carried out in the spring in the program, divided almost equally between was reflected in the name of the program that the fall and spring programs. remains today, Costa Rica: Field Research in Due to widespread interest in tropical the Environment, Social Sciences, & Humani- science in the United States at the time, the ties. In 2013, the fall semester was reoriented program received funding from the National to encompass new directions with the title Science Foundation and the Ford Founda- Costa Rica: Community Engagement in Public tion until 1967. Program director and agron- Health, Education, & the Environment. omy professor, Hunter, was accompanied by TSC professors Holdridge in ecology, Tosi CURRENT PROGRAM CURRICULA in geography and Jorge Lines in archaeol- ogy. At that time, students lived in a boarding The fall and spring ACM Costa Rica house in San José and language was not part programs share the common goal of exposing of the curriculum. students to Costa Rica’s natural and cultural

844 Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 843-858, September 2014 resources through an intensive language pro- concurrently deepen knowledge in particular gram complemented by educational field trips fields such as public health, education, and the and guest speakers by means of which they are environment. Although most of the semester exposed to projects, local experts and activ- is spent in the Central Valley in coursework ists in the environmental and social sciences, (www.acm.edu/costarica), an important part public health, and humanities. Students fur- of the program is a month-long rural intern- ther develop their language skills, intercultural ship or practicum, for which students live with competency and appreciation of the host coun- local “campesino” families, participate in local try by living with both an urban and a rural community life, and complete a project related Costa Rican host family. In both semesters, to public health, education and/or the environ- students often express their desire to make a ment. During the 12 week urban stay, divided contribution to the country through their intern- into an initial 5 week block and a 7 week block ship, practicum or field research. at the end, groups take field trips with course Since 1967, ACM´s commitment to greater professors and resource people around Costa student linguistic and cultural development Rica. This helps students appreciate the coun- through host family stays has led to the devel- try’s astounding biodiversity, and very interest- opment of a database of over 100 families in the ing cultural, educational, health and historical Rica program Costa Special section (ACM) San José area and approximately 300 families elements. Spanish language improvement is in rural areas throughout the country. Families continuous in small customized Spanish classes are selected based on their interest in integrat- and through living with Costa Rican families. ing students in their family life and culture; in Students are pressured and rewarded to speak exchange, they receive a monthly stipend to Spanish in courses, with family members, with cover the student´s expenses. Both students and peer groups, on field trips, and in other pro- families receive orientation and follow-up staff gram activities, leaving the country much more support to facilitate student adaptation to the fluent than they arrive. family, community and culture. For students, this experience has long been recognized as Costa Rica: Field Research in the a key component of the program and families Environment, Social Sciences, have expressed their desire to continue to & Humanities receive ACM students because of their stu- Similar to the fall program, the spring dents’ academic orientation and interest in the research program was created in 1964 to inves- culture. This component of the program has led tigate Costa Rica’s natural environment and its to the development of a culturally respectful people and especially how the country main- and mutually beneficial relationship essential tains a top ranking in Latin America in biodi- to the students´ personal and academic growth. versity conservation and human development. ACM student research is focused on applied Costa Rica: Community Engagement studies dealing with interrelated aspects of bio- in Public Health, Education, diversity, agriculture, public education, social- & the Environment ized medicine, and culture. An estimated 1 015 students have par- ticipated in the ACM Costa Rica fall semester ACM COSTA RICA RESEARCH program over the 50-year period. The present fall ACM Costa Rica program prepares stu- During the past 50 years, there have been dents for exploring, studying and working in approximately 1 088 ACM research projects Latin America (and developing countries in spanning the natural sciences (532) and social general) through coursework and other expe- sciences and humanities (556) (Table 1). The riences meant to develop language skills and research process at ACM Costa Rica has been

Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 843-858, September 2014 845 likened to conducting significant research and several drafts (Table 2). Finally, the advisor writing a “mini-thesis” because the student submits a grade and written evaluation. goes through all the steps that a graduate stu- dent undertakes for a master’s thesis. In almost Research Advisors all cases, the students have never conducted ACM research advisors (acm.edu/cradvi- research and Costa Rica is their first experi- sor) have been selected based on their profes- ence. Working closely with a local advisor, the sional fields, research experience and interest research coordinator and director, the student in working with ACM. These local advisors designs, carries out, analyzes and writes up a are an integral part of the research process, unique field research project. There are current- ensuring feasible research projects. Advi- ly semester and trimester research programs sors and research coordinators work with stu- during which students experience linguistic and dents in small groups and individually to cultural immersion living with urban and rural provide overall guidance in the development, host families. The semester research process implementation, data analysis, and preparation consists of three periods: a) one month (usually of final papers and presentations of student February) writing a proposal, while students research projects. Work of research coordina- study and improve Spanish in and outside of tors enhances the individual guidance pro- classes, b) two months of field research (March vided by the local research advisors. Several and April) when students collect data in a rural months before travelling to Costa Rica, stu- area following the guidelines established in the dents accepted into the program are assigned research proposal, and c) one month (May) of an advisor based on their interests and advisors’ data analysis when two paper drafts are written expertise. Based on initial contact between the before the final paper is turned in and pre- advisor and the student, students begin review- sented orally in the ACM building at a group ing and saving published articles pertinent to “semiprofessional” scientific symposium. the proposed research project. This database Since 2010, several changes to strengthen the can be improved in Costa Rica after meeting research program have been instituted. These with advisors at the ACM center or at the Uni- include: a) a field research course during the versity of Costa Rica (UCR) where they are first month, b) several drafts of the proposal visiting students. and final paper, c) use of Dropbox for student folders and the exchange of project-related Program Structure files, and d) required online training and in class review of ethical principles. The first month (February) involves learn- ing how to conduct research, writing a research Grading proposal outline, two preliminary drafts, the final research proposal and student and advi- For grading, this semester is divided into sor visiting the field site (Table 2). Cultural three courses: a) a Spanish course, b) a field and language immersion are also emphasized. research course and c) a field seminar and A workshop on research, proposal design and paper. For the field research course, the aca- methodology is given by the natural science demic work is evaluated based on the student’s: coordinator and director which focuses on a) use of field research methods in keeping such topics as research design, preparation of with the research proposal, b) understanding a proposal, and data collection in the field. and analysis of data, and c) dedication in the The proposal should follow guidelines given field. For the field seminar and paper, the stu- to students by the director and found on the dent is evaluated based on three oral presenta- following website (http://writing.colostate.edu/ tions and the research proposal and final paper. guides/processes/science/). A Dropbox folder The research proposal and final papers require for each student on his/her research project to

846 Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 843-858, September 2014 facilitate communication is used extensively introduction, methodology, results, discussion, by the student, advisor, research coordinator conclusions, and literature cited. Figures, tables and director with numerous folders dealing and graphs are presented at the end. These parts with drafts, final copies and oral presentations. were also followed during the proposal writing Students design a data sheet and/or interview process to facilitate final paper writing. The to organize data collection. Information on ACM director provides guidelines for writing ethical guidelines for human subject research the final paper and there are several classes is discussed for students dealing with humans given by the research coordinator and director in their research and they are required to take about writing the final paper. Also ACM Costa an online course given by the National Insti- Rica uses the following website http://writing. tutes of Health (NIH 2013). Students share colostate.edu/guides/processes/science to facili- their research project with their colleagues as a tate paper writing. An outline and minimum of PowerPoint presentation in a short joint meet- two drafts are written before the final paper is ing. Throughout the semester and during the turned in (Table 2). All documents are deposited midterm and final presentations, students share in Dropbox. All research papers are normally their research projects with colleagues, advi- written in English unless the director, Spanish sors and staff; so there is a joint progressive language coordinator and advisor agree to the Rica program Costa Special section (ACM) learning experience. student’s petition to write in Spanish. The final During the second and third month (March paper must be typed (including graphs, charts, and April) a minimum of 40 days are spent diagrams) and double-spaced, with final copies in data collection at a rural site, usually dur- saved in the student’s Dropbox folder, sent to ing weekdays while students live with a rural the director and advisor in Microsoft Word for- family. During the research period, students mat. Photographs, charts and appendices must maintain a data sheet and/or carry out inter- be incorporated in the main Word file. Students views, keep a daily activity record with a field may use the ACM Costa Rica research for a or spiral notebook, take pictures and carry out senior project, honor’s thesis, publication, Costa preliminary data analysis when in contact with Rican, U.S. or international symposium or as advisors or research coordinator. Advisors, the basis of graduate research. Eighteen (18) of director and research coordinator visit the site the ACM Costa Rica student research projects and consult by e-mail (all students have lap- have led to publications in the International tops and most have e-mail in their rural site) to Journal of Tropical Biology and Conservation ensure the project is proceeding as designed. (Appendix 1) and nine more are included in this Changes in experimental design may be made number of the journal. We are very proud to if necessary. Students return to San Jose for contribute to tropical biology and conservation a midterm data analysis and individual oral as well as the social sciences and humanities presentation to the group at the ACM office. through our students’ research. Bringing computerized data allows additional Three oral research reports using Power- review, analysis and design changes among Point are given: a) before leaving for the field student, advisor and research coordinator. As the student explains the proposed research mentioned, oral presentations permit informa- proposal, b) at the midterm there is a progress tion exchange and learning among all students. report, and c) a presentation is given based The research objective for the last month on the research project and final paper. This (May) is to analyze the data and write the final allows colleagues, advisors, public and staff to paper which is presented in an oral presenta- participate in the student’s research progress. tion. The final paper is written according to We consider the final oral presentation as a the format of the research journal International preliminary step to doing the same in national Journal of Tropical Biology and Conservation and international symposiums in the students’ and includes the following parts: title, abstract, professional fields (Table 3).

Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 843-858, September 2014 847 Student Applied Research Over the last half century, the ACM Costa and Its Contribution Rica experience has impacted many of its 1 000 research students and others in different forms. As stated, most of the ACM research This includes: a) learning to carry out struc- would be considered “applied research” and tured applied research, b) developing profes- objectives include: a) ensure that students learn to conduct all the steps of a “mini-the- sional skills in many fields, c) a strong sense of sis”, and b) focus on interdisciplinary themes student and local organization empowerment, combining natural and social sciences and and d) being able to generate knowledge of humanities. In most cases, students focus on benefit to Costa Rica with respect to past devel- current challenges in Costa Rican society or opment, current realities and future opportu- at a Central American or international level. nities within the country. We have included As an example of these projects, the May 22, comments from several former students dating 2013 symposium for the Spring 2013 students back to 1968 (Table 4). included research themes which involved: a) how tourism influences the artisanal identity in REFERENCES Brunca artwork, b) adolescent tobacco knowl- edge and perception in a rural community, c) Associated Colleges of the Midwest. (2013a, October 8). public health related to dengue knowledge and History. Retrieved from ACM Expand your world. perception in a rural community, d) interactions Associated Colleges of the Midwest: http://www. between humans and white-faced capuchins in acm.edu/ACM History a resort restaurant, e) rural community perspec- Associated Colleges of the Midwest. (2013b, October 8). tives of a national park, f) economic analysis Mission. Retrieved from ACM Expand your world. of the impact of “responsible” fishing on arti- Associated Colleges of the Midwest: http://www. sanal fishing in a coastal town, g) reduction of acm.edu/Mission Monilia roreri infection in an organic cacao Associated Colleges of the Midwest. (2013c, October 8). plantation using a natural fungus, and h) fac- Off-campus program history. Retrieved from ACM tors related to infection risk of coffee plants by Expand your world. Associated Colleges of the coffee leaf rust (Table 3). Midwest: http://www.acm.edu/Off-campus program history ACM fosters in students a strong sense of responsibility as researchers and commitment Brewer, L. & Cunningham, K. (2010). Capturing study abroad´s transformative potential. In L. Brewer & K. to the goal of making a meaningful contribu- Cunningham (Eds.). Integrating study abroad into the tion to the local communities (agricultural or curriculum. Sterling, VA: Stylus Publishing. fishing cooperative, tourist resort, protected Dennis, P. (Compiler) (1989). 25 Anniversary: A.C.M. area, etc.) where they carry out their research. Costa Rica 1964-1989. Associated Colleges of the Some students provide their local communities Midwest, Costa Rica Programs, San José, Costa Rica. and organizations with printed copies, disks National Institutes of Health. (2013, October 15). Protec- or posters of their research. A program goal is ting Human Research Participants. NIH Office of to increase the availability of student research Extramural Research. Retrieved from National Ins- findings and recommendations that can serve to titutes of Health: http://phrp.nihtraining.com/users/ empower both students and local organizations. login.php

848 Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 843-858, September 2014 Appendix 1 LIST OF ACM COSTA RICA SPRING STUDENT PUBLICATIONS IN THE REVISTA DE BIOLOGÍA TROPICAL

Updated: October 10, 2013

Myers, M., Wagner, J., & Vaughan, C. (2011). Long-term in neotropical agricultural landscapes during the comparison of the fish community in a Costa Rican dry season. Revista de Biología Tropical, 49(3-4), rocky shore marine reserve. Revista de Biología Tro- 1199-1206. pical, 59(1), 1-13. Willis, S. & Cortés, J. (2001). Mollusks of Manuel Antonio Spanier, M. (2010). Beach erosion and nest site selection National Park, Pacific Costa Rica. Revista de Biolo- by the leatherback sea turtle Dermochelys coriacea gía Tropical, 49 (Suplemento 2), 25-36. (Testudines: Dermochelyidae) and implications for Vaughan, C. & Foster Hawkins, L. (1999). Late dry season management practices at Playa Gandoca, Costa Rica. habitat use of common opossum (Didelphis mar- Revista de Biología Tropical, 58(4),1237-1246. supialis) in neotropical lower montane agricultural Guittar, J., Dear, F., & Vaughan, C. (2009). Scarlet Macaw areas. Revista de Biología Tropical, 47(1-2), 263-269. (Ara macao, Psittaciformes: Psittacidae) Nest Cha- Rica program Costa Special section (ACM) Vaughan, C. & Weis, K. (1999). Neotropical dry forest racteristics in the Osa Peninsula Conservation Area wildlife water hole use and management. Revista de (ACOSA), Costa Rica. Revista de Biología Tropical, Biología Tropical, 47(4), 1039-1044. 57(1-2), 387-393. Vaughan, C. & Shoenfelder, S. (1999). Dry season activity, Van Hulle, M. & Vaughan, C. (2009). The Effect of Human movement, habitat and den utilization of nine-banded Development on Mammal Populations of the Punta armadillo (Dasypus novemcinctus) in neotropical Leona Private Wildlife Refuge, Costa Rica. Revista de Biología Tropical, 57(1-2), 441-449. dry forest, Costa Rica. Revista de Biología Tropical, 47(4), 1117-1119. Vaughan, C., Bremer, M., & Dear, F. (2009). Scarlet Macaw (Ara Macao) (Pssitaciformes: Psittacidae) Vaughan, A., Cabrera, J., Moore, D., & Vaughan, C. Parental Nest Visitation in Costa Rica: Implications (1997). Water loss at twenty natural waterholes in for Research and Conservation. Revista de Biología Guanacaste National Park, Costa Rica. Revista de Tropical, 57(1-2), 395-400. Biología Tropical, 45(4), 1681-1684. Vaughan, C., Nemeth, N., & Marineros, L. (2006). Obser- Johnson, W. & Vaughan, C. (1993). Habitat use of small vations of Scarlet Macaw (Ara macao) diet in Central terrestrial rodents in the Costa Rican highlands. Pacific Costa Rica. Revista de Biología Tropical, Revista de Biología Tropical, 41(3), 521-527. 54(3), 919-926. Allsteadt, J. & Vaughan, C. (1988). Distress calls of Timmock, J. & Vaughan, C. (2002). A Census of Mammal Caiman (Caiman crocodilus fuscus) in Caño Negro Populations in Punta Leona Private Wildlife Refuge, Wildlife Refuge, Costa Rica. Revista de Biología Costa Rica. Revista de Biología Tropical, 50(3/4), Tropical, 36, 567-568. 1169-1180. Moscow, D. & Vaughan, C. (1987). Food habits and troop Paynter, C., Cortés, J., & Engels, M. (2001). Biomass, pro- movement of the White-faced Monkey (Cebus capu- ductivity and density of the seagrass Thalassia testu- cinus) in a tropical dry forest. Revista de Biología dinum at three sites in Cahuita National Park, Costa Tropical, 35(2), 287-297. Rica. Revista de Biología Tropical, 49, 265-272. Phillips, P. & Perez-Cruet, M. (1984). A comparative Williams, H. & Vaughan, C. (2001). White-faced monkey survey of reef fishes in Caribbean and Pacific Costa (Cebus capucinus) ecology and management Rica. Revista de Biología Tropical, 32(1), 95-102.

Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 843-858, September 2014 849 Appendix 2 – Tables

Table 1 Breakdown of research projects by field (1964-2013).

Natural sciences Field of Study Number of projects Percentage of total projects Agriculture 41 3.8% Botany 83 7.6% Chemistry 3 0.3% Entomology 37 3.4% Environmental Studies 89 8.2% Geography 11 1.0% Geology 27 2.5% Marine Biology 78 7.2% Ornithology 60 5.5% Zoology 103 9.5% Natural science total 532 48.9% Social sciences and humanities Field of Study Number of projects Percentage of total projects Anthropology 108 9.9% Archaeology 46 4.2% Economics 95 8.7% Education 24 2.2% Fine Arts 4 0.4% Health and Nutrition 90 8.3% History 5 0.5% Law 3 0.3% Literature 5 0.5% Music 2 0.2% Political Science 44 4.0% Psychology 16 1.5% Sociology 87 8.0% Women´s Studies 27 2.5% Social science and humanities totals 556 51.1% Total projects in all fields of study 1 088 100.00%

Table 2 Spring 2013 timetable. January - June, 2013

I. Orientation Period - January 29 - March 1. ACM Center in San Pedro ACM operations and logistics. Spanish language and Costa Rican culture course, directed by Mario Morera. Field research: project selection and research design preparation, consultation with advisors and methodology sessions, cultural activities and field trips.

January 29 Meeting with ACM Director at the ACM (10:00 - 11:30 a.m.) February 5 Meeting with ACM Director at the ACM (11:00 - 12:00 midday) Lunch and meeting with students (12 midday - 3 p.m.) February 7 Advisor meets with student (between 2 and 4:30 p.m.) Student submits paragraph describing project to advisor, ACM director and area coordinator (by 4:30 p.m.) February 11 Advisor meets with student to coordinate the visit to the research site and to discuss what equipment the student will need to carry out his/her research (2-4:40 p.m.).

850 Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 843-858, September 2014 TABLE 2 (Continued)

February 12 Advisor informs ACM Director where student will be carrying out field research for host family assignment. February 14 Student submits project outline of his/her research to advisor, ACM Director and area coordinator (by 4:30 p.m.) Student submits a list of materials and equipment needed for research to Administrative Coordinator. February 18 Advisor returns project outline with comments (12 midday). February 19 Student submits first draft of his/her research proposal to advisor, ACM Director and area coordinator (by 4:30 p.m.) Confirm information related to site visit with Administrative and Academic Services Coordinators (dates of visit, means of transportation, where the advisor and the student will stay, visit to host family, among other details). February 21 Advisor, ACM Director and area coordinator return first draft of research proposal with comments (by 4 p.m.) February 22, 23 and/or 24 Student visits research site with advisor. February 25 Advisor submits brief e-mail report on student progress including: research proposal,

experimental design, equipment needs, site visit and host family assignment (by 10 a.m.) Rica program Costa Special section (ACM) Hand in sheet with host family contact information to the Coordinator of Academic Services in the case of a new family Student submits second draft of his/her research proposal to advisor, ACM Director and area coordinator (by 4:30 p.m.) February 26 Advisor, ACM Director and area coordinator return second draft of research proposal with comments (by 4 p.m.) February 27 Advisor practices oral presentation of research proposal with student. Student submits final research proposal to advisor, ACM Director and area coordinator (by 4:30 p.m.) February 28 Sign up for interim oral presentation times (April 2) with the ACM Director. PowerPoint presentation of research proposal to ACM (12:30 - 4:30 p.m.)

II. Field Research Period - March 2 - April 26. Various Field Sites

March 2 or 3 Student travels to research site. March 4 - March 29 First period of field work. March 30 or 31 Students return to San José. April 1 Student meets with advisor, ACM Director and area coordinator for preliminary analysis of data and preparation of PowerPoint presentation. April 2 PowerPoint oral interim reports. April 3 Students return to research site. April 4 - April 26 Second period of field work. April 24 Advisors meet with Director at the ACM (11 a.m. - 12 midday). April 27 Students return to San José.

III. Analysis and Report Period - April 29 - May 23 ACM Center, San Pedro

April 29 Student consults with advisor about data analysis and report writing (2 - 4 p.m.) May 2 Student sends outline of final paper format to advisor, ACM Director and area coordinator (by 4:30 p.m.) May 3 Advisor returns outline with comments via Dropbox (by 4 p.m.) May 8 First draft of final paper turned in to advisor, ACM Director and area coordinator (by 4 p.m.) May 9 Advisor, ACM Director and area coordinator return first draft of paper with comments (by 4 p.m.)

Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 843-858, September 2014 851 TABLE 2 (Continued)

May 10 Student consults with advisor May 13 Second draft of paper handed in to advisor, ACM Director and area coordinator (4 p.m.) The Director will contact you to coordinate the final oral presentation schedule Student submits Spanish abstract of final paper to the Spanish Language Coordinator (by 11 a.m.) May 15 Advisor, ACM Director and area coordinator return second draft of paper with comments. May 15, 16 and 17 Final written and oral language exams with the Spanish Language Coordinator (8 - 10:00 a.m.) May 17 Final draft of paper handed in to advisor, ACM Director and area coordinator (by 4 p.m.) May 20 Practice for oral presentation with advisor May 22 Final oral reports. May 23 End of the program.

IV. Final Evaluation Period - May 27 - June 7.

Advisors hand in final grades and written evaluation of students´ work on or before Friday, June 7. Please keep in mind that the ACM building closes at 4:30 p.m., Monday through Friday.

Table 3 Spring 2013 symposium.

Acm Spring 2013 Research Presentations May 22, 2013 9:30 a.m. OPENING WORDS Christopher Vaughan, PhD 9:35 a.m. INFLUENCE OF TOURISM ON ARTISANAL IDENTITY IN BRUNCA ARTWORK Brandylyn Arredondo, Kenyon College 9:50 a.m. USES AND PERCEPTIONS OF MEDICINAL PLANTS AND NATURAL MEDICINE BY THE ORGANIZATION PROAL-HOLOSALUD AND THE COMMUNITY OF LLANO BONITO, COSTA RICA IN MARCH AND APRIL OF 2013 Jenny Asparro, St. Olaf College 10:05 a.m. CONOCIMIENTOS Y PERCEPCIONES SOBRE EL TABAQUISMO Y SUS EFECTOS DAÑINOS EN LA SALUD ENTRE LOS ESTUDIANTES ADOLESCENTES DE VENECIA Y AGUAS ZARCAS, COSTA RICA Nicholas Bohrer, St. Olaf College 10:20 a.m. KNOWLEDGE, PERCEPTIONS AND PRACTICES WITH RESPECT TO THE PREVENTION OF DENGUE IN A MID-PACIFIC COASTAL VILLAGE OF COSTA RICA Victoria Egedus, Lake Forest College 10:35 a.m. FISHING EFFORT AND PREDATORY BEHAVIOR OF BROWN PELICANS (Pelecanus occidentalis) ON THE CENTRAL PACIFIC COAST OF COSTA RICA Lauren Heber, Colorado College 10:50 a.m. INTERACTIONS BETWEEN HUMANS AND WHITE-FACED CAPUCHINS (Cebus capucinus) IN A MID-PACIFIC COASTAL RESORT Elizabeth Landry, Colorado College 11:05 a.m. BREAK 11:20 a.m. EFFECTS OF HUMAN DISTURBANCE ON GHOST CRABS (Ocypode) IN COSTA RICA

852 Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 843-858, September 2014 Michael Maurer, Colorado College 11:35 a.m. COMMUNITY PERSPECTIVES OF ARENAL NATIONAL PARK: RESOURCE MANAGEMENT AND RELATIONSHIPS Margaret McKeon, Colorado College 11:50 a.m. CHARACTERIZATION OF VOLCANIC ASH, PRODUCED BY THE 1963-1965 ERUPTIONS OF IRAZÚ VOLCANO; COSTA RICA Margo Regier, Beloit College 12:05 a.m. EFECTOS ECONÓMICOS DEL ÁREA MARINA DE PESCA RESPONSABLE PARA LOS PESCADORES ARTESANALES DE TÁRCOLES EN EL PACÍFICO DE COSTA RICA Emily Rhoades, Grinnell College 12:20 p.m. LUNCH 1:30 p.m. A COMPARISON OF BIRD POPULATIONS IN LIVE FENCES, RIPARIAN FORESTS, SHADE- GROWN CACAO, AND PRESERVED TROPICAL RAINFOREST IN NORTH-EASTERN COSTA RICA Skye Greenler, Colorado College Special section (ACM) Costa Rica program Costa Special section (ACM) 1:45 p.m. USE OF TRICHODERMA FUNGI IN SPRAY SOLUTIONS TO REDUCE Monilia roreri INFECTION OF Theobroma cacao FRUITS IN NORTHEASTERN COSTA RICA John Seng, Grinnell College 2:00 p.m. AN ANALYSIS OF FACTORS AFFECTING INFECTION RISK OF COFFEE PLANTS BY COFFEE LEAF RUST FUNGUS (Hemileia vastatrix) IN THE LOS SANTOS REGION OF COSTA RICA, 2013 Justine Decker, Macalester College 2:15 p.m. SPECIES DIVERSITY AND ACTIVITY OF INSECTIVOROUS BATS IN THREE HABITATS IN LA VIRGEN DE SARAPIQUÍ, COSTA RICA Amanda Cormier, Colorado College 2:30 p.m. DRY SEASON HABITAT USE AND HOME RANGE OF THE WOOLLY OPOSSUM (Caluromys derbianus) IN SARAPIQUÍ RAINFOREST, COSTA RICA Danica Lewis, Knox College 2:45 p.m. CLOSING WORDS Christopher Vaughan, PhD

Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 843-858, September 2014 853 Table 4 Former ACM Costa Rica student comments.

Peggy Barlett, PhD (8 in cover) Goodrich C. White Professor of Anthropology, Emory University ACM Costa Rica Spring 1968 The ACM program experience was life-changing for me. I was always interested in Latin America, but I didn’t know how much I was drawn to rural communities until I lived with a family in Nicoya, Guanacaste, during the fall of my senior year. In our ACM classes, I was introduced to development theories—which has become a lifelong focus—and in particu- lar the work of Boserup on the role of population and agriculture. I still introduce my own graduate students to Boserup’s important influence in social science and development perspectives. Our fieldtrips to sugar and coffee plantations, benefi- cios, cattle farms, and small, diversified householders gave me knowledge of commodity systems and smallholders before those terms were in use. I have always loved the way I can picture in my mind’s eye the movement of coffee from bush to bean to processing to shipment to roasting to cup. I even learned how to roast beans on a wood stove, grind them, and enjoy the resulting beverage in a rural farmhouse, in the company of encouraging and welcoming people. The ACM experiences were not only an introduction to my life as an anthropologist—and I applied for graduate training by candlelight in longhand from the cot in the home of my host family—but also an introduction to our human commonalities behind the differences in our lifestyles and consumer habits. I learned to shed a lot of things I thought I couldn’t live without, learned to eat many new foods, learned a new language for communication, and learned the habits of observation, listening, careful attention to details, and the joy of posing a question and finding an intriguing answer. The ACM group led to lifelong friendships, which I value highly. And it cemented my commitment to anthropology, economic and development processes, and what today we call sustainability, but back then was simply, “how can Costa Rican families meet the economic, ecological, demographic, political, and cultural challenges they face and move toward a more just and livable future?”

Kathleen L. Shea, PhD (6 in cover) Professor of Biology and Environmental Studies, Curator of Natural Lands, St. Olaf College ACM Costa Rica Spring 1969 Over forty years ago I spent the spring semester of my junior year (1969) as an undergraduate at Grinnell College in Costa Rica as part of the ACM Field Research Program. That experience changed my life in numerous ways. The program gave me the opportunity to do an in-depth research project on sugar cane varieties and yields. Through this project I realized that I loved the research process and wanted to go to graduate school. I liked project planning and data collection as well as the analysis parts of research. I gave my first presentation at a scientific meeting, the Iowa Academy of Science Meeting in the spring of my senior year, on my project in Costa Rica. While in graduate school my experiences in Costa Rica helped me be accepted to an Organization for Tropical Studies graduate course on tropical ecology and a Missouri Botanical Garden course on the Flora of Panama. My experiences in Costa Rica have resulted in a lifelong interest in this country. Living and working with local people was especially valuable in learning Spanish and learning about the culture of Costa Rica. As a professor at St. Olaf College I have taken students to Costa Rica for a January Interim course several times and I am the campus adviser for the ACM Field Research Program. My interests in conservation and sustainable agriculture really began in Costa Rica. I have focused on the issue of how to balance conservation and the need people have to make a living since I was a student in Costa Rica. I still show my current students pictures from my Costa Rica experiences. With my current students we study tropical ecology, sustainable land use, and how conservation can be successful over the long-term. Costa Ricans are proud of their natural resources and have established a viable national park and reserve system because local people see its value and potential to provide jobs through ecotourism. While students today have more options for off-campus study this is one program that has stood the test of time and should be considered. My life has been much richer for having this tropical connection!

Christopher Vaughan, PhD (Background up in cover) Director, ACM Costa Rica Programs; Honorary Associate Fellow, Department of Forestry and Wildlife Ecology, University of Wisconsin-Madison; Retired Director, International Program for Wildlife Conservation and Management, Universidad Nacional, Costa Rica ACM Costa Rica Spring 1969 ACM provided me with the ultimate education in experiential learning- a structured program, with a study site and advisor for ensuring correct and timely field data collection to complete university requirements. But the ACM experience also allowed me time to think outside the normal university setting, interact with friendly ticos, explore and observe tropical nature on a daily basis and how many humans were impacting nature while a few were trying to conserve it with a multitude

854 Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 843-858, September 2014 of ideas. While in Costa Rica with ACM, I decided that my mission in life would be to try to understand and work with local institutions and communities to conserve/manage tropical nature. Once this decision was made, my life became very focused, meaningful, and challenging. I am continuously learning from nature, people and literature/documentaries and it is a pleasure to work with the younger generation and expose them when possible to an ACM type experiential learning.

Ken Salazar, JD (14 in cover) Partner in the law firm WilmerHale Secretary of the Interior of the United States, 2009-2013 U.S. Senator for Colorado, 2005-2009 Attorney General of Colorado, 1999-2005 Executive Director of the Colorado Department of Natural Resources, 1990-1994 Chief legal counsel to the Governor of Colorado, 1986-1990 ACM Costa Rica Spring 1975 In 1975, I participated in the Associated Colleges of the Midwest Program in Costa Rica. The experience shaped much of my future including a deep interest in Latin America. Many times in the United States Senate and at the White House, I have spoken with Presidents and Senators about the imperative of strong mutually supportive relationships with Central and South America. I am rooted in this view in part because of the ACM program and my study of President Kennedy’s

Alliance for Progress. I believe the future of the United States, and Latin America, are inextricably tied together because of Rica program Costa Special section (ACM) our shared history and shared demographics.”

Robin Chazdon, PhD Full Professor, Department of Ecology and Evolutionary Biology, University of Connecticut ACM Costa Rica Spring 1976 My experience as an ACM student opened the doors to the world of tropical biology and Latin America for me. I went during my sophomore year and had only had 3 semesters of Spanish. The immersion into Spanish was very effective; I have been a fluent Spanish speaker since. The course gave me exposure to the geographic, cultural, and biological heterogeneity of Costa Rica, stimulating me to pursue doctoral research in Plant Ecology in Costa Rica and to spend over thirty years of fieldwork in the wet tropical lowlands. I have always felt comfortable travelling throughout Latin America and have develo- ped close professional and personal relationships with Latin Americans. I attribute much of my ease to my early experiences as an ACM student and to the early bonds that I formed with Costa Rican scientists and friends. I went on to pursue a Ph.D. in Ecology from Cornell University, studying understory palms at La Selva Biological Sta- tion in the Atlantic lowlands. After starting my faculty position at the University of Connecticut in 1988, I became interested in studying the process of secondary forest regrowth in wet tropical regions, and once again returned to La Selva where I am now involved in a long-term project on vegetation dynamics in secondary and mature forests. I am also starting a similar project in the Osa Peninsula with several Costa Rican collaborators. As a mentor, I encourage all students to study abroad and to immerse themselves in other cultures and biological worlds. My faculty mentors at Grinnell, Laverne and Leonore Durkee, played a major role in inspiring me to pursue a career in field- work and tropical biology and I am forever grateful for the wonderful opportunities that the ACM course offered me during my most “impressionable” years. To date, I have trained 11 graduate and 2 undergraduate students; all have conducted their research in Costa Rica. I am strongly committed to helping Costa Ricans protect and manage their precious biological heritage, to restore forests, and to improve their own livelihoods and education.

Warren Johnson, PhD Staff Scientist, Laboratory of Genomic Diversity, National Cancer Institute, NIH ACM Costa Rica Spring 1982 As a Junior at Oberlin College, I spent a semester abroad in Costa Rica as part of the ACM program. It remains one of the most memorable and formative experiences of my life and was crucial in helping me down my career path. I have many memories of that time, but perhaps one of the most illustrative was a trip to Volcan Arenal with fellow students: one who kept pointing out the interesting geology, one that kept commenting on the plants, and a third that had a keen eye for all of the birds. We spent that night on the top of Volcan Arenal with an Italian volcanologist, and I, as the mammalogist, wondered at the sight of the howler monkey several thousand meters above the tree line. In Costa Rica, I had the opportunity to greatly improve my very rusty Spanish, to learn from fellow students from a variety of backgrounds and interests, to learn and feel a part of a different culture, and to have my first taste of field research. The mix of academic study, full immersion with local

Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 843-858, September 2014 855 families, and publishable research is what sets the program apart from others. I have been many places since then, but can truly say that the ACM program helped launch my career in international research which I continue today.

Thomas Sisk, PhD Olajos-Goslow Professor of Environmental Science and Policy, School of Earth Sciences and Environmental Sustainability, Northern Arizona University ACM Costa Rica Spring 1982 The ACM experience was my first introduction to what research was all about. It opened my eyes to the discovery that I could use my education to learn something new, both fascinating in itself and useful to others, realizing it changed my life. The people I met during my ACM semester in Costa Rica are still with me after all these years. Remembering their faces brings back some of the most important lessons I ever learned.

Lisa Naughton, PhD Professor, Department of Geography, University of Wisconsin-Madison ACM Costa Rica Spring 1984 My ACM semester in Costa Rica forever changed me and deepened my understanding of tropical ecology and Latin American society. I enrolled in the semester to fulfill my dream to “be Jane Goodall.” Thanks to the guidance and mentoring of my ACM field biology professor (Dr. Chris Vaughan), I did indeed spend two months studying squirrel monkey behavior. I loved the research, but after witnessing rapid deforestation and uncontrolled tourism development, my career interests shifted. From that semester on, I pursued a career in applied conservation research, mainly in tropical countries.

Lic. Judith Magnan (10 in cover) Coordinator of Academic Services, ACM Costa Rica Official Translator, Ministerio de Relaciones Exteriores y Culto de Costa Rica ACM Costa Rica Spring 1985 In 1985, the ACM program gave me the unique opportunity of learning how to carry out field research in the extraor- dinarily biodiverse setting of Playa Nancite in the Santa Rosa National Park under the guidance of conservation expert and author, Christopher Vaughan. Intense language instruction combined with cultural immersion led me to make Costa Rica my home, while developing a career in intercultural education and communication in order to facilitate the exposure of other students to such a challenging and enriching experience.

Michael Pignone, MD, MPH Professor of Medicine and Chief of the Division of General Internal Medicine, University of North Carolina- Chapel Hill ACM Costa Rica Spring 1987 The ACM Costa Rica Field Research program has had a profound effect on my career. The research I did in Costa Rica served as the basis for my undergraduate honors thesis and helped me decide to pursue a career in academic medicine. My language training has allowed me to continue to serve Spanish-speaking patients here in the U.S. throughout medical school, residency, and as a faculty member now at the University of North Carolina.

R. John Kantner, PhD, RPA Vice President and Professor, Office of Research and Sponsored Programs, University of North Florida ACM Costa Rica Spring 1988 My experience in the ACM Costa Rica program still stands out as one of the most influential years in my life. I dis- covered my passion for archaeology and anthropology during that year, a passion that I still pursue today as a university professor. Perhaps just as importantly, I learned a tremendous amount about myself, for my time in Costa Rica taught me that I can achieve any goal I set my mind to, whether it was learning Spanish, traveling through the countryside alone, or climbing Chirripo to view both the Pacific and Atlantic Oceans. I still read my diary from the ACM program, recalling all the fond memories from years ago.

856 Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 843-858, September 2014 Nicole Nemeth, DVM, PhD (4 in cover) Assistant Professor, Department of Pathobiology, Ontario Veterinary College, University of Guelph ACM Costa Rica Spring 1994 ACM really helped cultivate and develop my interests in a very unique and special setting with excellent guidance and long-lasting mentors that have continued to have a presence in my life. ACM provided me with my first opportunity to work in wildlife ecology and conservation, which subsequently motivated me to complete both veterinary and graduate school, followed by a wildlife pathology residency, leading to a position as Assistant Professor at the University of Guelph, where I will help students develop their interests in wildlife health.

Mark Myers, PhD (11 in cover) Assistant Professor, University of Northern Iowa, ACM Costa Rica Spring 1995 I considered a number of study abroad programs before choosing ACM, and what appealed to me about the Costa Rica Field Research program was that it allowed me to pursue independent research at a single site over several months, rather than “touring” many sites for just a few days at a time with larger groups of students, as was typical of many other programs. The educational “pay-off” of this approach, both in terms of insights gained from the many hours spent in the field and in terms of the language skills and cultural experiences gained from living in a small, rural community, was huge. My expe- rience in Costa Rica was a strong influence in my decision to pursue a graduate degree in Conservation Biology, and many

of the experiences and contacts I made laid the groundwork for future projects in Costa Rica. Rica program Costa Special section (ACM)

Jonathan Henn (5 in cover) Fulbright Scholarship recipient working on plant regeneration, Patagonia, Argentina ACM Costa Rica Spring 2011 My semester in Costa Rica was a critical period in my professional and personal development. The cultural immersion, Spanish language practice, and independent research experience has opened many doors as I have continued working toward my goal of working in conservation. For me, the most important part was the independence and freedom you were allowed when developing your project. Instead of being handed an already developed project, the chance to think creatively and critically about what question you would ask and how you would go about answering it was a unique insight into the process of science that is not usually part of an undergraduate education. The feeling of ownership that comes with developing a project from proposal to presentation was very useful. These skills and experiences that I developed during my semester in Costa Rica led to a Fulbright grant in Argentina and I’m sure they’ll continue serving as I advance in my career.

Jeffrey Nadel, BA (15 in cover) Deans Scholar and Medical Student, University of Michigan Medical School Formerly: Research Assistant at the Engelberg Center for Health Care Reform at the Brookings Institution Post-baccalaureate IRTA Fellow, National Institutes of Health ACM Costa Rica Spring 2011 My experience as a student in the ACM Costa Rica Field Research Program was transformative and changed me both personally and professionally. My research on elderly mental health was the primary force that pushed me towards a career serving at the nexus of clinical medicine, public health, and policy, working to help eliminate social and health disparities within and between communities. The experience also taught me that no challenge is too great, and that compassion and thoughtfulness can enable amazing connections between people. I still talk to my host families regularly and have been back to Costa Rica to visit them both.

Emma Cornwell (12 in cover) Fulbright Scholarship recipient, ACM Costa Rica Spring 2012 I am so grateful for the opportunity to study abroad in Costa Rica. The ACM program, established in the country for decades, has gathered an incredibly strong network of host families and research advisors. This provided me with expe- riences that would be hard to find in other study abroad programs: two Costa Rican families with whom I have formed unforgettable lifelong bonds, and an extensive research project of which I can feel proud.

Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 843-858, September 2014 857 Justine Decker Macalester College, ACM Costa Rica Spring 2013 The organization and preparedness demonstrated by the ACM Costa Rica staff was absolutely incredible. The program has established hugely valuable connections within the country that make it much easier for students to quickly become engaged within Costa Rican culture and society. These connections were what allowed me to become a part of two wonder- ful host families and to have the opportunity to work side-by-side with Costa Rican coffee farmers.

John Seng (13 in cover) Grinnell College, ACM Costa Rica Spring 2013 My semester in Costa Rica with the ACM was extremely empowering. As a student, I learned to manage and take advan- tage of the freedom that comes with conducting independent research. My research skills and Spanish fluency improved greatly, more than I could have imagined. As an individual, I became more confident and adventurous. I wouldn’t trade this experience for anything.

Amanda Cormier (2 in cover) Colorado College, ACM Costa Rica Spring 2013 ACM Costa Rica was an amazing opportunity. I got to spend two months doing research in the jungle, I got to work with amazing people, and I got to experience a different culture. Everyone who works at ACM was wonderful along with both of my host families and the advisors. I felt like I was part of multiple families and I would love to return to see all of these excellent people again.

1. Mike McCoy. 3. Lauren Carlson. 7. Shasta Webb. 9. Victoria Egedus. Knowledge, perceptions, and practices with respect to the prevention of dengue in a mid-Pacific coastal village of Costa Rica

Victoria L. Egedus1,2, José Morales Ortega3 & Anabelle Alfaro Obando4 1. Associated Colleges of the Midwest, San Pedro, Costa Rica 2. Lake Forest College, Lake Forest, IL 60045 USA; [email protected] 3. Ministerio de Salud, Área Rectora Garabito Jacó, Puntarenas, Costa Rica; [email protected] 4. Universidad Nacional, Costa Rica; [email protected]

Received 20-iii-2014. Corrected 12-iv-2014. Accepted 13-v-2014.

Abstract: Dengue fever is the fastest spreading, most prevalent and deadly arthropod-borne viral disease world- Rica program Costa Special section (ACM) wide, present in over 125 countries. In 2013, Costa Rica is experiencing the highest number of Dengue infec- tions since the virus’s reintroduction to the country in 1993. This study evaluated the Dengue-related knowledge of 320 community members, and polled opinions on how to better educate and mobilize the community on Dengue prevention in Quebrada Ganado, Costa Rica during March-May, 2013. In addition, property inspections were used to find relationships between knowledge or opinions and presence of Aedes breeding sites. Results showed that while citizens knew the virus is transmitted by mosquitoes that reproduce near water, they lacked knowledge on Aedes habits and confused it with other mosquitoes. Eighty-one percent of respondents assumed some responsibility for dengue prevention. Suggestions for improved education included consistent and con- tinual information on the risks and dangers dengue poses to an individual. Characteristics relating to households with more positive breeding sites were: lower education level, higher prevalence of dengue infections in the household, lesser knowledge of dengue symptoms, a lower rating on the dangers of dengue and a lower rating on the importance of preventative actions. While a range of prevention methods are implemented, active com- munity involvement is highly important for successful Dengue prevention. Continual evaluation is necessary to make more immediate, long-term behavioral societal changes, and to maximize the economic resources spent on Dengue prevention. Rev. Biol. Trop. 62 (3): 859-867. Epub 2014 September 01.

Key words: dengue fever, Costa Rica, social communication, Aedes aegypti, arbovirus, KAP.

Dengue fever is a viral illness caused by World War II when a global dengue pandemic any one of the four DENV flavivirus serotypes began (reviewed in Gubler, 1998). To this day, (DENV1-DENV4). It is the most prevalent and no dengue vaccine has been approved, and deadly arthropod-borne human illness world- prevention methods rely heavily on diminish- wide (Guzman & Kouri, 2002; Farrar et al., ing the vector. In 1947 the Pan American Sani- 2007; Ross, 2010) and is transmitted by mos- tary Bureau initiated a campaign to eradicate quitoes of the genus Aedes, primarily Aedes Aedes aegypti from the Western Hemisphere, aegypti. Infection with one serotype may cause and by 1961 Aedes was completely eradicated dengue with or without warning signs, or from many Latinamercan countries, includ- severe dengue. Subsequent acquisition of dif- ing Costa Rica (Severo, 1955; Soper, 1963). ferent serotypes is associated with a weakened While dengue continued to be a major health immune response and worsens the perspective problem in Asia, the Americas remained clear for recuperation. approximately between 1950-1970 (Gubler, Dengue epidemics were reported in 1779, 1998). The eradication campaign ended, and but no persistent infection patterns existed until during the 1970s the Pacific Islands were

Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 859-867, September 2014 859 re-infected with the virus, leading to the spread needed to optimize community involvement of dengue across the Western Hemisphere once (Toledo-Romaní, Baly-Gil, Ceballos-Ursula, again (Gubler, 1998). Today the number of Boelaert, & Van der Stuyft, 2006). incidences of dengue is 30 times greater than it In order to combat dengue and eradicate was 50yrs ago, with an estimated 390 million the vector, the Pan-American Health Organiza- dengue infections annually (Bhatt et al., 2013) tion, national health care system and the Min- and 2.5 billion people living in areas of risk istry of Health of Costa Rica designed a plan in (WHO, 2009). 2005. The plan, EGI (Spanish for Integrative Exacerbated by increases in climatic tem- Management Strategy), has five integrated peratures, frequent intercontinental airline components, one of which is social commu- travel, booming, uncontrolled urbanization, nication. A goal of social communication is inadequate municipal services, and vector to mobilize public participation in eliminating adaptability, (Parks & Lloyd, 2004; Farrar et the vector; however, attaining and maintaining al., 2007; Miller, 2012) dengue is a pressing long-lasting behavioral changes in a commu- global health concern. Relatively successful, nity is a difficult task (Parks & Lloyd, 2004). the country of Costa Rica remained free of den- It was reported that public education regarding gue until an epidemic outbreak in 1993. Since dengue knowledge is relatively strong in Costa 1993, infections have persisted and posed an Rica, yet education on attitudes and beneficial increasing health threat in Costa Rica, and practices has not been well provided (Solís, there is a severe gap in published dengue Quesada, & Gutiérrez, 2010). research in the country (Troyo, Porcelain, This study aimed to evaluate and report on Calderón-Arguedas, Chadee, & Beier, 2006). the current status of the residents’ knowledge, In 2013 Costa Rica was declared to have the perceptions and actions in regards to dengue in largest dengue epidemic since 1993, and by the the small, mid-Pacific coastal village Quebrada end of August 2013 around 27 000 infections Ganado, Puntarenas, Costa Rica. were reported. Aedes aegypti maintains a lifestyle depen- MATERIALS AND METHODS dent on humans, and lives exclusively in human-populated areas (Trpis, Hausermann, Study site: The study site was Quebra- & Craig, 1995). The female mosquito lays her da Ganado, Puntarenas, Costa Rica (hereafter eggs in clean water, and eggs can lay dormant referred to as Quebrada, or QG), (9.72ºN for over a year in a dry state. While a number - 84.63ºW), with a population of 1 916 (967 of options for diminishing the vector have males, 949 females) (INEC, 2011). Before been considered in the past 20yrs, including the study began, QG had 40 confirmed and predation to kill the mosquitoes, fumigation, reported cases of dengue in the first 16wks of and shorter-lived organophosphate insecticides, 2013 (Ministry of Health). The town has 754 these methods are not successfully executed residences including 583 occupied and 171 and are difficult to manage without com- unoccupied (INEC, 2011). Ninety-nine percent munity participation. Aedes breeding is often of the houses (582/583) have internal running propagated by the failure to remove breeding water, and 99% (577/583) have a television sites, such as old tires, flower vases, random (INEC, 2011). In general, houses are very close garbage, and even pop bottle caps that should together and space is concentrated, including a be properly disposed of. Studies show that in community center, soccer field, kindergarten, working to eliminate dengue, community edu- pharmacy, primary and secondary schools, cation and involvement are crucial (Espinoza- and public health clinic. The most commonly- Gómez, Hernández-Suárez, & Coll-Cárdenas, reported resident occupations are personal ser- 2002; Parks & Lloyd, 2004; Martín & Prado, vices, cleaning services, sales and construction 2004; Nam et al., 2005), yet much effort is still (INEC, 2011).

860 Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 859-867, September 2014 Study design: To correlate resident between various groupings of participants were knowledge and attitudes with their respective estimated with one-way parametric ANOVA. property, and to evaluate the behavior of QG Assumptions of homogeneity of variance were citizens, only citizens at residential properties tested and posterior comparisons between were interviewed. Our 25-item interview was means were carried out with LSD, Tukey or adapted from a similar interview designed to Scheffe tests. Statistical procedures followed aid the Ministry of Health (MH) in a 2004 Sokal & Rohlf (1995). Analysis of variance study (R. Baxter, unpublished). All interviews and linear regression analyses were conducted took place in March and April, 2013. Inter- using Statgraphics Centurion XVI (Statpoint view procedures, including interviewee verbal Technologies, Inc., 2011). consent protocol, were approved by the Ethics Review Board of the Associated Colleges of RESULTS the Midwest, Chicago, Illinois. Household inspections were carried out Tables and additional comments appear as during March through May, 2013 by employ- Digital Appendices. The rest of the results are ees of the Vector Control sector of the Garabito detailed below. Ministry of Health of Costa Rica. As part of Rica program Costa Special section (ACM) the EGI branch of Entomological Surveillance, Sample demographics: Of the 19 city MH workers inspect properties for possible blocks in the study site, 18 blocks were vis- and positive Aedes breeding sites, and educate ited and 320 citizens were interviewed. This owners about dengue prevention. included 242 (76%) females and 78 (24%) All answers were directly recorded on an males. Median age was 40 with a range from interview sheet. Independent of how many 18 to 86. Twenty-three (7%) participants had people were present in the house, only one per- no education, 167 (52%) had received educa- son was asked to respond. All respondents were tion at the primary level, 102 (32%) at the at least 18 years of age, lived in the house, and secondary level and 28 (9%) up to university as often as possible, managed household deci- level. Fifty-three percent of participants were sions. To avoid influence by recent in-house stay-at-home mothers or not employed, 14% education by MH employees, we interviewed were blue-collar workers (construction, pack- 182 households with no previous recent MH ing, maintenance, fishing, among others), 13% visit. These properties were inspected after the worked in a restaurant, tourism, or as a cashier interview. An additional 37 interviews were and the other 20% were employed in 10 other conducted directly after a MH inspection, sectors including childcare, independent work and 45 interviews were conducted in house- and security. Sixty-five percent of those inter- holds that had an inspection two months prior. viewed owned their home or paid mortgage and Fifty-six interviews were conducted in house- 35% rented. holds with no previous inspection, and no MH inspection data following the interview were Citizen history with dengue and famil- available for these houses. iarity: Sixty-one percent of participants report- All data were entered in Microsoft Excel ed someone in their house had been previously and rechecked. Dependency of type of water infected with dengue and 21 (11%) of these used for mosquito breeding with prior educa- participants reported it to be “hemorrhagic”. tion by MH was tested with a contingency According to the previous WHO guidelines table. The relationship between Likert scale for dengue classification, community mem- or other types of responses for various inde- bers were more familiar with this terminology pendent and dependent variables of interest than with the newer classification of “severe were estimated by simple linear regression. dengue.” Thirty-nine percent had no house- Differences in mean Likert scale responses member with previous infection. Of those who

Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 859-867, September 2014 861 reported a previous infection in the house, an Community concerns: When asked for average of two people in the house had con- the “most severe problem in the community”, tracted dengue and on average one time rang- 20 people (6%) reported a lack of available ing from a reported one to six times. appointments in the clinic, 12 people (4%) Participants were asked to list the symp- reported garbage thrown in the streets and river, toms of dengue fever, and 89% of participants nine (3%) reported poor management of water could state three or more symptoms. We desig- in the streets, five (1.6%) reported inconsider- nated three or more symptoms as an adequate ation for the environment, four (1.3%) reported understanding of the symptoms of dengue. non-efficient political organizations, and den- The most common symptoms mentioned were gue or excess of mosquitoes was reported by fever (285 participants, 89%), headache (209 13 (5%). The most common answer was illegal participants, 65%), body pain (134 participants, drug use (29%, 94 people). 42%), vomiting (122 participants, 38%), and When asked for the “principal health prob- rash (105 participants, 33%). Of those who lem in QG”, an overwhelming 160 (50%) par- had an adequate understanding, participants ticipants reported dengue. Six (2%) reported mentioned up to nine symptoms (n=2); with their concern for contaminated, stagnant water a mean, median and mode of four symp- in the streets, and four (1.3%) reported a plague toms. Participants with previous acquisition of mosquitoes. were more knowledgeable of the symptoms, (F=76.6; df=1, 318; p<0.0001, R2=19.41%). Citizen perceptions: When asked: How dangerous it is for a person to get infected with dengue? 157 (52%) respondents said it is very Citizen knowledge: Three hundred fif- dangerous, 125 (42%) said it is dangerous, and teen (98%) participants knew that dengue is 18 (6%) said it is not dangerous, or they did transmitted by a mosquito and 98% knew that not know. Mean response for danger level was the mosquito reproduces in stagnant water. greater for respondents that had a house mem- It became evident that the town has common ber with previous dengue infection ( =2.52 on knowledge that mosquitoes reproduce in water, scale from 1-3) when compared to the mean but they grossly generalize all mosquitoes, response for respondents who had no house specifically those around dirty street water, to member previously infected ( =2.32) (F=5.97; carry dengue. When this generalization became df=1, 298; p=0.015). evident, we began asking participants (n=253) Two hundred twelve (71%) respondents what kind of water is preferred by the mosqui- reported that “actions on dengue prevention” toes that transmit dengue. As previously noted, are very important, 75 (25%) reported the participants were interviewed either before or actions are important, three (1%) reported after an inspection by the Ministry of Health the actions are not important, and 10 (3%) workers (Table 1) (Appendix 1). Participants reported they did not know. Mean response reported clean water more often if their house for perceived “importance of actions on den- had been inspected by a Ministry of Health gue prevention” were greater for participants worker 5wks prior, compared with the par- with a previously infected house member ( ticipants with no previous house inspection =2.74) than for those with no previous house- (x2=8.88; df=1; p=0.0029). Education from hold infection ( =2.46) (F=13.23; df=1, 298; MH inspections had an effect on community p=0.0003). Similarly, respondents’ rating on the knowledge 5 weeks later. When asked how “importance of actions on dengue prevention” to decrease the number of dengue infections was greater for respondents who could name in the community, 280 participants (88%) “three or more dengue symptoms” ( =2.66) mentioned eliminating stagnant water (Table compared to those who could not ( =2.38) 2) (Appendix 1). (F=5.25; df=1, 298; p=0.0227). Similarly, those

862 Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 859-867, September 2014 who could name “three or more dengue symp- p=0.0196; R2=2.07). Houses with more pos- toms” perceived a greater “danger of infection” sible breeding sites had more positive breeding ( =2.46) than those who could not ( =2.29) sites for A. aegypti larva/pupa (F=142; df=1, (F=1.77; df=1, 298; p=0.18). 263; p<0.0001; R2=35.17).

Citizen opinions on responsibility and Citizen trends and predictive factors: effective dengue education in the commu- Having a previous infection or knowledge nity: The last three questions in the interview of dengue symptoms did not have a signifi- were used to gather community views, opin- cant relationship with the number of breeding ions and advice on dengue prevention within sites found during inspections. Among the their community. In response to the question: knowledge and demographic factors collected, Who has the responsibility of protecting the participant education level had the greatest community from dengue? 167 (52%) report- relationship to number of positive breeding ed it was their responsibility entirely, 29% sites found in inspections (negative relation- responded the community has responsibility ship) (F=5.92; df=1, 262; p=0.0156; R2=2.21). along with other organizations, and 19% did The number of positive breeding sites was not include the community in their response also greater in homes where a) the respondent Rica program Costa Special section (ACM) (Table 3) (Appendix 1). rated a lower importance on the actions of den- We asked participants: How the com- gue prevention (F=5.08; df=1, 244; p=0.0251; munity can be educated on dengue and its R2=2.05) and b) the rating for dengue dan- prevention? and 406 responses were received. ger was lower (F=5.08; df=1, 244; p=0.0223; Response frequencies from the 320 participant R2=2.13). total were calculated as percentages (Table 4) (Appendix 1). Many comments on dengue DISCUSSION education were also offered, and some of the most common comments were recorded As with other studies, most often women (Table 5) (Appendix 1). We also asked: How of the house were available for interviewing community members can be convinced to (de María Cáceres-Manrique, Vesga-Gómez, eliminate breeding sites on their properties? Perea-Florez, Ruitort, & Talbot, 2009; Zuleta (Table 6) (Appendix 1), and additional com- et al., 2011). Based on our total sample size ments were recorded (Table 7) (Appendix 1). of 42.4% of the residences in QG, the majority of townspeople reported having had a dengue Citizen practices: Based on 264 prop- infection in their household (61%). Nine in erties inspected by MH employees, all 264 ten citizens (90%) have an adequate under- properties (100%) had possible breeding sites. standing of the fever’s symptoms, indicating Mean number of possible breeding sites per they understand dengue fever well, which is house was 8.5 (median=7, mode=5, and ranged comparatively high (Panagos, Lacy, Gubler, from 2-51 per house). Nine percent of the & Macpherson, 2005). An overwhelming 98% properties inspected had at least one breeding understand that dengue is transmitted by a site positive for Aedes aegypti larvae or pupas mosquito, but think erroneously that the mos- (n=24). The quantity of positive breeding sites quito proliferates in stagnant water. However, found on each property ranged from 0 to 6, a portion of participants reported false positive and included a variety of receptacles, includ- dengue cases. They reported up to six distinct ing small buckets, flowerpots, and appliances dengue infections, which is not possible with not in use. The maximum number of times a four serotypes. Citizens also had a defined house member had been infected by dengue confusion on the characteristics of the dengue was positively related to the number of possible vector. Previous studies have shown a similar breeding sites at that house (F=5.51; df=1, 261; trend where communities are familiar with

Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 859-867, September 2014 863 dengue and mosquitoes, but have a poorer reported the best ways to convince commu- understanding of the relationship between the nity members to eliminate stagnant water were mosquito, human behavior and transmission through: campaigns, information on risks and (Panagos et al., 2005). dangers, talks and lectures and the work by the Citizens were not well educated on the MH. That is, the number one way to convince differences in mosquito niches, leaving them members to act is through more information to perceive “a mosquito is a mosquito,” hence, and education. all mosquitoes transmit dengue. For example, The next most popular response was that citizens were preoccupied that dengue was contracting dengue is the only way to convince spread by mosquitoes that are often found in members to act; indicating personal experi- areas where Aedes does not reproduce. They ence is the best teacher. About one in three failed to focus on the primary breeding sites respondents expressed doubts in convincing the dengue vector prefers: receptacles or other other community members to take preventa- items with solid, flat surfaces that hold clean tive actions. This stems from a long-standing water in residential areas. Previous studies culture that community members expect their similarly reported participants failed to rec- neighbor will not take action and remove ognize mosquitos in their house as the dengue breeding sites on their property. For example, vector (Pérez-Guerra, Seda, García-Rivera, & in the town it is “normal” to litter, therefore Clark, 2005; Torres-López, Guerrero-Cordero, watching a neighbor throw trash on the street is & Salazar-Estrada, 2012). Reported social bar- acceptable. In order to change this norm, each riers include a lack of responsibility and a member has to not only take individual owner- muted concern of the disease (Pérez, Zielinski, ship, but collaborate with others. This attitude Vargas, & Clark, 2009), especially due to hazy of pointing the finger at one’s neighbors is not information (Torres-López et al., 2012). We unique to this community (Pérez-Guerra et al., believe the inability to differentiate and iden- 2005, Torres-López et al., 2012). But mem- tify practices between mosquitoes may lead to bers need to take ownership, challenge norms, this lack of responsibility and a muted concern, and even expand their personal responsibili- as well as the failure to mobilize in QG. ties. Some community members are physically The primary role communities play in unable to independently review their lawns reducing the risk for dengue is to remove adequately, such as the elderly. These people receptacles on their property (San Martin & need family members or neighbors to step in Prado, 2004), and the role of community par- for them. ticipation needs to be unified within a com- Respondents also suggested dengue edu- munity (Toledo-Romaní et al., 2006). A blurry cation needs to be frequent, continual and understanding of Aedes habits mutes the rel- non-stop, because community members are evancy and urgency for community members human and therefore tend to “forget” if infor- to play their part. By comprehending the dis- mation is not brought up often. This report tinctions of the dengue vector, community reflects one of the tenets of Communication members will understand the reason they must for Behavior Impact (COMBI) put out by review their lawns diligently, and unify on their the WHO: sustained appropriate advertising participatory role. which is massive, repetitive, intense and per- Regarding community views on dengue sistent (Parks & Lloyd, 2004), which has been responsibility, eight in ten (81%) respondents supported in other studies as well (Khun & said the community is at least partially respon- Manderson, 2007). sible for dengue prevention; however, about 20 Regarding MH vector control inspections, percent do not claim any responsibility. This is MH workers inspect properties to remove and a concern because a mindset of responsibility prevent Aedes breeding sites, and to educate and ownership leads citizens to act. Citizens citizens on dengue and its prevention. The MH

864 Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 859-867, September 2014 visits and inspections were of the most com- To avoid “preaching to the choir,” dengue mon suggestions on how to educate and even education needs to target the right audience. to convince community members on dengue According to our findings, we compiled the and its prevention. Other communities have following characteristics which suggest the similarly requested house visits (Zuleta et al., target audience should include people who: do 2011) to explain specific control measures not live with anyone previously infected by (Pérez-Guerra et al., 2005). The use of dialogue dengue, do not understand the symptoms of has been stressed in the transition from critical dengue, have more frequent infections in the understanding to critical action (Freire, 1973), household (relates to more possible breeding and a house visit provides a great opportunity sites), have a low education level (Rosenbaum for one-on-one dialogue with MH educators et al., 1995, Syed et al., 2010), rate the impor- and residents. tance of actions on dengue prevention lower House visits and family education require (which relates to more positive breeding sites), many trained vector control workers, and a rate dengue to be less dangerous (which relates lot of time. Additionally, some citizens report to more positive breeding sites), and have more that the MH has abandoned their position in possible breeding sites on their property (which Special section (ACM) Costa Rica program Costa Special section (ACM) dengue control because they fumigate much relates to more positive breeding sites). less than before. The request for fumigation has been reported in other communities as well ACKNOWLEDGMENTS (Pérez-Guerra et al., 2005). However, an over- dependence on fumigation or repeated house A special thank you goes to Chris Vaughan, visits diminishes the community’s ownership Michael McCoy, Mario Morera, Judith Mag- and perceived responsibility to review their nan, Jaime Rodríguez Arguedas, José Valerín own properties for breeding sites. There is a Cordero, Maynor Peraza Solas, Zulema Conejo misunderstanding that fumigation equals sup- Mora, Luis Guillermo Chaverri, Nick Bohrer, port in fighting dengue. This aim should be and above all, the community members of QG. openly communicated with community mem- bers so they do not get the incorrect impres- RESUMEN sion that the MH is retracting its position or support in dengue control by decreasing the Conocimientos, percepciones y prácticas para frequency of fumigation. prevenir el dengue en un pueblo costero del Pacífico central de Costa Rica. Entre las enfermedades virales Regarding target audience characteristics transmitidas por artrópodos, el dengue es la más mortal to consider, dengue perceptions have been y de difusión más rápida en el mundo. Es transmitida por shown to vary based on previous acquisition los mosquitos Aedes, y está presente en más de 125 países. (Pérez et al., 2009). We similarly found that En 2013, Costa Rica tuvo el mayor porcentaje de casos de dengue desde la reintroducción del virus al país en 1993. respondents who have had a dengue infection Este estudio evaluó el conocimiento de 320 miembros de in the household perceive dengue to be more la comunidad costera de Quebrada Ganado, Costa Rica, dangerous, and they consider the actions of quienes dieron sus opiniones acerca de cómo mejorar la prevention to be more important. Respondents educación y movilizar la comunidad en la prevención del who were familiar with dengue symptoms dengue durante marzo y abril 2013. Además, mediante inspecciones de propiedades se trató de hallar conexio- were also stronger supporters of preventa- nes entre el conocimiento u opiniones y la presencia de tive actions. This indicates if people are more criaderos de Aedes. Aunque los habitantes sabían que el aware of dengue symptoms, they may be more virus es transmitido por los mosquitos que se reproducen inclined to prevent the disease. This supports cerca del agua, no reconocían los hábitos del mosquito y lo confundían con otros mosquitos. El 81% acepta la respon- the educational goal to highlight the symptoms sabilidad por la prevención del dengue. Sugerencias para of dengue and risks it poses to an individual mejorar la educación incluyeron: información continua y (Pérez et al., 2009). frecuente sobre los riesgos, peligros y daños a la salud. Las

Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 859-867, September 2014 865 características relacionadas con las propiedades que tenían Miller, A. (2012). UN atlas links climate change to disea- más criaderos fueron: menor nivel de educación, mayor se spread. Canadian Medical Association Journal, frecuencia de infecciones con dengue, menor entendimien- 185(1), E38. to de los síntomas del dengue, menor calificación dada Nam, V. S., Yen, N. T., Phong, T. V., Ninh, T. U., Mai, L. a los peligros del dengue y menor calificación dada a la Q., Lo, L.V., & Kay, B. (2005). Elimination of dengue importancia de las acciones preventivas. Mientras se pone by community programs using mesocyclops (cope- en práctica una gama amplia de métodos de prevención pod) against Aedes aegypti in central Vietnam. The del dengue, la participación activa de la comunidad es American Journal of Tropical Medicine and Hygiene, sumamente importante para el éxito de estas actividades. 72(1), 67-73. La evaluación continua es necesaria para lograr cambios inmediatos y a largo plazo en el comportamiento social y Panagos, A., Lacy, E. R., Gubler, D. J., & Macpherson, C. para aprovechar mejor los recursos económicos invertidos N. (2005). Dengue in Grenada. Revista Panamerica- en la prevención del dengue. na de Salud Pública, 17(4), 225-229. Parks, W. & Lloyd, L. (2004). Planning social mobilization Palabras clave: dengue, Costa Rica, comunicación social, and communication for dengue fever prevention and Aedes aegypti. control: A step-by-step guide. Geneva, Switzerland: World Health Organization. REFERENCES Pérez, C. L., Zielinski, E., Vargas, D., & Clark, G. G. (2009). Community beliefs and practices about den- Bhatt, S., Gething, P. W., Brady, O. J., Messina, J. P., gue in Puerto Rico. Revista Panameña de Salud Farlow, A. W., Moyes & Hay, S. I. (2013). The Publica, 25, 218-226. global distribution and burden of dengue. Nature, Pérez-Guerra, C. L., Seda, H., García-Rivera, E. J., & 496(7446), 504-507. Clark, G. G. (2005). Knowledge and attitudes in de María Cáceres-Manrique, F., Vesga-Gómez, C., Perea- Puerto Rico concerning dengue prevention. Revista Florez, X., Ruitort, M., & Talbot, Y. (2009). Conoci- Panameña de Salud Pública, 17(4), 243-253. mientos, actitudes y prácticas sobre dengue en dos Rosenbaum, J., Natahan, M. B., Ragoonanansingh, R., barrios de Bucaramanga, Colombia. Revista de Salud Rawlins, S., Gayle, C., Chadee, D. D., & Lloyd, L. S. Pública, 11(1), 27-38. (1995). Community participation in dengue preven- Espinoza-Gómez, F., Hernandez-Suarez, C. M., & Coll- tion and control: A survey of knowledge, attitude and Cárdenas, R. (2002). Educational campaign versus practice in Trinidad and Tobago. American Journal malathion spraying for the control of Aedes aegypti of Tropical Medicine and Hygiene, 53(2), 111-117. in Colima, Mexico. Journal of Epidemiology and Ross, T. M. (2010). Dengue virus. Clinics in Laboratory Community Health, 56(2), 148-152. Medicine, 30(1), 149-160. Farrar, J., Focks, D., Gubler, D., Barrera, R., Guzman, M. San Martín, J. L. & Prado, M. (2004). Percepción del riesgo G., Simmons, C., & Kroeger, A. (2007). Editorial: Towards a global dengue research agenda. Tropical y estrategias de comunicación social sobre el dengue Medicine & International Health, 12(6), 695-699. en las Américas. Revista Panameña de Salud Públi- ca, 15(2), 135-139. Freire, P. (1973). Education for critical consciousness. New York, NY, USA: Seabury. Severo, O. P. (1955, September), Eradication of the Aedes aegypti mosquito from the Americas. Paper presented Gubler, D. J. (1998). Dengue and dengue hemorrhagic at “Yellow fever, a symposium in commemoration fever. Clinical Microbiology Review, 11(3), 480-496. of Carlos Juan Finlay” Paper 6 of Jefferson Medi- Guzman, M. & Kouri, G. (2002). Dengue: An update. The cal College, Philadelphia. Retrieved from http://jdc. Lancet Infectious Diseases 2(1), 33-42. jefferson.edu/yellow_fever_symposium/6 INEC (Instituto Nacional de Estadística y Censos) (2011). Sokal, R. R. & Rohlf, F. J. (1995). Biometry: The principles Estadística Censo 2011. Retrieved from Área de and practice of statistics in biological research. New Servicios de Información y Divulgación. San José, York, NY, USA: W. H. Freeman & Co. Costa Rica. Solís, M. I., Quesada, M., & Gutiérrez, I. (2010). Knowled- Khun, S. & Manderson, L. (2007). Community and school- based health education for dengue control in rural ge, attitudes and practices against the dengue fever Cambodia: A process evaluation. PLoS Neglected promoted by Costa Rican media. Revista Costarri- Tropical Diseases, 1(3), e143. cense de Salud Pública, 19, 88-92. Martín, J. L. & Prado, M. (2004). Risk perception and Soper, F. L. (1963). The elimination of urban yellow fever strategies for mass communication on dengue in in the Americas through the eradication of Aedes the Americas. Revista Panameña de Salud Pública, aegypti. American Journal of Public Health and the 15(2), 135-139. Nation’s Health, 53(1), 7-16.

866 Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 859-867, September 2014 Statpoint Technologies, Inc. (2011). Statgraphics Centurion Troyo, A., Porcelain, S. L., Calderón-Arguedas, O., Cha- [computer software]. Retrieved from http://www. dee, D. D., & Beier, J. C. (2006). Dengue in Costa statgraphics.com. Rica: The gap in local scientific research. Revista Syed, M., Saleem, T., Syeda, U., Habib, M., Zahid, R., Bas- Panameña de Salud Public, 20(5), 350-360. hir, A., & Saleem, S. (2010). Knowledge, attitudes Trpis, M., Hausermann, W., & Craig, G. B. (1995). Esti- and practices regarding dengue fever among adults mates of population size, dispersal, and longevity of of high and low socioeconomic groups. Journal of domestic Aedes aegypti aegypti (Diptera:Culicidae) Pakistan Medical Association, 60(3), 243-247. by mark-release-recapture in the village of Shauri Toledo-Romaní, M. E., Baly-Gil, A., Ceballos-Ursula, E., Moyo in eastern Kenya. Journal of Medical Entomo- Boelaert, M., & Van der Stuyft, P. (2006). Participa- logy, 32(1), 27-33. ción comunitaria en la prevención del dengue: Un WHO (World Health Organization). (2009). Dengue: abordaje desde la perspectiva de los diferentes actores Guidelines for diagnosis, treatment, prevention and sociales. Salud Pública de México, 48(1), 39-44. control-new edition. Geneva, Switzerland: WHO. Torres-López, T. M., Guerrero-Cordero, J. L., & Salazar- Zuleta, L. P., Garzón, A., Pérez, R., Rodríguez, E., Fonseca, Estrada, J. G. (2012). Dimensiones culturales del J., & Cano, F. (2011). Caracterización de conductas dengue que favorecen o dificultan su prevención en relacionadas con dengue, Yopal, Casanare, Colombia, México. Revista Panameña de Salud Pública, 31(3), 2010. Informe Quincenal Epidemiológico Nacional, 197-203. 16(17), 203-215. Special section (ACM) Costa Rica program Costa Special section (ACM)

Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 859-867, September 2014 867

Incidence and risk factors for cognitive impairment in rural elderly populations in Costa Rica

Jeffrey L. Nadel & Diana Ulate Associated Colleges of the Midwest, San Pedro Montes de Oca, Costa Rica; [email protected], [email protected]

Received 20-iii-2014. Corrected 20-iv-2014. Accepted 15-v-2014.

Abstract: Risk factors for the onset of cognitive impairment in Costa Rica are not well understood, despite a substantial elderly population stemming from a higher than average life expectancy for the western hemisphere.

To investigate the risk factors that predict the onset of cognitive impairment in the rural elderly of Costa Rica, Rica program Costa Special section (ACM) a modified version of the Mini Mental State Exam—designed for illiterate populations—was administered to 90 elderly inhabitants of San Carlos, Alajuela, Costa Rica between April and May of 2011. Subsequently, each participant took a structured interview assessing viability of risk factors and behaviors potentially contributing to a diagnosis of cognitive impairment. Results showed strong dependencies between age (p=0.0001), education level (p=0.0095), the ability to read (p=0.0001) and write (p=0.0153), frequency of reading (p=0.0011), use of puzzles and mind games (p<0.0001), vocation (p=0.0225), area of residence (p<0.0001), comorbid mental dis- eases (p=0.0005), history of stroke or brain trauma (p=0.0104), urinary or renal problems (p=0.0443), consistent cooking practices (p=0.0262) and number of living companions (p=0.0299) in susceptibility for developing cognitive impairment. The study concluded that high intellectual use, or lack thereof, during the lifetime of a person was a predictor for cognitive status later in life. In addition, comorbid mental disorders, including neu- rological trauma due to stroke, impeded normal cognitive function. Future research should examine incidence and risk factors of cognitive impairment in urban, more educated populations. Rev. Biol. Trop. 62 (3): 869-876. Epub 2014 September 01.

Key words: cognitive impairment, dementia, education, elderly, mini mental state exam, Costa Rica.

Although widely applicable to the Costa TR), and include memory loss, attention defi- Rican public health sector, risk factors for cog- cits, a loss of problem solving skills, mood and nitive impairment in Costa Rica are not well attitude changes, and trouble producing and understood. With a substantial elderly popula- understanding language (American Psychiatric tion stemming from a higher than average life Association, 2000). Dementia is a character- expectancy for the western hemisphere, and a istic symptom of several mental health dis- portion of this population living in rural areas orders also diagnosable through DSM-IV TR of the country, incapacitation attributable to criteria, including Alzheimer’s disease (AD). cognitive malfunction in the rural elderly is In fact, people diagnosed with Mild Cognitive prevalent (Cáceres-Rodas, 2004). Cognitive Impairment (MCI) are at significant risk for impairment is a key indicator for a diagnosis developing AD and other age-related neurode- of dementia, defined as a progressive, global generative disorders (Belleville, Sylvain-Roy, or local loss of proper mental function (Yen de Boysson, & Ménard, 2008; Li et al., 2011; et al., 2010). The primary symptoms associ- Petersen et al., 1999; Petersen et al., 2001). ated with dementia are characterized in the With the current rate at which globally Diagnostic and Statistical Manual of Mental applicable biomedical research is advancing, Disorders, 4th Edition, Text Revision (DSM-IV the average human life span is dramatically

Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 869-876, September 2014 869 increasing. Costa Rica is no exception. Accord- structured interview following the Mini Mental ing to the 1984 census, there were 158 144 State Exam; and c) estimate the correlations people over the age of 60 in Costa Rica, com- between risk factors and presence and severity prising approximately 6.5% of the population. of cognitive impairment. The 2000 census revealed an increase to 301 474 people over the age of 60, making up 7.9% MATERIALS AND METHODS of the Costa Rican population. Current projec- tions for the year 2015 place the estimated Study site: This study was carried out in elderly population at 557 076, which would two locations: Florencia, San Carlos, Alajuela, correlate to more than 11.1% of the population Costa Rica (10°22’0” N and 84°29’0” W) and (Cáceres-Rodas, 2004; Centro Centroamerica- Ciudad Quesada, San Carlos, Alajuela, Costa no de Población, 2011). Additionally, portions Rica (10°20’0” N and 84°26’0” W) (Centro of Costa Rica have been designated a Blue Centroamericano de Población, 2011). Zone, which are defined as special areas of the world with high numbers of living centenarians Study population: The study population (Buettner, 2010). With this rapidly aging popu- was anyone over the age of 60 as defined as lation, it is becoming increasingly important elderly for rural research by the World Health to assess the prevalence of cognitive impair- Organization (2011). All participants fulfilled ment, identify risk factors associated with such one or more of the following three criteria to impairment, and educate the public on ways to create a random sample: a) being part of a non- preserve their mental status as they age. residential, elderly support group in the above Numerous studies in other countries have districts; b) living in an elderly institution shown that age, gender, education level, cardio- providing around-the-clock medical and social vascular health, daily nutritional value, living assistance; c) in their home when interviewed situation, pollution, consumption of tobacco by the researcher in neighborhoods of Ciudad and alcohol, and other comorbid mental and Quesada and Florencia, San Carlos, Alajuela, non-mental diseases are significant risk factors Costa Rica. for a diagnosis of cognitive impairment (Yen et al., 2010; Heun & Hein, 2005; Lee, Hong, Interview criteria: Human participants Cheong, & Oh, 2009; Sánchez-Rodríguez, San- and treatment during the interview was in tiago, Arronte-Rosales, Vargas-Guadarrama, accordance with the International Ethical Mendoza-Núñez, 2006; van Exel et al., 2002). Guidelines for Biomedical Research Involving However, a unique climate, economy, culture, Human Subjects as set forth by the Council for location, and greater than average life expec- International Organizations of Medical Sci- tancy contribute to risk factors specific to Costa ences in collaboration with the World Health Rica. Cáceres-Rodas (2004) explored the rela- Organization in 2002. Prior to the interview, tionship between physical health and mental all participants signed a standard consent form health in the outer lying rural and quasi-urban explaining information about the purpose of the parts of the Central Valley of Costa Rica. This study, the rights of the participant, and the risks study, though, did not specifically investigate of involvement. The form was read to each behavioral, social, or cultural risk factors for participant orally to ensure that all participants cognitive impairment. fully understood the implications of their par- The objectives of this study were to: a) ticipation. Literate subjects signed their name individually assess elderly patients in Floren- and illiterate participants supplied their finger- cia and Ciudad Quesada for differences in the print to confirm their consent. incidence rate of cognitive impairment using Subsequently, a modified version of the the Mini Mental State Exam; b) assess poten- Mini Mental State Exam (MMSE) was admin- tial risk factors for cognitive impairment in a istered to determine whether or not there was

870 Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 869-876, September 2014 an incidence of cognitive impairment and if their own meals; l) living companions—partici- so, to what degree. This rural population had pant lives alone, lives with 1-3 others, or lives a high percentage of illiterate and uneducated with four or more other people; m) urinary elderly persons. As such, an adjusted version problems—participant suffers from urinary of the MMSE designed for uneducated and problems, or does not suffer from urinary prob- illiterate populations was utilized (Kabir & lems; n) tv watching time—participant watches Herlitz, 2000). less than one hour of television per day, 1-3hrs After administration of the MMSE, all of television per day, or more than 3 hours of patients underwent a structured interview sur- television per day; o) diabetes—participant vey aimed at assessing risk factors for cognitive suffers from either Type I or Type II Diabetes, impairment. Both of these examinations were or does not suffer from diabetes; p) living sib- administered orally in an interview format. lings—participant has three or fewer siblings still living, or has four or more living siblings; Assessment survey variables: Variables q) cholesterol—participant is currently suffer- examined in the risk-factor assessment sur- ing from high cholesterol, has suffered from vey included: a) mental exercises—participant high cholesterol in the past, or does not and has regularly completes puzzles, word searches, never suffered from high cholesterol; r) num- Rica program Costa Special section (ACM) crossword puzzles, plays cards, dominoes, or ber of children—participant’s has reared three other mental puzzles or exercises; b) location or fewer children, or has reared four or more of residence—participant lived only in a rural children; s) number of surgeries—participant region, in both a rural and urban region, or only has had 2 or more surgeries, or has had 1 or in an urban region during their life; c) ability fewer surgeries; t) death of a child—participant to read—a self report of whether the partici- has reared a child who has subsequently died, pant can read well, can barely read, or cannot or all of the children of the participant are still read at all; d) age—participant is 60-70yrs old, living; u) colon problems—participant suffers 71-80yrs old, or 81+yrs old; e) comorbid men- from any number of intestinal problems includ- tal issues—self report of whether the patient ing, but not limited to, colitis, or the participant has been diagnosed with mental or neurological does not suffer from colon problems; v) smoker conditions such as depressed continuously for or past smoker—participant currently smokes more than one month, issues with anxiety or or has smoked in the past, or the participant nervousness, or takes or has taken treatments has never smoked; w) family mental issues— for mental problems; f) frequency of reading— someone in the participants family has suffered participant reads everyday, reads 2-6d of the from mental disease, stroke, or brain trauma, or week, or reads less than one time per week; g) no one in the family has suffered from mental formal education—participant received no for- or neurologic problems; x) gender—participant mal education, completed 1-2 grade, completed is male or female; y) eats dinner—participant 3-4 grade, completed 5-6 grade, or received eats dinner regularly, or participant does not eat more than six years of formal education; h) dinner; z) pregnancy frequency—female only: stroke—participant has had a stroke, or another participant has been pregnant four or fewer form of vascular or cerebral trauma; i) ability times, or five or more times; aa) drinker or past to write—participant can write more than their drinker—participant currently or habitually name, or the participant can write only their consumed alcoholic beverages now or in the name, or the participant cannot write; j) voca- past, or has never consumed alcoholic bever- tion—participant was an agricultural worker, ages; ab) tv companions—participant watches was a home-maker, or worked in business, TV alone, or watches with other people; ac) commerce, or city-oriented jobs; k) cooked for heart attack—participant has had a heart attack, themselves—participant was responsible for or the participant has never had a heart attack; preparing their own meals, or does not prepare ad) hypertension—participant is currently

Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 869-876, September 2014 871 suffering from hypertension, or has suffered 48.3%. Of those impaired, 48.8% were mildly from hypertension in the past; ae) cardiovas- impaired, 48.8% were moderately impaired, cular—participant self reports that a doctor has and 2.3% were severely impaired. previously told them that they have heart prob- Results of bi-variate contingency table lems, or that they have never been informed analysis of independence between individual of heart problems; and af) gastric problems— risk factor variables and the incidence of cog- participant suffers from any number of gastric nitive impairment are summarized in table 1. problems including, but not limited to, gastritis, Specific elucidation of three statistically signif- or does not suffer from gastric problems. icant risk factor variables—formal education, The MMSEs were scored, and overall vocation, and age—can be seen in a categorical prevalence of cognitive impairment was esti- breakdown in Figs. 1-3, respectively. mated. Participants were considered cogni- There was a significant dependency of the tively impaired by scoring below 25 out of 30 selection group sampled on the incidence of possible points on the MMSE. A score of <9 cognitive impairment (X2= 8.9, df=3, p=0.030). points indicated severe cognitive impairment, a score of 10-20 points indicated moderate 18 Normal cognitive impairment, and a score of 20-24 16 Incapacitated indicated mild cognitive impairment. 14 This measured independence between 12 individual risk factors and incidence of cogni- 10 tive impairment through contingency tables. 8 Each severity category was not analyzed with 6 each variable. Only the overall incidence (all 4 severity categories combined) or lack of cog- Number of individuals 2 nitive impairment was measured against the 0 individual risk factor variables. Statistical tests No 1-2 grade 3-4 grade 5-6 grade 7+ grade followed Sokal & Rohlf (1995) and were run education with Statgraphics Centurion XVI.I software Educational Status (Statpoint Technologies, Inc., 2011). Fig. 1. Frequency of participants and their obtained educational level as it corresponds to mental state, April- May 2011, San Carlos, Costa Rica. RESULTS

25 This study included a total of 90 partici- Normal pants. One participant underwent administra- tion of a non-modified MMSE for illiterate 20 populations, was unable to be re-tested with an unbiased form of the exam, and was thus 15 excluded from analysis. Of those included, 60.6% were women. Mean age of participants 10 was 76.8+8.1 (Mean+S.D.) years; 33.7% were

Number of individuals 5 over the age of 80. In terms of educational status, 5.6% received more than a primary 0 education, with 41.6% of participants having Agricultural House work City work completed the second year of formal primary work education or less. The majority of the study Vocational Type population (96.6%) was born in Costa Rica. Fig. 2. Frequency of participants and their vocational type Overall incidence of cognitive impairment as it corresponds to mental state, April-May 2011, San (mild, moderate and severe combined) was Carlos, Costa Rica.

872 Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 869-876, September 2014 TABLE 1 Contingency table analysis of independence between individual risk factor variables and the incidence of cognitive impairment. San Carlos, Costa Rica. April-May 2011

Risk Factor Variables X2 (Chi-Squared) Degrees of Freedom p Mental Exercises 18.417 1 <0.0001* Location of Residence 20.509 2 <0.0001* Ability to Read 17.896 2 0.0001* Age 19.297 2 0.0001* Comorbid Mental Issues 12.105 1 0.0005* Frequency of Reading 13.555 2 0.0011* Formal Education 13.406 4 0.0095* Stroke 6.565 1 0.0104* Ability to Write 5.822 1 0.0153* Vocation 7.591 2 0.0225* Cooked for Themselves 4.945 1 0.0262* Living Companions 4.715 1 0.0299*

Urinary Problems 4.044 1 0.0443* Rica program Costa Special section (ACM) TV Watching Time 2.740 1 0.0979 Diabetes 1.927 1 0.1651 Living Siblings 1.857 1 0.1730 Cholesterol 1.829 1 0.1762 Number of Children 1.084 1 0.2979 Number of Surgeries 0.934 1 0.3337 Death of Child 0.757 1 0.3842 Colon Problems 0.685 1 0.4078 Smoker or Past Smoker 0.440 1 0.5072 Family Mental Issues 0.440 1 0.5072 Gender 0.435 1 0.5097 Eats Dinner 0.116 1 0.7332 Pregnancy Frequency 0.092 1 0.7613 Drinker or Past Drinker 0.054 1 0.8163 TV Companions 0.036 1 0.8499 Heart Attack 0.028 1 0.8675 Hypertension 0.020 1 0.8867 Cardiovascular 0.015 1 0.9012 Gastric Problems 0.014 1 0.9050

*Denotes statistical significance at the p<0.05 level.

Residents of the Hogar de Ancianos had the Higher levels of brain input and stimulation most impairment (77.8%), followed by Ciudad appeared to exhibit neuroprotective effects Quesada (42.3%), and then Florencia (41.7%). as the study population aged. Acquiring and/ The lowest proportion of impairment was noted or sustaining formal education by frequently in the group collected house to house (33.4%). reading, writing, and completing puzzles also appeared to aid in staving off brain atrophy. DISCUSSION Additionally, both a city-oriented vocation that requires enhanced daily thought processing and Nearly all of the significant risk factors a home in a suburban or city atmosphere where leading to increased impairment shared the more visual, auditory, and interpersonal stimu- common theme of lack of mental stimulation. lation exists had seemingly similar effects.

Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 869-876, September 2014 873 25 cognition and memory, among others. Further- Normal more, only about half of stroke patients regain 20 enough function to independently care for themselves (PubMed Health, 2011). Thus, it is 15 logical that stroke survivors tested significantly for increased cognitive impairment. 10 Regarding implications of dangerous behaviors in impairment, behaviors such as chronic smoking and drinking showed no sig-

Number of individuals 5 nificant effect on cognitive impairment in this study, debate still exists about the roles that 0 60-70 Y.O. 71-80 Y.O. 80+ Y.O. smoking and alcohol consumption play in Age general cognition and incapacity (Hill, 1989; Fig. 3. Frequency of individuals and their age as it Reitz, Luchsinger, Tang, & Mayeux, 2005; corresponds to mental state, April-May, 2011, San Carlos, Kalmijn, van Boxtel, Verschuren, Jolles, & Costa Rica. Launer, 2002). Gender played an insignificant role as an indicator for cognitive impairment in this study. Interestingly, Yen et al. (2010) reported females Numerous studies have found strong at higher risk for cognitive impairment. How- trends suggesting education level as a predictor ever, van Exel et al. (2001) showed women less for the onset of cognitive impairment, and that prone to the development of cognitive impair- participation in life-long intellectual activity ment. Obvious debate exists, prompting the can be an important factor in staving off such necessity for unbiased sampling. The present symptoms (Callahan et al., 1996; Kliegel, Zim- study is prone to a similar type of sample bias prich, & Rott, 2004; Marcopulos, McLain, & given the gender demographics of the sample Giuliano, 1997; White et al., 1994). Similarly, population. In this sample population, greater intellectual stimulation can be seen as a tech- than 60% of participants were female. The nique to attenuate the harsh effects of cognitive majority of male participants were interviewed impairment when it acts as a characteristic in the Hogar de Ancianos, a hospital-type insti- symptom of neurologic disorders such as mul- tution where more than 70% of residents inter- tiple sclerosis (Sumowski, Wylie, Chiaraval- viewed exhibited cognitive impairment. Thus, loti, & DeLuca, 2010). As such, it is becoming with a higher than representative sample of clearer that higher intake of life-long intellectu- impaired males, and a larger than 60% sample al stimuli is essential for resiliency in cognitive of females, it is possible that the sample bias status, regardless of age or comorbid disease. contributed to a non-significant role of gender Regarding implications of brain traumas in in cognitive impairment. impairment, not surprisingly, patients who had Additionally, the majority of health-orient- previously sustained brain trauma, whether in ed comorbidities in this study did not appear the form of stroke, or other, were significantly as significant indicators for a diagnosis of more impaired than their non-affected coun- cognitive impairment. Significant correlations terparts. A stroke is any interruption of blood were previously reported for diabetes and car- supply to the brain, either by blockage or burst- diovascular disease as predictors of cognitive ing (PubMed Health, 2011). The implications impairment (Yen et al., 2010; van Exel et al., of a stroke, according to the same source, can 2002). However, the study population assessed cause lasting brain damage and mal-effects in in Yen et al. (2010) was nearly 10 yrs younger speech production and comprehension, move- on average than in the present study. Further, ment, bladder and bowl problems, issues with van Exel et al. (2002) implemented the use of

874 Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 869-876, September 2014 electrocardiogram technology, and in-depth (p=0.0225), el lugar de residencia (p<0.0001), las enfer- analysis of patient charts. Each of these meth- medades mentales comórbidas (p=0.0005), un historial de derrame o de trauma cerebral (p=0.0104), los trastornos odological disparities can contribute to differ- urinarios o renales (p=0.0443), la preparación de alimentos ences in findings. en forma consistente (p=0.0262), el número de personas Finally, self-reported data often provides con quienes convive (p=0.0299) y la susceptibilidad de a skewed view of the medical history of the desarrollar la discapacidad cognitiva. El estudio concluyó patient. Patients may not understand exactly que un alto nivel de uso del intelecto, o la falta del mismo, durante la vida es un indicador del estatus cognitivo en what a doctor is saying during consultations, etapas más avanzadas de la vida. Además, se encontró que especially in a highly uneducated, rural popula- las enfermedades mentales comórbidas, como el trauma tion. Thus, readers must take into account the neurológico provocado por un derrame cerebral, impiden means by which the data was collected when la función cognitiva normal. Se recomienda que futuras generalizing about the apparent insignificance investigaciones examinen la incidencia y los factores de riesgo asociados con la discapacidad cognitiva en pobla- of existing health comorbidities. ciones urbanas de mayores niveles educativos.

Palabras clave: discapacidad cognitiva, demencia, edu- ACKNOWLEDGMENTS cación, adulto mayor, mini examen del estado mental,

Costa Rica. Rica program Costa Special section (ACM) We thank Chris Vaughan and Michael McCoy for their tireless efforts to make this project a success, help on statistical analysis, REFERENCES and manuscript edits. We also thank Olga American Psychiatric Association. (2000). Diagnostic and Mora, Albino Riveras, and Dayma Morales, Statistical Manual of Mental Disorders. (DSM-IV- directors and overseers of the elderly institu- TR). Arlington, VA, USA: American Psychiatric tions in San Carlos de Alajuela, Costa Rica. Association. And finally, thanks are given to all study par- Belleville, S., Sylvain-Roy, S., de Boysson, C., & Ménard, ticipants who selflessly shared their stories, M. C. (2008). Characterizing the memory changes in persons with mild cognitive impairment. Progress in insights, fears, and lives during this experience. Brain Research, 169, 365-375. Buettner, D. (2010). The blue zones: Lessons for living lon- RESUMEN ger from the people who’ve lived the longest. Wash- ington, D. C., USA: National Geographic Books. Incidencia y factores de riesgo para la discapaci- Cáceres-Rodas, R. I. (2004). Estado de salud físico y dad cognitiva en poblaciones rurales de tercera edad mental de los adultos mayores del área rural de en Costa Rica. Los factores de riesgo asociados con el Costa Rica. Población y salud en Mesoamérica, 1(2). inicio de la discapacidad cognitiva en Costa Rica son poco Retrived from http://revistas.ucr.ac.cr/index.php/psm/ comprendidos, a pesar de la existencia de una considerable article/view/13931 población de la tercera edad que ha resultado de una expec- Callahan, C. M., Hall, K. S., Hui, S. L., Musick, B. S., tativa de vida mayor que el promedio en el hemisferio occi- Unverzagt, F. W., & Hendrie, H. C. (1996). Relation- dental. Para investigar los factores de riesgo que indican el ship of age, education, and occupation with dementia inicio de la discapacidad cognitiva en adultos mayores de among a community-based sample of African ameri- zonas rurales en Costa Rica, se les administró una versión cans. Archives of Neurology, 53(2), 134-140. modificada del Mini Examen del Estado Mental diseñado Centro Centroamericano de Población. (2011). Estima- para poblaciones analfabetas a 90 adultos mayores resi- ciones y Proyecciones de Población por sexo y dentes en San Carlos de Alajuela, Costa Rica, entre abril edad 1950-2100 (cifras actualizadas). Universidad y mayo 2011. Posteriormente cada participante recibió una de Costa Rica. Retrieved from http://censos.ccp.ucr. entrevista estructurada para evaluar la viabilidad de los fac- ac.cr /cgi-bin/login.pl tores de riesgo y comportamientos que podrían contribuir a un diagnóstico de discapacidad cognitiva. Los resultados Heun, R. & Hein, S. (2005). Risk factors of major depres- mostraron una fuerte relación de dependencia entre la edad sion in the elderly. European Psychiatry, 20(3), (p=0.0001), el nivel educativo (p=0.0095), la capacidad 199-204. de leer (p=0.0001) y escribir (p=0.0153), la frecuencia Hill, R. D. (1989). Residual effects of cigarette smoking on de lectura (p=0.0011), la utilización de rompecabezas y cognitive performance in normal aging. Psychology juegos de estímulo intelectual (p<0.0001), la vocación and Aging, 4(2), 251-254.

Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 869-876, September 2014 875 Kabir, Z. N. & Herlitz, A. (2000). The Bangla Adaptation Sánchez-Rodríguez, M. A., Santiago, E., Arronte-Rosales, of Mini-mental State Examination (BAMSE): An A., Vargas-Guadarrama, L. A., & Mendoza-Núñez, V. instrument to assess cognitive function in illiterate M. (2006). Relationship between oxidative stress and and literate individuals. International Journal of cognitive impairment in the elderly of rural vs. urban Geriatric Psychiatry, 15(5), 441-450. communities. Life Sciences, 78(15), 1682-1687. Kalmijn, S., van Boxtel, M. P. J., Verschuren, M. W. M., Sokal, R. R. & Rohlf, F. J. (1995). Biometry: The principles Jolles, J., & Launer, L. J. (2002). Cigarette smoking and practice of statistics in biological research. New and alcohol consumption in relation to cognitive York, USA: W. H. Freeman and Co. performance in middle age. American Journal of Epidemiology, 156(10), 936-944. Statpoint Technologies, Inc. (2011). Statgraphics Centurion [computer software]. Available from http://www. Kliegel, M., Zimprich, D., & Rott, C. (2004). Life-long statgraphics.com. intellectual activities mediate the predictive effect of early education on cognitive impairment in cen- Sumowski, J. F., Wylie, G. R., Chiaravalloti, N., & DeLu- tenarians: A retrospective study. Aging and Mental ca, J. (2010). Intellectual enrichment lessens the Health, 8(5), 430-437. effect of brain atrophy on learning and memory in Lee, K. S., Hong, C. H., Cheong, H. K., & Oh, B. H. multiple sclerosis. Neurology, 74(24), 1942-1945. (2009). Difference in nutritional risk between mild van Exel, E., Gussekloo, J., de Craen, A. J. M., Bootsma- cognitive impairment group and normal cognitive van der Wiel, A., Houx, P., Knook, D. L., & Westen- function elderly group. Archives of Gerontology and dorp, R. G. J. (2001). Cognitive function in the oldest Geriatrics, 49(1), 49-53. old: Women perform better than men. Journal of Li, H., Li, J., Li, N., Li, B., Wang, P., & Zhou, T. (2011). Neurology, Neurosurgery & Psychiatry, 71(1), 29-32. Cognitive intervention for persons with mild cogni- van Exel, E., Gussekloo, J., Houx, P., de Craen, A. J. M., tive impairment: A meta-analysis. Ageing Research Macfarlane, P. W., Bootsma-van der Wiel, A., Blauw, Reviews, 10(2), 285-296. G. J., & Westendorp, R. G. (2002). Atherosclerosis Marcopulos, B., McLain, C. A., & Giuliano, A. J. (1997). and cognitive impairment are linked in the elderly. Cognitive impairment or inadequate norms? A study The Leiden 85-plus study. Atherosclerosis, 165(2), of healthy, rural, older adults with limited education. 353-359. The Clinical Neuropsychologist, 11(2), 111-131. White, L., Katzman, R., Losonczy, K., Salive, M., Wallace, Petersen, R. C., Doody, R., Kurz, A., Mohs, R. C., Morris, R., Berkman, L., Taylor, J., Fillenbaum, G., & Havlik, J. C., Rabins, P. V., Ritchie, K., Rossor, M., Thal, R. (1994). Association of education with incidence of L., & Winblad, B. (2001). Current concepts in mild cognitive impairment. Archives of Neurology, 58(12), cognitive impairment in three established populations 1985-1992. for epidemiologic studies of the elderly. Journal of Clinical Epidemiology, 47(4), 363-374. Petersen, R. C., Smith, G. E., Waring, S. C., Ivnik, R. J., Tangalos, E. G., & Kokmen, E. (1999). Mild cogni- World Health Organization. (2011). Definition of an older tive impairment: Clinical characterization and outco- or elderly person. Retrieved from http://www.who. me. Archives of Neurology, 56(3), 303-308. int/healthinfo/survey/ageingdefnolder/en/index.htm PubMed Health. (2011). Stroke. Retrieved from http:// Yen, C. H., Yeh, C. J., Wang, C. C., Liao, W. C., Chen, S. www.ncbi.nlm.nih.gov/pubmedhealth/PMH0001740/ C., Chen, C. C, Liang, J., Lai, T. J., Lin, S. H., & Lee, Reitz, C., Luchsinger, J., Tang, M. X., & Mayeux, R. M. C. (2010). Determinants of cognitive impairment (2005). Effect of smoking and time on cognitive over time among the elderly in Taiwan: Results of the function in the elderly without dementia. Neurology, national longitudinal study. Archives of Gerontology 65(6), 870-875. and Geriatrics, 50(suppl 1), S53-S57.

876 Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 869-876, September 2014 Knowledge of cervical cancer pathology of high school students in San Carlos, Costa Rica

Lauren M. Carlson1 & Sam Gonzalez2 1. St. Olaf College, Northfield, MN, USA; [email protected] 2. Macalester College, St. Paul, MN, USA; [email protected]

Received 20-iii-2014. Corrected 20-iv-2014. Accepted 26-v-2014.

Abstract: In Costa Rica, cervical cancer is the second most common female cancer, despite accessibility of screening through the universal health care system. There is little understanding of knowledge levels of the

adolescent population on cervical cancer, although this population is the most exposed to risk factors. This study Rica program Costa Special section (ACM) sought to learn about male and female adolescent knowledge of preventative methods, infection acquisition, clinical manifestations and medical services. A total of 533 students from grades 7-12 from five public high schools in the rural San Carlos region participated in the study from March through April 2012. Students were found to lack knowledge, as only 30.8% of students stated that they knew what cervical cancer is. Additionally, a connection was lacking between cervical cancer concepts, as for example 75.2% of students had heard of Human Papiloma Virus, but only 33.9% of those students knew that HPV is related to cervical cancer. Age had a positive relationship with knowledge of main concepts (p≤0.001). More women than men had heard of cervical cancer and the Papanicolaou (p=0.025, p≤0.0001), but otherwise no significant difference in mean response between genders was found. Students were found to have a limited awareness of the educational cervical cancer campaign (7.7% of the student population) and the ability to go alone to the doctor to receive medical attention (30.6% of students). Additionally, the public education system does not require cervical cancer to be included in the present curriculum. Hence, as students lack education on prevention and risk factors, the majority of prevention responsibility falls only on the universal health care system to regularly perform Papanicolaou exams to detect pre-cancerous or cancerous changes. The findings indicated the importance of including cervical cancer in the sexual education curriculum of the public education system in Costa Rica to educate the at-risk population of preventative methods, infection development, clinical manifestations and medical services in order to lower the presence of cervical cancer. Rev. Biol. Trop. 62 (3): 877-886. Epub 2014 September 01.

Key words: adolescents, cervical cancer, Costa Rica, Papanicolaou, sex education.

In order to prevent cervical cancer one Goodman, 2011). About 90% of the infections should limit exposure to Human Papilloma naturally clear, yet to avoid HPV exposure or Virus, especially by avoiding the main risk to receive a HPV vaccination is highly recom- factors: an early onset of sexual activity and an mended (Maine, Hurlburt, & Greeson, 2011). elevated number of sexual partners (American The Papanicolaou (Pap smear) is the most Cancer Society, 2011). Additional risk factors common screening test, or method to detect include sexually transmitted diseases, smoking, treatable pre-cancerous formations (Feldman obesity, contraceptive medications and unpro- et al., 2011). Screening is recommended every tected sexual intercourse (American Cancer other year for sexually active women at least Society, 2011). 20yrs old (Feldman et al., 2011). Cervical cancer develops due to grad- Symptoms of cervical cancer, particularly ual changes in the cervix after persistent, abnormal vaginal bleeding and pain during repeated HPV infections (Feldman, Sirovich, & sexual intercourse, generally arise in the latest

Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 877-886, September 2014 877 stages of the disease (American Cancer Soci- The “Caja Costarricense de Seguro Social” ety, 2011). Development from localized to (CCSS), the universal health care system of metastasized cancer typically takes 10-15yrs Costa Rica, works to prevent cervical cancer (RHO Cervical Cancer, 2009). The survival through education and conducting the Papani- rate for localized cancer is 90%, yet only colaou (A. Muñoz, personal communication, 16.5% for metastasized cancer (American Can- April 16, 2012). One week during each Octo- cer Society, 2011). This emphasizes the impor- ber or November is dedicated to the “Semana tance of detecting cervical cancer early through de Salud” (“Health Week”). During this time, the Papanicolaou before symptoms arise. stands are placed outside the EBAIS, the Cervical cancer is the second most common local public clinics, each with different health cancer facing Costa Rican women (Aguado, themes. One stand is dedicated to cervical can- Beauvais, Byrne, & Gacic-Dobo, 2010). Every cer and is geared towards teenagers, as they are year, an estimated 403 women are positively the most vulnerable population since they are diagnosed and 158 die due to cervical cancer, less likely to use condoms. They also work to in a population of over 2 million (Aguado et al., regularly screen the female population. 2010; Instituto Nacional de Estadística y Censo Each adolescent that comes to the clinic [INEC], 2011). Screening methods are avail- receives a “risk classification.” When a patient able for this preventable disease through Costa indicates they are sexually active, the doctor Rica’s universal health care system (Kivistik, speaks with the patient without their par- Lang, Baili, Antitila, & Veerus, 2011). How- ents. In addition to receiving a discussion ever, only an estimated 44.8% of Costa Rican about reducing risk factors and receiving free women are yearly screened and 17% of women contraceptives, females are advised to sched- over 65yrs-old have never received a Pap ule a Pap smear for another day when they smear (Aguado et al., 2010). Meanwhile, 52% can come alone. of the sexually active Costa Rican population The objectives of this study were to mea- faced the onset of sexual activity between 14 sure: a) the level of knowledge of the adolescent and 17yrs of age, putting them at a greater risk population concerning: preventative measures, of contracting cervical cancer (Picado, 2007). infection development, metastasis and avail- Few studies have investigated the level of able medical services; b) the level of knowl- cervical cancer knowledge of males, teenag- edge related to age, gender and school type; c) ers or Latinos. A South African study focused other potential barriers inhibiting women from on the knowledge of males of cervical cancer, receiving screening tests. found an extremely limited level of knowl- edge towards the disease and risk factors MATERIALS AND METHODS (Maree, Wright, & Makua, 2011). In another study, Nigerian women did not utilize screen- Study area: The study took place in ing exams, as they lacked knowledge of the Costa Rica in the province of Alajuela, within disease and screening accessibility (Nwankwo, the towns of Aguas Zarcas, Río Cuarto, La Aniebue, Aquwa, Anarado, & Aqunwah, 2011). Palmera and Venecia. Aguas Zarcas (10°20′N - Another South African study on the female 84°26′W) and Venecia (10°22’0” N - 84°17’0” population found basic preventative education W) are located 15km and 28km, respectively, should be given, in a manner appropriate to from Ciudad Quesada, the largest city of the the cultural context (Francis et al., 2011). In San Carlos region (Souza, 2011). Aguas Zar- socially-deprived areas of the United States cas is 97km North of San José, the capital of acceptance of the Papanicolaou exam is lacking Costa Rica (Souza, 2011). Aguas Zarcas has and these women not only were found to have an estimated population of 13 651 residents a limited knowledge, but negative attitudes (INEC, 2011). There are three high schools in towards screening (Logan & Mcifatrick, 2011). the Aguas Zarcas district, including one in La

878 Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 877-886, September 2014 Palmera. Venecia has an estimated population campaign in their community or not, and to of 7 394 residents (INEC, 2011). There are four assess the level of its presence. We also includ- high schools within the Venecia district, includ- ed questions to measure the knowledge of ing one school in Río Cuarto. teenagers about the medical services available, such as the Papanicolaou and whether unac- Study population: The study included companied doctor visits were possible. The public high school students from five schools final question asked students to identify poten- within the Venecia and Aguas Zarcas school tial barriers to going alone to the doctor. districts. Only students from grades 7-12 who consented to research were eligible to partici- Receiving approval: Before beginning pate. High school sizes ranged from approxi- research, we received approval from the St. mately 200 to 1 200 students (Table 1). A total Olaf Institutional Review Board. Then we of 533 students participated. The technical high received approval from the: a) regional direc- schools ranged from grades 7-12, while aca- tor, Lic. Oscar Oviedo Valerio, at the Ministry demic high schools ranged from grades 7-11. of Education in Ciudad Quesada; b) district supervisor of Venecia, Rodolfo González, to Procedure/Creating questionnaire: First, receive permission to work within the Vene- Rica program Costa Special section (ACM) we created a questionnaire to analyze the level cia school district; c) Rodolfo Salas, the dis- of knowledge of cervical cancer of male and trict supervisor of Agua Zarcas, and d) high female adolescents. To test for a basic level of school principals. understanding, we included initial questions pertaining to cervix identification, mortality Initial field research: We randomly chose and symptoms, as well as stating whether or classes to start distributing parental consent not they know what cervical cancer is. Addi- forms. We gave each class a brief explanation tionally, we added a series of questions about of the study, along with instructions to have risk factors, HPV and infection acquisition, to their parent/guardian sign the consent form that measure the understanding of students of cervi- night. We recorded each class section visited cal cancer prevention. and their class schedule for the next day. The following day, we returned to the same class We asked students to state whether they sections to distribute a student consent form had heard of an educational cervical cancer to minors that returned the signed parental consent form and to those at least 18yrs old. TABLE 1 We then administered the questionnaire to the Age and total number of students participating in the students with both completed forms. We trans- study for each age sampled. San Carlos, Costa Rica. ferred the questionnaire results onto an elec- March-April 2012 tronic data sheet. We returned a third day to the Age (years) Number of Participants same class sections to distribute questionnaires 12 44 to any additional students who had brought the 13 82 signed parental consent form. We repeated this 14 73 process with every class section possible. 15 76 16 58 For most questions we used a condensed 17 56 Likert scale with 0 and 1; 0= an incorrect or 18 102 negative response and 1= a correct or posi- 19 26 tive response. We constructed frequency his- 20 9 tograms of median answers with Microsoft 21 3 Excel. Relationships between several depen- 22 1 dent variables from questionnaire responses

Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 877-886, September 2014 879 with age of respondent (independent variable) 12-22yrs old and from grades 7-12 (Table 1). were estimated with simple linear parametric A total of 315 respondents were identified as regression. Dependencies of responses between females, and 216 identified as males. pairs of certain questions were estimated with The individual return rates of each school contingency tables. Differences between mean varied from 13.8% to 39.5%. The overall return responses of questions grouped by several par- rate was 27.5%. The technical schools had an ticipant characteristics were carried out with overall return rate of 14.3%, while the academ- one-way, parametric ANOVA. Homogeneity of ic schools had an overall return rate of 32.0%, variance was tested before applying paramet- excluding the Colegio Técnico de Aguas Zar- ric ANOVA. Statistical procedures followed cas, as not all of the classes were reached. Sokal and Rohlf (1995). Statistical tests were run with Statgraphics Centurion XVI (Statpoint Preventative measures: Only 7.7% of Technologies, Inc., 2011) students had heard of the presence of an educa- tional cervical cancer campaign in their com- Return to the field: Following data analy- munity. A total of 30.8% of students stated they sis, we returned to San Carlos to discuss the knew what cervical cancer is, with the mean findings and leave copies of the final paper response for females ( =0.35) greater than with the medical doctor of EBAIS 2 of Vene- for males ( =0.25) (F=5.04, d.f.=1, p=0.025; cia, the regional director in Ciudad Quesada, n=529). We found that 41.5% of the students as well as the district supervisors, principals knew cervical cancer is preventable. In iden- and interested teachers of each high school. We tifying the two key risk factors, 30.8% of stu- additionally left the answers to the question- dents identified an early onset of sexual activity naires at each high school with enough cop- and 37.1% identified an elevated number of ies for each teacher and class section, so the sexual partners. A total of 92.1% of students students could receive the correct information. identified at least one correct risk factor. A total of 67.9% of students had heard of RESULTS the Papanicolaou (Table 2). There was a strong positive relationship between age and likeli- Participants: A total of 533 students hood of hearing of the PAP (F=27.1, d.f.=1, completed the questionnaire from the five p≤0.0001, r2=4.88%; n=528, Table 3, Fig. 1). high schools studied. Participants ranged from Additionally, mean female response ( =0.73)

Fig. 1. Relationship between student age (12-22 years old) and response when asked, “Have you heard of the Papanicolaou (PAP)?” Values of “0” and “1” represent responses of “no” and “yes”, respectively. Equation is: “Heard of PAP?”=- 0.0544+0.0472*Age. Concave lines are 95% confidence intervals of the estimated line. San Carlos, Costa Rica. March- April 2012.

880 Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 877-886, September 2014 TABLE 2 Questionnaire responses and the percentage of students who gave each response. San Carlos, Costa Rica. March-April 2012

Response Percentage of Responses (%) Prevention & Risk Factors Had heard of an educational cervical cancer campaign 7.7 Stated cervical cancer is preventable 41.5 Identified an early onset of sexual activity as a risk factor 30.8 Identified a high number of sexual partners as a risk factor 37.1 Had heard of the Papanicolaou 67.9 Infection Had heard of the Human Papilloma Virus 75.2 Stated cervical cancer is acquired only through an infection 72.2 Had heard of HPV and stated it has an association with cervical cancer 33.9 Metastasis Identified vaginal bleeding as a symptom 69.4

Identified pain during sexual intercourse as a symptom 48.2 Rica program Costa Special section (ACM) Stated cervical cancer is fatal 63.8 Additional Information Correctly identified the cervix 39.4 Contributory Facts Stated adolescents can go to the doctor alone 30.6 Stated they would go to the doctor alone 51.0 was greater than male response ( =0.60) There was also a positive relationship between (F=9.37, d.f.=1, p=0.0023; n=528). There also age and likelihood of answering a correct was a positive dependency between know- symptom (F=5.18, d.f.=1, p=0.0232, r2=0.97%; ing cervical cancer is preventable and having n=528). heard of the Papanicolaou (X2=9.22; d.f.=1; p=0.0024). Additional information: Overall, 39.4% of students correctly identified the cervix from Infection development: We found 75.2% an image of the female reproductive system of the students had heard of Human Papilloma (Table 2). There was a strong positive relation- Virus (Table 2). A strong positive relationship ship between age and responding correctly existed between age and likelihood of hearing (F=28.1, d.f.=1, p<0.0001, r2=5.06; n=528). of HPV (F=61.9, d.f.=1, p<0.0001, r2=10.5%; n=528, Fig. 2). There was a positive depen- Medical services: We found 30.6% of dency between having heard or not about cervical cancer, with having heard or not about students knew adolescents can go alone to the HPV (X2=5.267; d.f.=1; p=0.0217). Yet, of the clinic for medical services (Table 2). Yet, 51% students who stated they knew what cervical of students responded they would go alone to cancer is and had heard of HPV, only 40.3% the clinic (Table 2). There was a strong posi- stated HPV has a relation to cervical cancer. tive relationship between age and the willing- ness to go alone (F=74.8, d.f.=1, p≤0.0001, Clinical manifestations: In recognizing r2=12.53; n=522, Fig. 3). Additionally, mean cervical cancer symptoms, 69.4% of students male response ( =0.59) was greater than mean identified vaginal bleeding and 48.2% identi- female response ( =0.45) (F=10.96, d.f.=1, fied pain during sexual intercourse (Table 2). p=0.0010; n=529).

Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 877-886, September 2014 881 TABLE 3 Question, F-Value, Degrees of Freedom, P-Value and R2 of each significant regression analysis. San Carlos, Costa Rica. March-April 2012

Question F-Value df n P-Value R2 (%) Preventative Measures Have you heard of the Papanicolaou (PAP)? 27.10 1 528 <0.0001 4.88 Infection Development Have you heard of HPV? 61.88 1 528 <0.0001 10.49 Metastasis Identify the symptoms of cervical cancer 5.18 1 528 =0.0232 0.97 Additional Information Identify the cervix 28.14 1 528 <0.0001 5.06 Contributory Facts Can an adolescent go without their parent/guardian to the 32.36 1 528 <0.0001 5.78 clinic for medical attention? Would you go to the clinic without your parent/guardian 74.78 1 522 <0.0001 12.53 for medical attention?

Fig. 2. Relationship between student age (12-22yrs old) and response when asked, “Have you heard of the Human Papilloma Virus?” Values of “0” and “1” represent responses of “no” and “yes”, respectively. Equation is: “Heard of HPV?”=- 0.2353+0.0640*Age. Concave lines are 95% confidence intervals of the estimated line. San Carlos, Costa Rica. March-April 2012.

The most common reasoning for unwill- DISCUSSION ingness to go alone to the doctor consisted of “fear of parental punishment”, “embarrassment A disconnection existed between the knowledge of adolescents with various cervi- of sharing personal information with parents” cal cancer concepts. For example, nearly all and “fear of the doctor”, while the two least of the students identified at least one risk fac- common reasons were “embarrassment with tor, but less than half stated cervical cancer is friends” and “embarrassment with boyfriend/ preventable. Hence, the majority of students girlfriend”. Mean female response ( =0.30) did not understand that avoiding risk factors was greater than male response ( =0.13) for prevents the acquisition of the disease. Addi- choosing “fear of the doctor” (F=20.1, d.f.=1, tionally, only less than half of the students who p≤0.0001; n=529). knew what cervical cancer is and had heard

882 Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 877-886, September 2014 Fig. 3. Relationship between student age (12-22 years old) and response when asked if they personally would be willing to go to the clinic alone to receive medical attention. Values of “0” and “1” represent responses of “no” and “yes”, respectively. Equation is: “Would Go Alone to Clinic?”=-0.7207+0.0791*Age. Concave lines are 95% confidence intervals of the estimated line. San Carlos, Costa Rica. March-April 2012. Special section (ACM) Costa Rica program Costa Special section (ACM) of HPV, also stated that HPV has a relation to of cervical cancer and medical services, specif- cervical cancer. ically: identifying the cervix, identifying symp- The questionnaire included two open- toms, hearing of the Papanicolaou, hearing of ended questions to allow for students to dem- HPV and knowing adolescents can go alone onstrate their level of knowledge of cervical to the doctor. A study in Nigeria found only cancer and the Papanicolaou. While less than 15% of participants knew of the availability of one-third of students stated they knew what cervical cancer screening, which had a positive cervical cancer is, when asked to define the correlation with the level of education attain- disease, the majority of responses consisted of ment (Nwankwo et al., 2011). The number of “cancer of the cervix” and possibly included students in our study that knew they could go to information describing it as sexually trans- the doctor alone, which had a positive correla- mitted or a female disease. This showed the tion with the level of education attainment, was students were lacking a depth of knowledge. nearly double the amount in the Nigerian study. Only three students were able to connect the As students aged, they were more likely to Papanicolaou with cervical cancer and only be willing to go alone to the doctor. With age, five students mentioned prevention. This dem- the students likely feel more independent. Also onstrated a lack of understanding between con- as they aged, the reasoning of students for not necting cervical cancer topics. going to the clinic alone changed from “fear of According to a study conducted in the parental discipline” to “embarrassment of shar- United Kingdom, women living in socially- ing personal information with their parents”. deprived areas also had a limited knowledge of The older the student, the more likely they cervical cancer risks and prevention strategies were to have been sexually active and, there- (Logan & Mcifatrick, 2011). Researchers of a fore, could be associated with a reason to feel similar study in South Africa, where cervical embarrassed (Picado, 2007). cancer is also the second-leading cause of death Women were more likely to state they among women, stated prevention methods and knew what cervical cancer is or have heard of materials are necessary to inform women about the Papanicolaou. Yet, there was no significant HPV, cervical cancer, screening and discussing correlation between gender and knowledge the topic with children (Francis et al., 2011). of HPV or cervical cancer risk factors. A With a few principal questions, a positive study conducted on South African men found correlation existed between age and knowledge men lack knowledge of cervical cancer, as no

Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 877-886, September 2014 883 participants mentioned the importance of HPV school or the clinic, so the majority likely and only a few knew having multiple sexual isn’t receiving the key preventative measure, partners was a risk factor (Maree et al., 2011). which is education. In contrast, about three-fourths of both Costa Hence, the majority of the responsibility Rican male and female students stated they had of cervical cancer prevention falls upon CCSS heard of HPV, with no significant difference intervention after the onset of sexual activity between genders. Maree et al. (2011) stated and potential exposure to HPV, by providing men need to be informed of cervical cancer Papanicolaou exams in order to detect any as they can help protect women from risk fac- pre-cancerous or cancerous lesions. There are tors and encourage screening. Although Costa also potential barriers preventing women from Rican men had stated they had heard of HPV, receiving routine Papanicolaou, as only 44.8% few could relate the virus to cervical cancer, of Costa Rican women are yearly screened showing an overall lack of connection between (Aguado et al., 2010). A study conducted in the these concepts. United States found distrust in the healthcare Although Costa Rican females were slight- system was negatively correlated with utiliza- ly more informed of cervical cancer and the tion of cervical cancer screening (Tse-Chuan, Papanicolaou, males were more likely to be Matthews, & Hillemeier, 2011). Although we willing to go alone to visit the doctor. At the did not measure health care system distrust, same time, females were more likely to be nearly half of students with reasoning for not deterred by their “fear of the doctor”. Hence, going alone to the clinic chose, “fear of the males could help support females in receiving doctor”. Hence, this could prevent utilizing the screening exam. the accessible screening exam. Therefore, it is Regarding responsibility for cervical can- even more crucial to educate the population on cer prevention, in order to prevent cervical risk factors. cancer, one can limit exposure to HPV infec- As regards intervention strategies, this tion through limiting risk factors and routinely study showed students were lacking informa- receiving a Papanicolaou in order to prevent tion not only on cervical cancer. High school pre-cancerous infections from forming cancer (Feldman et al., 2011). Ideally, one would students would benefit from a specific sexual receive a HPV vaccination, but this is not education course. Otherwise, cervical cancer covered through the CCSS, which means it could be added to the already required curricu- is unrealistic for much of the population to lum pertaining to sexually transmitted diseases. receive. (A. Muñoz, personal Communication, Currently health discussions are given 2-3 April 16, 2012). In order to limit risk factors, times a year by the CCSS to the student popula- education of risk factors needs to reach the tion, but this is dependent on funding. Hence, population before they become sexually active, we stress the importance of keeping these as this exposes them to HPV (American Cancer health discussions a priority of the community. Society, 2011). Less than half of the students who com- Unfortunately, the current, formal, required pleted the questionnaires were willing to go to curriculum by the MEP does not require infor- the doctor alone, nearly one quarter of these mation on cervical cancer to be included. students because of “embarrassment of shar- Fewer than one-third of students both know ing personal information with their parents”. they can go alone and are willing to go alone. Hence, it is also possible that students accom- This shows few students are going alone to the panied by their parents would avoid correctly clinic and, therefore, are probably less likely filling out the risk classification sheet in the to receive information on sexual education case they are fearful of the response of their from doctors. Therefore, few students receive parents. We think it would be ideal for the information from educational campaigns, CCSS to offer yearly physicals for students at

884 Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 877-886, September 2014 the high schools so students would feel more types are most effective, in order to ensure comfortable going alone to visit the doctor. teenagers are receiving the information in an The limitations or obstacles we faced appropriate manner. included a low return rate of parental consent forms, from forgetfulness or unwillingness of ACKNOWLEDGMENTS students and/or parents. When there were only a few students of a classroom with a signed We would like to thank our advisors, Diana permission slip, it was more likely they would and Freddy Ulate, who helped us design and choose not to participate. In the Colegio Téc- carry out a realistic study that is beneficial nico de Venecia, the principal distributed a to Costa Rica. Additionally, we would like to consent form, containing incorrect information, acknowledge the support of ACM staff, espe- to students before we went to the high school. cially Michael McCoy and Chris Vaughan, who Therefore, students with these consent forms helped us along every step of the research pro- could not participate either. cess. Thanks also go to the student participants, In a few classes, teachers would discuss as without their willingness and initiative we cervical cancer after their students asked ques- would not have a study. The warm welcome Rica program Costa Special section (ACM) tions about the study. Students also discussed and above and beyond patience of the superin- the questions between themselves, during and tendent, directors, principals, teachers and sec- after taking the questionnaire. This could dis- retaries within the Ministry of Public Education tort the data, as the discussions were only were utterly appreciated. prompted by our presence. With regard to future research, this study created new questions concerning the bar- RESUMEN riers of cervical cancer screening. As in the Conocimiento sobre cáncer de cérvix entre cole- one EBAIS we visited, the male doctor had a giales de San Carlos, Costa Rica. En Costa Rica, el significantly lower number of Papanicolaou cáncer de cérvix es el segundo tipo de cáncer femenino más exams administered in 2011 than the female común, a pesar de la accesibilidad del Papanicolaou gracias doctor; it would be important to analyze the al sistema de salud universal. No se ha evaluado el nivel de conocimiento de la población adolescente sobre el cáncer mean number of Pap smears administered de cérvix, aunque este grupo es el más expuesto a los fac- between male and female providers. Alongside, tores de riesgo. Este estudio investigó cuánto sabe el ado- it would be beneficial to compare women’s lescente sobre métodos de prevención, riesgo de infección, comfort levels in receiving the exam by doctors manifestaciones clínicas y servicios médicos. En total, 533 of each gender. estudiantes de séptimo a doceavo año de cinco colegios Three formal cervical cancer education públicos de la zona rural de San Carlos participaron en esta investigación durante marzo y abril 2012. Se observó que methods exist for adolescents in the San Car- los estudiantes carecen de información, ya que solamente los region: Semana de Salud once a year, el 30.8% sabían qué es el cáncer de cérvix. Nuestros resul- discussions by medical doctors in the high tados indican que es importante incluir el cáncer de cérvix schools and visits with the doctor. The trend en los programas de educación sexual del sistema de edu- that males were more likely to hear of an cación pública de Costa Rica, para educar a la población en educational cervical cancer campaign in their riesgo acerca de los métodos de prevención, desarrollo de la infección y servicios médicos. community lacks an explanation and would be a source for future study. Additionally, it is Palabras clave: adolescentes, cáncer de cérvix, Costa important to know which of these education Rica, educación sexual, Papanicolaou.

Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 877-886, September 2014 885 REFERENCES the only issue. American Journal of Public Health, 101(9), 1549-1555. Aguado, T., Beauvais, O., Byrne, S., & Gacic-Dobo, M. Maree, J. E., Wright, S. C., & Makua, T. P. (2011). Men’s (2010). Human papillomavirus and related cancers: lack of knowledge adds to the cervical cancer burden Costa Rica. World Health Organization. Retrieved in South Africa. European Journal of Cancer Care, from http://apps.who.int/hpvcentre/statistics/dyna- 20(5), 662-668. mic/ico/country_pdf/CRI.pdf?CFID=4005139&CFT OKEN=50211239 Nwankwo, K. C., Aniebue, U. U., Aquwa, E. N., Anarado, American Cancer Society. (2011). Cervical cancer: Pre- A. N., & Aqunwah, E. (2011). Knowledge attitudes vention and early detection. Retrieved from http:// and practices of cervical cancer screening among www.uptodate.com/contents/screening-for-cervical- urban and rural Nigerian women: A call for education cancer-rationale-and-recommendations and mass screening. European Journal of Cancer Care, 20(3), 362-367. Feldman, S., Sirovich, B. E., & Goodman, A. (2011). Screening for Cervical Cancer: Rationale and recom- Picado, K. (2007). Primera encuesta nacional de juventud, mendations. Retrieved from http://www.uptodate. Costa Rica. Fondo de Población de las Naciones com/contents /screening-for-cervical-cancer-rationa- Unidas. Retrieved from http://www.unfpa.or.cr/com- le-and-recommendations ponent/content /225?task=view Francis, S. A., Battle-Fisher, M., Hipple, L., Liverpool, J., RHO Cervical Cancer. (2009). About cervical cancer. Mofammere, N., Mosavel. M., & Soogun, S. (2011). Retrieved from http://www.rho.org/about-cervical- A qualitative analysis of South African women’s cancer-quickfacts.htm knowledge, attitudes and beliefs about HPV and Sokal, R. R. & Rohlf, F. J. (1995). Biometry:The principles cervical cancer prevention, vaccine awareness and and practice of statistics in biological research. New acceptance, and maternal-child communication about York, NY: W. H. Freeman and Co. sexual health. Vaccine, 29(47), 8760-8765. Instituto Nacional de Estadística y Censos [INEC]. (2011). Souza, C. (2011). Media screen time, physical activity and Población total por zona y sexo, según provincia, nutritional indicators as risk factors for childhood canton y distrito. Retrieved from http://www.inec. and adolescent overweight and obesity in Aguas go.cr/Web/Home/GeneradorPagina.aspx Zarcas, Costa Rica. Semester research project. Field Research in the Environment, Social Science & Kivistik, A., Lang, K., Baili, P., Antitila, A., & Veerus, P. Humanities (Spring 2011). San Pedro de Montes de (2011). Women’s knowledge about cervical cancer risk factors, screening, and reasons for non-partici- Oca, Costa Rica: Associated Colleges of the Midwest. pation in cervical cancer screening programme in Statpoint Technologies, Inc. (2011). Statgraphics Centurion Estonia. BMC Women’s Health, 11, 43. [computer software]. Available from http://www. Logan, L. & Mcifatrick, S. (2011). Exploring women’s statgraphics.com knowledge, experiences and perceptions of cervical Tse-Chuan, Y., Matthews, S. A., & Hillemeier, M. M. cancer screening in an area of social deprivation. (2011). Effect of health care system distrust on breast European Journal of Cancer Care, 20(6), 720-727. and cervical cancer screening in Philadelphia, Penns- Maine, D., Hurlburt, S., & Greeson, D. (2011). Cervical ylvania. American Journal of Public Health, 101(7), cancer prevention in the 21st century: Cost is not 1297-1305.

886 Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 877-886, September 2014 Effects of different agricultural systems on soil quality in Northern Limón province, Costa Rica

Emma Cornwell Departments of Biology and Environmental Studies, St. Olaf College, 1500 St. Olaf Ave, Northfield, MN 55057 USA; [email protected]

Received 20-iii-2014. Corrected 20-iv-2014. Accepted 26-v-2014.

Abstract: Conversion of native rainforest ecosystems in Limón Province of Costa Rica to banana and pine- apple monoculture has led to reductions in biodiversity and soil quality. Agroforestry management of cacao

(Theobroma cacao) is an alternative system that may maintain the agricultural livelihood of the region while Rica program Costa Special section (ACM) more closely mimicking native ecosystems. This study compared physical, biological and chemical soil quality indicators of a cacao plantation under organic agroforestry management with banana, pineapple, and pasture systems; a native forest nearby served as a control. For bulk density and earthworm analysis, 18 samples were collected between March and April 2012 from each ecosystem paired with 18 samples from the cacao. Cacao had a lower bulk density than banana and pineapple monocultures, but greater than the forest (p<0.05). Cacao also hosted a greater number and mass of earthworms than banana and pineapple (p<0.05), but similar to forest and pasture. For soil chemical characteristics, three composite samples were collected in March 2012 from each agroecosystem paired with three samples from the cacao plantation. Forest and pineapple ecosystems had the lowest pH, cation exchange capacity, and exchangeable nutrient cations, while cacao had the greatest (p<0.05). Total nutrient levels of P and N were slightly greater in banana, pineapple and pasture than in cacao; probably related to addition of chemical fertilizer and manure from cattle grazing. Forest and cacao also had greater %C, than other ecosystems, which is directly related to soil organic matter content (p<0.0001). Overall, cacao had more favorable physical, biological and chemical soil characteristics than banana and pineapple monocultures, while trends were less conclusive compared to the pastureland. While organic cacao was inferior to native forest in some soil characteristics such as bulk density and organic carbon, its soil quality did best mimic that of the native forest. This supports the organic cultivation of cacao as a desirable alternative to banana and pineapple monoculture. Rev. Biol. Trop. 62 (3): 887-897. Epub 2014 September 01.

Key words: agroforestry, banana (Musa acuminata), cacao (Theobroma cacao), Costa Rica, organic, pineapple (Ananus comosus), pasture, soil quality.

In the Caribbean coastal plain of Costa Gordon, Thevathasan, Oppong, & Quashie- Rica, conversion of native rainforest ecosys- Sam, 2005). tems to monocrop agriculture production has One crop that is commonly grown under led to habitat loss and biodiversity reductions agroforestry management is cacao (Theobroma (Vaughan, Ramírez, Herrera, & Guries, 2007). cacao). Previous studies indicated that, in com- Agroforestry systems, consisting of shade trees parison to monocrop systems, cacao cultivation integrated among primary crops, can more has larger nutrient inputs, reduced nutrient closely mimic the floral and faunal biodiversity losses, and better soil physical properties such of native forests. Past research has shown that as penetrability and bulk density (Young, 1989; after many years of agroforestry management, Isaac et al., 2005; McNeely & Schroth, 2006; fields can attain levels of soil quality equal to Vallejo, Roldan, & Dick, 2010; Tully & Law- those of an adjacent native forest system (Isaac, rence, 2011).

Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 887-897, September 2014 887 Soil health is very important to success- agroforestry management with conventional ful agriculture. A high quality soil provides banana and pineapple monocultures, as well as an environment for optimum root growth that a neighboring pasture and fragment of primary enhances crop health and productivity (Bald- rainforest. I also compared the soil quality of win, 2006). Beyond merely enhancing the organic and non-organic cacao cultivation. By growth of one desired crop, a balanced, healthy adding to the body of research surrounding soil can sustain the productivity of other ani- this topic, I aimed to strengthen the case for mals and plants, support human health, and cacao agroforestry as an alternative agriculture improve the quality of air and water (Doran, method that can mitigate the negative effects of Sarrantonio, & Liebig, 1996; Bonanomi et deforestation while still providing a livelihood al., 2011). Because healthy soil can benefit for area farmers. whole ecosystems, analyzing the soil qual- ity of a certain system of agriculture can MATERIALS AND METHODS provide a strong case for or against its envi- ronmental sustainability. Study area: The study area is located in Soil quality is a complex entity that cannot Pueblo Nuevo de Villa Franca de Guácimo, in be measured directly. Instead, physical, chemi- Limón Province, Costa Rica (10º20’ N - 83º20’ cal and biological indicators must be consid- W). This area receives a mean annual rainfall ered. Physical factors such as bulk density give of 6 000mm with a mean air temperature of an indication of the quality of soil for adequate 26ºC and is located 40m above sea level, in root growth and movement of water and air the Premontane Wet Forest Life Zone (Hold- (USDA, 1999). In addition, macroinvertebrate ridge, 1964; Vaughan et al., 2007). Major soil and microbial communities are essential factors types (sub-group/great group/suborder/order) to consider in the maintenance of soil fertility, on this farm and surrounding areas include because they play crucial roles in influencing typic tropoquept, aeric tropoquept, aquic eutro- physical and chemical properties of the soil pept, aquic dystrandept, and oxic dystrandept (Sánchez-de Leon, de Melo, Soto, Johnson- (Obando, 1982). Maynard, & Lugo-Perez, 2006). It is also very The primary study site is a 110ha cacao important, especially in tropical ecosystems, plantation that has been under organic manage- to maintain high levels of soil organic matter, ment since 2002. This land was originally pri- because this is the primary source of energy, mary rainforest, but was managed as livestock nitrogen and soil structure for soil and plant pasture for many years before its conversion ecosystems (Sánchez-de Leon et al., 2006). to cacao agriculture in the 1980s. The primary Compared to monocrop systems, agroforestry tree crop, Theobroma cacao, is maintained has been shown to have more abundant and under shade trees Eucalyptus deglupta, Cocos diverse microfauna, and greater levels of soil nucifera, Leucaena leucocephala, and Musa organic matter (Sánchez-de Leon et al., 2006). acuminata (R. Wieme, unpublished). Organic One can also note differences in the soil fertilizer made of composted coffee husks is quality of farms under organic vs. conventional applied biannually, and since 2011 an organic management. The use of chemical fertilizers herbicide has been applied for weed control; and pesticides on “conventionally” managed no chemical fertilizers or pesticides have been farms can lead to polluted runoff that has det- used in the prior 13yrs. A subterranean system rimental effects on the organisms in the wider of drainage tubes was installed in 1997 to help ecosystem (Castillo et al., 2006). These chemi- maintain adequate soil moisture content (G. cals can also affect soil health. Herrera, pers. comm., [email protected]). The purpose of this study was to compare There is a strip of cacao farmland bordering the physical, biological, and chemical indicators adjacent banana plantation on which the cacao of soil quality on an organic cacao farm under is officially classified as conventional.

888 Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 887-897, September 2014 A large banana (Musa acaminata) mono- x Wdsub)/Wfsub, where Wd=oven dried weight culture borders the cacao plantation on the of full sample, Wf =fresh weight of full sample, west and is maintained with regular applica- and Wdsub and Wfsub are the dry and fresh tions of chemical fertilizers, insecticides, and weights of the subsample. I then calculated fungicides. Pasture for cattle grazing borders bulk density using the formula (Wd/V), where the north side of the cacao plantation. These V=soil core volume. pastures consist mainly of African star grass (Cynodon nlemfuensis) with some large iso- On-site analysis of earthworm abun- lated trees like Coussapoa villosa, Ocotea dance: Using a 25x 25cm quadrant frame, I sinuata, Erythrina poeppigiana and Hura crep- excavated the soil at each paired sampling site itans and are not chemically treated (Vaughan to a depth of 20cm and searched by hand for et al., 2007). A fragment of native primary earthworms (Gregory, Shea, & Bakko, 2005; forest is located approximately 2km northeast Sánchez-de Leon et al., 2006). I obtained three of the cacao plantation. I also studied a pine- replicates at each sampling site and determined apple monoculture that is maintained conven- the total number and mass of earthworms. tionally with chemical fertilizers, insecticides, Because earthworm abundance varied greatly Special section (ACM) Costa Rica program Costa Special section (ACM) fungicides and an intensive system of open depending on weather, earthworms from paired drainage ditches. This pineapple monoculture sites were collected on the same day; all existed on very similar soil types as those of the samples were collected between March 8 and cacao plantation. April 19, 2012. Selection of sampling sites: I selected Laboratory analysis of soil samples: For sampling sites on similar soil types located chemical laboratory analysis, I took samples at on roughly the same topographical features, three of the six paired sampling sites for each according to an available soil types map (Oban- agricultural ecosystem on March 27, 2012. At do, 1982). To compare banana and pasture each site, I combined five cores, 10cm deep by with cacao, I selected six paired plots along 4cm wide. Samples were sent to the Centro de the cacao plantation with each plot located Investigaciones Agronómicas at the University approximately 15m perpendicular to the border between cultivation types. of Costa Rica, San Pedro de Montes de Oca I also designated six study sites in the pine- for laboratory analysis. Chemical laboratory apple plantation and three control sites in the analysis included pH, acidity, cation exchange natural primary forest. For analysis, these sites capacity, concentrations of macro- and micro- were paired with six cacao sites to minimize nutrients, %C and %N. variation due to sampling time; samples for I used Statgraphics Centurion XVI soft- bulk density and earthworm abundance from ware (Statpoint Technologies, Inc., 2012) to paired sites were collected on the same day. analyze each soil quality indicator separately for significant differences between crop types. On-site analysis of bulk density: At each For analysis of the forest ecosystem, I used sampling site described above, I extracted three one-way ANOVA tests. The other ecosystems core samples of known volume to approximate- were compared to their corresponding paired ly 10cm depth at each sampling site; I weighed sites in the cacao using paired comparisons these samples to determine their fresh weight. two-way parametric ANOVA tests, in order to A subsample of each soil core was weighed, account for some of the inherent microscale dried in a microwave oven until mass no longer variation in soil characteristics. All ANOVAs decreased, and then weighed again to determine were performed after testing for normality dry weight. To calculate the oven-dried weight and homogeneity of variance (Sokal & Rohlf, of the full sample, I used the formula Wd = (Wf 2012). Unplanned comparisons of means were

Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 887-897, September 2014 889 estimated with least significant, Tukey or cacao site ( =0.61g/cm3) (F=1.73; df=1, 24; Scheffe confidence intervals. p=0.20, Fig. 1).

Earthworm abundance: Mean number of RESULTS earthworms did not differ significantly between Bulk density: Among all ecosystems, forest ( =13.9) and cacao ( =15.1) (F=0.16; mean soil bulk density was greatest in pine- df=1, 16; p=0.69; Fig. 2). There was also no significant difference in earthworm mass apple ( =0.79g/cm3) and lowest in forest between forest ( =2.93g) and cacao ( =3.55g) ( =0.48g/cm3) (F=13.1; df=7, 127; p≤0.0001, (F=0.76; df=1, 16; p=0.3965). There were Fig. 1). Forest soil bulk density was less than fewer earthworms in the banana plantation ( any other crop, including its paired cacao 3 =1.8) than in its corresponding paired sites of site (0.66g/cm ) (p<0.05, Fig. 1). Banana ( cacao ( =8.6) (F=38.86; df=1, 24; p<0.0001; 3 =0.69g/cm ) had greater bulk density than its Fig. 2). There was also a lower earthworm corresponding paired cacao sites ( =0.61g/ mass in the banana ( =0.39g) than correspond- 3 cm ) (F=19.92; df=1, 24; p=0.0002; Fig. 1). ing cacao sites ( =1.73 g) (F=7.36; df=1, 3 Pineapple ( =0.79g/cm ) also had greater bulk 16; p=0.0153). In the pineapple plantation, density than corresponding paired cacao sites there was a lower number ( =1.5) and mass ( =0.66g/cm3) (F=28.17; df=1, 24; p≤0.0001; ( =0.24g) of earthworms than in the corre- Fig. 1). The pasture system ( =0.60g/cm3) sponding paired sites in the cacao (number: did not differ significantly from its paired =8.0; mass: =1.86g) (number: F=48.03;

0.85

) 0.75 3

0.65

0.55 Bulk Density (g/cm 0.45

0.35 Cacao Forest Cacao Cacao Cacao Pasture Banana Pineapple Organic Cacao Organic Conventional Cacao Conventional

a b c b b

Ecosystem

Fig. 1. Mean bulk density by ecosystem. Adjacent bars indicate mean bulk density from paired sampling sites. Error bars represent 95% LSD confidence interval. a: Bulk density was significantly different between forest and cacao (p<0.0001). One-way ANOVA, Fisher’s LSD. b: Bulk density differed significantly (p<0.0001). Two-way paired-comparison ANOVA, Fisher’s LSD. c: Bulk density was not significantly different (p>0.05). Two-way paired-comparison ANOVA, Fisher’s LSD. Pueblo Nuevo de Guácimo, Limón Province, Costa Rica. March-April 2012.

890 Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 887-897, September 2014 25

20

15

10 Earthworm Number 5

0 Cacao Forest Cacao Cacao Cacao Pasture Banana Pineapple Organic Cacao Organic Conventional Cacao Conventional Special section (ACM) Costa Rica program Costa Special section (ACM)

a b c b b

Ecosystem

Fig. 2. Earthworm mean numbers by ecosystem. Adjacent bars indicate mean earthworm number from paired sampling sites. Error bars represent 95% LSD confidence interval. a: Earthworm number did not differ significantly between forest and cacao (p>0.05). One-way ANOVA. Fisher’s LSD. b: Earthworm number differed significantly (p<0.01). Two-way paired- comparison ANOVA, Fisher’s LSD. c: Earthworm number did not differ significantly (p>0.05). Two-way paired-comparison ANOVA, Fisher’s LSD. Pueblo Nuevo de Guácimo, Limón Province, Costa Rica. March-April 2012.

df=1, 24; p≤0.0001; mass: F=26.19; df=1, 20; =5.7) did not differ significantly from its cor- p=0.0001) (Fig. 2). The pasture ecosystem had responding paired sites in the cacao ( =6.0) a greater number of earthworms ( =11.9) than (Table 1). the corresponding cacao paired sites ( =6.6) Cation exchange capacity (CEC) was greater in the cacao ( =22.55cmol+/L) than (F=7.91; df=1, 24; p=0.0096; Fig. 2). However, the forest ( =10.45cmol+/L) (p<0.05; Table 1). pasture did not have a significantly different Cacao CEC was also greater ( =22.55cmol+/L) earthworm mass ( =2.30g) than corresponding than in corresponding paired sites of pineapple paired cacao sites ( =1.80g) (F=0.61; df=1, ( =16.74cmol+/L) (F=53.9; df=1, 2; p=0.0087; 16; p=0.45). Table 1). Cacao also had a greater CEC ( =21.16cmol+/L) than corresponding paired Physicochemical soil characteristics: sites in pasture ( =17.20cmol+/L) (F=113; Mean pH was lower in the forest ( =4.3) than df=1, 2; p=0.0087; Table 1). However, banana + the cacao ( = 5.8) (p<0.01; Table 1). Pineapple CEC ( =15.10cmol /L) did not differ signifi- cantly from corresponding cacao paired sites ( =4.6) also had a lower pH than its corre- ( =15.93 cmol+/L) (Table 1). sponding cacao paired sites ( =5.8) (F=61.7; df=1, 2; p=0.0158; Table 1). In addition, pas- Exchangeable nutrient cations: For- ture sites had lower pH ( =5.2) than corre- est had less mean concentration of Ca2+ ( sponding cacao paired sites ( =5.6) (F=24.14; =3.90cmol+/L) than cacao ( =17.10cmol+/L) df=1, 2; p=0.0390; Table 1). Banana pH ( (p<0.01; Table 1). Pineapple also had less Ca2+

Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 887-897, September 2014 891 TABLE 1 Mean soil physicochemical characteristics and exchangeable nutrient cations compared between different agroecosystems

Agroecosystems compared pH CEC (cmol+/L) Ca2+ (mol/L) Mg2+ (mol/L) K+ (mol/L) Forest vs. Cacaoa Forest 4.3** 10.45* 3.90** 2.19* 0.16* Cacao 5.8 22.55 17.10 4.81 0.52 Banana vs. Cacaob Banana 5.7 15.1 10.33 3.77 0.78 Cacao 6.0 15.9 11.85 3.39 0.52 Pineapple vs. Cacaob Pineapple 4.6* 16.74* 9.80* 4.81 0.50 Cacao 5.8 22.55 17.10 4.81 0.52 Pasture vs. Cacaob Pasture 5.2* 17.20** 11.72** 4.69 0.28 Cacao 5.6 21.16 16.13 4.39 0.50 a One-way ANOVAs among all ecosystems Tukey HSD. b Two-way paired-comparison ANOVA, Fisher LSD. * means comparison significant at <0.05; **at <0.01. Data obtained from three composite samples taken from paired sampling sites in Pueblo Nuevo de Guácimo, Limón Province, Costa Rica. March-April 2012.

TABLE 2 Mean soil total nutrient and organic carbon levels in different agroecosystems

Ecosystems compared P (mg/L) % N % Cc Forest vs. Cacaoa Forest 25.6* 0.49* 5.59* Cacao 12.3 0.37 3.85 Banana vs. Cacaob Banana 72.6* 0.41 4.06 Cacao 18.0 0.47 4.39 Pineapple vs. Cacaob Pineapple 21.6 0.15* 1.34** Cacao 12.3 0.37 3.85 Pasture vs. Cacaob Pasture 35.6 0.46 4.28 Cacao 15.0 0.41 4.05 aOne-way ANOVA among all ecosystems Tukey HSD. bTwo-way paired-comparison ANOVA, Fisher LSD. c%C is directly correlated with % organic matter (OM) in the soil. To determine estimated OM, multiply values by 1.43. *means comparison significant at <0.05; **at <0.01. Data obtained from three composite samples taken from paired sampling sites in Pueblo Nuevo de Guácimo, Limón Province, Costa Rica. March-April 2012.

( =9.80 cmol+/L) than corresponding paired ecosystems varied from 3.39cmol+/L to sites in the cacao ( =17.10cmol+/L) (F=53.6; 4.81cmol+/L, and none differed significantly df=1, 2; p=0.0182; Table 1). Similarly, Ca2+ from corresponding paired sites in the cacao was lower in pasture ( =11.72cmol+/L) than (Table 1). cacao ( =16.13cmol+/L) (F=121; df=1, 2; Reflecting the trends in Mg2+, the only p=0.0081; Table 1). Banana, however, did not significant difference in K+ concentration was have a significantly different level of Ca2+ ( between forest ( =0.16cmol+/L) and cacao =10.33cmol+/L) than corresponding cacao ( =0.52cmol+/L) (p<0.05; Table 1). While paired sites ( =11.85cmol+/L) (Table 1). no other ecosystems differed significantly Forest ( =2.19cmol+/L) had less Mg2+ from corresponding cacao paired sites, K+ than cacao ( =4.81cmol+/L) (p<0.05; Table concentration in banana ( =0.78cmol+/L) 1). Mean concentrations of Mg2+ in the other tended to be greater than that in cacao (

892 Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 887-897, September 2014 =0.52cmol+/L). The K+ concentration in pas- The lower bulk density in the forest and cacao ture ( =0.28cmol+/l) tended to be less than that indicated that the soil structure in these systems in cacao ( =0.50cmol+/L) (Table 1). was more favorable for plant growth than that of the banana and pineapple monocultures. This Total nutrients: Mean phosphorus lev- could be associated with the greater abundance els were significantly greater in forest ( of earthworms and more complex root structure =25.6mg/L) than cacao ( =12.3mg/L) (Table in these agriculture systems. Abundant trees in 2). In the other ecosystems, there was great these systems also add organic material to the variation in P levels among sampling stations, soil through litterfall which may contribute to even within the same ecosystem. Although lesser soil compaction. Ten years of organic there were no significant differences between compost addition to the cacao soil probably ecosystems, the organic cacao consistently had was the greatest factor in improving its soil a lower mean P level than its corresponding bulk density. In contrast, all natural vegetation paired sites in banana, pineapple, and pasture- is removed from the banana and pineapple land (Table 2). plantations for “monocrop” cultivation, leav- Mean percentage of N did not differ signif- ing the soil exposed to the elements. Intensive icantly between forest ( =0.49%) and cacao ( cropping, trampling of soil by workers, lack of Rica program Costa Special section (ACM) =0.37%) (p>0.05; Table 2). Other ecosystems organic material addition, and pounding of rain showed very similar mean %N values which can compact the soil, leading to an undesir- did not differ significantly from corresponding ably high bulk density (Hajabbasi, Jalalian, & paired sites in the cacao; values ranged from Karimzadeh, 1997). 0.37% to 0.47% with the exception of pine- The soil in the cacao had a greater bulk apple, which had a much lower %N ( =0.15%) density than that in the forest, suggesting than its paired cacao sites ( =0.37) (F=43.6; that while preferable to banana and pineapple df=1, 2; p=0.0222; Table 2). monoculture, agroforestry management did not attain a soil structure as favorable as the natu- Soil organic carbon: Mean percentage of ral forest. A study of livestock pasture found organic carbon (%C) did not differ significantly a similar relationship between cacao and for- between the forest ecosystem ( =5.59%) and est bulk density, showing that an agroforestry cacao ( =3.85) (p>0.05; Table 2). However, system had intermediate levels of soil compac- mean percentage of C was lower in the pine- tion and bulk density between a conventional apple ( =1.34%) than corresponding paired monocropped pasture and natural forest (Valle- sites in the cacao ( =3.85%) (F=82.6; df=1, 2; jo et al., 2010). While I didn’t find significant p=0.0119; Table 2). Percentage of C in banana differences between cacao and pasture, the and pasture did not differ significantly from cacao plantation was previously managed as that of corresponding paired sites in organic pasture before its conversion 30 years ago; pre- cacao, ranging from 3.88% to 4.39% (Table 2). vious research has found that, over time, soil All results of the above comparisons under agroforestry management can continue (Tables 1, 2) for “conventional cacao” plots to decrease in bulk density (Isaac et al., 2005). did not differ significantly from those of their Earthworm abundance was greater in the paired sites in the organic cacao. cacao than in the monoculture banana and pineapple systems, and was similar between DISCUSSION the cacao and forest. This suggests that agro- forestry management of cacao better mimics Soil bulk density is an indication of soil the macroinvertebrate population of a natural physical structure. A lower bulk density facili- rainforest. This could be related to the presence tates unrestricted root growth, oxygen and of shade trees in the cacao and forest, which water flow through the soil (USDA, 1999). create a favorable environment for earthworms.

Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 887-897, September 2014 893 A study of earthworms in coffee plantations in forest was the lowest of all ecosystems studied. Turrialba, Costa Rica found that earthworm Forests in the humid tropics of Latin America density and fresh weight were greater in plan- have a greater percentage of acid soils (82%) tations under shade vs sun, and organic vs. compared to those in Africa (56%) or Asia conventional management (Sánchez-de Leon (38%) (Sánchez, 1987). et al., 2006). Cation exchange capacity (CEC) indicates Earthworms usually have positive effects the nutrient retention capability of a soil by on soil quality: as they feed and burrow through measuring the quantity of available negative the soil, they aid in decomposition of organic charges to which nutrient cations such as K+, matter, nutrient cycling, microbial activity, soil Ca2+, and Mg2+ can bind. A greater CEC pro- porosity and bulk density (Fragoso et al., 1997; vides a potential bank of nutrients to replace Brown et al., 1999). However, they can also those taken up by plants or leached by rain- negatively affect agricultural production by water (Camberato, 2001). Soil with greater competing with plants for water and nutrients quantities of clay and organic matter tends to (Brown et al., 1999). The specific benefits or have greater CEC (Camberato, 2001). In my drawbacks to earthworm presence depend on study, CEC was lowest in the native forest, the type of agricultural system and the species even though it had a greater amount of organic of earthworms present (Fragoso et al., 1997, carbon. This is probably associated with the Brown et al., 1999). Exotic earthworms can low pH in the forest; more acidic soils have replace native species, reducing the productiv- fewer negative charges, and many of the bind- ity of the ecosystem (Fragoso et al., 1997). ing sites are occupied by H+ and Al3+ cations Exotics can tolerate a wider range of soil con- (Al3+ is present in cation form at pH lower than ditions, while native earthworms usually can’t 5.0) (Lines-Kelly 1993). This trend of increas- survive in disturbed ecosystems like banana ing CEC with greater pH carried overall among or pineapple monocultures (Lee, 1983). Case all ecosystems. Of the agricultural systems, studies have shown that in tropical ecosystems CEC was lowest in the banana and pineapple native earthworm species are more likely to be monocultures, intermediate in the pasture, and maintained when there are many trees, chemi- greatest in the cacao. This is consistent with cals are not used, and organic inputs are domi- the results of a past study which found sig- nant in the system (Fragoso et al., 1997). It was nificantly greater CEC under an agroforestry not within the scope of this study to classify the system than pastoral land, and consequently earthworms found in the different agricultural greater CEC under pastoral than arable land systems into species or functional groups, but (Sharma et al., 2009). future studies might find that the species com- Regarding exchangeable nutrient cations, position of earthworms differs depending on such cations as Ca2+, Mg2+, and K+ occupy the agricultural system. the negative binding sites in the soil indicated Regarding physicochemical soil charac- by CEC. As with CEC, native forest had the teristics, the more neutral pH that I found in lowest levels of Ca2+, Mg2+, and K+ of all eco- cacao as compared to the banana and pine- systems. This contradicts past research which apple monocultures is consistent with previous suggests that decomposition of litterfall and research showing greater pH in agroforestry neutral soil reactions lead to greater nutrient than monocultures (Sharma et al., 2009). In cation concentrations in tree-based systems tropical forest and agroforestry systems, the (Lal, 1989; Sharma et al., 2009). Greater nutri- deposition of bases over many years of tree ent cation concentrations in the agricultural growth can increase the pH of the soil, buffer- systems may be attributed to the application of ing the effects of nutrient leaching (Young, fertilizer, whether organic (such as the compost 1989; Nair, Buresh, Mugendi, & Latt, 1999; used in the cacao) or chemical (for example, it Sharma et al., 2009). The pH in the native appeared that calcium carbonate had recently

894 Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 887-897, September 2014 been applied to the banana plantation when %C as well. Previous research has also found soil samples were taken). The cacao had sig- greater SOM in native forest and agroforestry nificantly more Ca2+ than pasture or pineapple systems than monocropping, which has been sites, and while I did not find a significant dif- attributed to greater root biomass and litterfall ference between banana and cacao, this may (Szott, Fernandes, & Sánchez, 1991; Mad- have been related to the recent application of humitha, Singh, & Khan, 1997; Sharma et al., calcium carbonate in the banana. Differences 2009). The pasture had a slightly greater %C among agricultural systems in Mg+ and K+ than its paired sites in the cacao, but not sig- were not as great, although previous research nificantly so. Past studies comparing pasture has found that agroforestry systems have great- and agroforestry systems have had conflicting er amounts of these nutrients than pastoral or results: some found that pasture has the capaci- arable land (Sharma et al., 2009). ty to sequester more carbon, while others found Regarding total nutrients, cacao tended to more carbon in agroforestry systems (Lantz, have lower levels of phosphorus (P) than its Lal, & Kimble, 2001; Sharma et al., 2009). corresponding paired sites in the banana and This study indicates that organic agrofor- pineapple although these differences were not estry management of cacao results in soil with significant. P was also significantly greater greater overall quality for plant and animal Rica program Costa Special section (ACM) in the forest than the cacao. Nitrogen (N) health than banana and pineapple monocul- percentage did not differ as greatly among tures. Shade trees add organic material through ecosystems as P, but was significantly lower in litterfall and provide a haven for wild animals, the pineapple than the cacao. These results con- which in turn add their own nutrients and tradict previous studies which have shown that organic matter to the soil. The lack of chemi- levels of N, P, and potassium were all greater cal additives allows native macroinvertebrates under tree cover than in open field monocul- such as earthworms to flourish, which also tures (Aggarwal, 1980; Radwanski & Wickens, improves bulk density and nutrient content. 1981; Young, 1989). The greater levels of P When agriculture is managed to mimic a in banana and pineapple may be explained by native rainforest ecosystem, the results of this chemical fertilizer application which artifi- study showed that many factors work together cially increases nutrient levels. Future studies in a complicated net of interactions to main- analyzing nutrient content of rainwater runoff tain soil fertility. would help determine if nutrients such as N Banana and pineapple monocultures, in and P are being applied in excess, an economi- contrast, maintain high levels of productivity cally ineffective and environmentally damag- only with the addition of chemical fertilizers ing practice. and pesticides. As this study shows, soil qual- Cacao also had slightly lower P and %N ity is reduced; it is unlikely that this soil will than pasture. Sharma et al. (2009) attributed support such aggressive management for many higher levels of N and P in pastoral land than years without a drastic decline in productivity. an agroforestry system to nutrients added by Because many residents of Limón Prov- manure from grazing cattle. ince rely on agriculture as their livelihood, it Regarding soil organic carbon, the per- is essential to maintain the fertility of the soil centage of organic carbon in the soil is directly so that it can support the healthy crop growth related to the soil organic matter (SOM), and in future generations. Banana and pineapple can therefore be used to model SOM levels. growers may try to mimic the agroforestry While not significantly different, the forest did approach of cacao production, imitating as have greater %C than the cacao and the great- closely as possible the native ecosystem. This est %C overall. Although the only ecosystem might involve allowing ground cover to prevent with a significantly lower %C than organic soil compaction; finding organic alternatives to cacao was pineapple; banana had slightly lower fertilizers and pesticides; or even developing a

Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 887-897, September 2014 895 more integrated agroforestry approach like that Palabras clave: banano (Musa acuminata), cacao (Theo- of the cacao plantation. There are examples broma cacao), calidad del suelo, Costa Rica, manejo agro- of the successful cultivation of banana and forestal, cultivo orgánico, piña (Ananus comosus), potrero. pineapple alongside other crop plants (such as cacao) in agroforestry systems (Khaleque & REFERENCES Gold, 1993; Jose, 2009). Agroforestry manage- ment of not just cacao but banana and pine- Aggarwal, R. K. (1980). Physico-chemical status of soils apple as well may indeed be the best way to under “khejri” (Prosopis cineraria Linn.). In H. S. Mann & S. K. Saxena (Eds.), Khejri (Prosopis cinera- maintain soil quality in the region. ria) in the Indian Desert-Its Role in Agroforestry (pp. 32-37). Jodhpur, India: Central Arid Zone Research Institute (CAZRI). ACKNOWLEDGMENTS Baldwin, K. R. (2006). Soil quality considerations for orga- Thanks to Michael McCoy and Christo- nic farmers. Center for Environmental Farming Sys- tems. Raleigh, North Carolina, USA: North Carolina pher Vaughan for logistical assistance, sup- Cooperative Extension Service. port with fieldwork, and help with manuscript Bonanomi, G., D’Ascoli, R., Antignani, V., Capodilupo, revisions. Thanks also to Hugo Hermelink for M., Cozzolino, L., Marzioli, R., & Zoina, A. (2011). providing the use of his cacao plantation, as Assessing soil quality under intensive cultivation and well as supplies and funding for soil analysis. tree orchards in Southern Italy. Applied Soil Ecology, Geovanny Herrera, Oldemar Miranda, Marco 47, 184-194. Antonio Herrera, Manuel Herrera and several Brown, G. G., Pashanasi, B., Villenave, C., Patron, J. C., other employees of Finmac cacao plantation Senapati, B. K., Giri, S., & Boyer, J. (1999). Effects provided assistance with field work, supplies, of earthworms on plant production in the tropics. In P. Lavelle, L. Brussaard, & P. Hendrix (Eds.), and background information. The Associated Earthworm management in tropical agroecosystems Colleges of the Midwest Costa Rica provided (pp. 87-148). Wallingford, United Kingdom: CABI financial support for soil analysis. Publishing. Camberato, J. J. (2001). Cation exchange capacity: Everything you want to know and much more. South RESUMEN Carolina Turfgrass Foundation News. Columbia, SC, USA: Clemson University. Retrieved from http:// Efectos de sistemas agrícolas diferentes sobre files.clino.webnode. com/200000022-0bdf60cd97/ calidad de suelos en el norte de la provincia de Limón, What%20is%20CEC.pdf Costa Rica. Muchos de los ecosistemas boscosos naturales de la provincia de Limón, Costa Rica, se han convertido Castillo, L. E., Martínez, E., Ruepert, C., Savage, C., Gilek, en monocultivos de banano y piña, lo que ha reducido la M., Pinnock, M., & Solis, E. (2006). Water quality biodiversidad y la calidad de los suelos. El manejo agro- and macroinvertebrate community response following forestal del cacao (Theobroma cacao) es un sistema alter- pesticide applications in a banana plantation, Limon, nativo, que puede generar ingresos para los agricultores Costa Rica. The Science of the Total Environment, de la región mientras imita mejor los ecosistemas nativos. 367, 418-432. En esta investigación se compararon los indicadores físi- Doran, J. W., Sarrantonio, M., & Liebig, M. A. (1996). cos, biológicos, y químicos de la calidad del suelo en un Soil health and sustainability. In L. S. Donald (Ed.), cacaotal orgánico y agroforestal, una plantación bananera, Advances in Agronomy (pp. 1-54). Lincoln, NE, USA: una piñera, un potrero y un bosque natural. En general, el Academic Press. suelo del cacaotal tuvo características físicas, biológicas y Fragoso, C., Brown, G. G., Patrón, J. C., Blanchart, E., químicas mejores que los monocultivos de banano y piña, Lavelle, P., Pashanasi, B., & Kumar, T. (1997). mientras las tendencias en comparación con el potrero fueron menos marcadas. Aunque algunas características Agricultural intensification, soil biodiversity and del suelo en el cacaotal orgánico, como densidad y cantidad agroecosystem function in the tropics: The role of de carbono, fueron inferiores al bosque nativo, la calidad earthworms. Applied Soil Ecology, 6, 17-35. del suelo en el cacaotal imitó mejor la del bosque nativo. Gregory, M. M., Shea, K. L., & Bakko, E. B. (2005). Com- Los resultados de esta investigación apoyan el cultivo de paring agroecosystems: Effects of cropping and tillage cacao como una alternativa deseable al monocultivo de patterns on soil, water, energy use and productivity. banano y piña. Renewable Agriculture and Food Systems, 20, 81-90.

896 Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 887-897, September 2014 Hajabbasi, M. A., Jalalian, A., & Karimzadeh, H. R. (1997). Obando, O. A. (1982). Mapa semi-detallado de suelos Deforesatation effects on soil physical and chemical [map]. 1:10,000. San Jose, Costa Rica: HDA Agrope- properties, Lordegan, Iran. Plant Soil, 190, 301-308. cuaria Tica S.A. Holdridge, L. (1964). Life Zone Ecology. San José, Costa Radwanski, S. A. & Wickens, G. E. (1981). Vegetative Rica: Tropical Science Center. fallows and potential value of the neem tree (Azadi- Isaac, M. E., Gordon, A. M., Thevathasan, N., Oppong, S. rachta indica) in the tropics. Economic Botany, 35, K., & Quashie-Sam, J. (2005). Temporal changes in 398-414. soil carbon and nitrogen in west African multistrata Sánchez, P. A. (1987). Management of acid soils in the agroforestry systems: A chronosequence of pools and humid tropics of Latin America. In P. A. Sánchez, fluxes. Agroforestry Systems, 65, 23-31. E. R. Stoner, & E. Pushparajah (Eds.), Management Jose, S. (2009). Agroforestry for ecosystem services and of Acid Tropical Soils for Sustainable Agriculture. environmental benefits: An overview. Agroforestry International Board for Soil Research and Mana- Systems, 76, 1-10. gement Proceedings No. 2 (pp. 64-107). Bangkok, Thailand: International Board for Soil Research and Khaleque, K. & Gold, M. A. (1993). Pineapple agroforestry: Management. An indigenous system among the Garo community of Bangladesh. Society and Natural Resources, 6, 71-78. Sánchez-de Leon, Y., de Melo, E., Soto, G., Johnson-May- nard, J., & Lugo-Perez, J. (2006). Earthworm popula- Lal, R. (1989). Potential of agroforestry as a sustaina- tions, microbial biomass and production in different ble alternative to shifting cultivation: Concluding

experimental coffee agroforestry management sys- Rica program Costa Special section (ACM) remarks. Agroforestry Systems, 8, 239-242. tems. Caribbean Journal of Science, 42, 397-409. Lantz, A., Lal, R., & Kimble, J. (2001). Land use effects Sharma, K. L., Raju, K. R., Das, S. K., Rao, B., Kulkarni, B. on soil carbon pools in two major land resource areas S., Srinivas, K., & Gajbhiye, P. N. (2009). Soil fertility of Ohio, USA. In D. E. Stott, R. H. Mohtar, & G. and quality assessment under tree-, crop-, and pasture- C. Steinhardt (Eds.), Sustaining the Global Farm based land-use systems in a rainfed environment. (pp.499-502). West Lafayette. In: Proceedings of the Communications in Soil Science and Plant Analysis, 10th International Soil Conservation Organization, 40, 1436-1461. Purdue University and the USADA-ARS National Sokal, R. R. & Rohlf, F. J. (2012). Biometry: The principles Soil Erosion Research Laboratory, USA. and practice of statistics in biological research. New Lee, K. E. (1983). Earthworms of tropical region. Some York, NY, USA: W. H. Freeman & Co. aspects of their ecology and relationships with soils. Statpoint Technologies, Inc. (2012). Statgraphics Centurion In J. E. Satchell (Ed.), Earthworm Ecology: From [computer software]. Available from http://www.stat- Darwin to Vermiculture (pp. 449-466). London, Uni- graphics.com. ted Kingdom: Chapman & Hall. Cited in: Fragoso, C., Brown, G. G., Patrón, J. C., Blanchart, E., Szott, L. T., Fernandes, E. C. M., & Sánchez, P. A. (1991). Lavelle, P., Pashanasi, B., Senapati, B., & Kumar, T. Soil-plant interactions in agroforestry systems. Forest (1997). Agricultural intensification, soil biodiversity Ecology Management, 45, 127-152. and agroecosystem function in the tropics: The role of Tully, K. L. & Lawrence, D. (2011). Closing the loop: earthworms. Applied Soil Ecology, 6, 17-35. Nutrient balances in organic and conventional coffee Lines-Kelly, R. (1993). Cation exchange capacity. In Soil agroforests. Journal of Sustainable Development, 35, Sense Leaflet. Wollongbar, Australia: Wollongbar 671-695. Agricultural Institute. USDA (United States Department of Agriculture). (1999). Madhumitha, D., Singh, B. P., & Khan, S. K. (1997). Effect Soil quality test kit guide. Washington, D.C., USA: of major land uses on soil characteristics of alfisols United States Department of Agriculture Soil Quality in Meghalaya. Journal of the Indian Society of Soil Institute. Science, 45, 547-553. Vallejo, V. E., Roldan, F., & Dick, R. P. (2010). Soil enzy- McNeely, J. & Schroth, G. (2006). Agroforestry and bio- matic activities and microbial biomass in an integrated diversity conservation: Traditional practices, present agroforestry chronosequence compared to monocul- dynamics, and lessons for the future. Biodiversity and ture and a native forest of Colombia. Biology and Conservation, 15, 549-554. Fertility of Soils, 46, 577-587. Nair, P. K. R., Buresh, R. J., Mugendi, D. N., & Latt, C. Vaughan, C., Ramírez, O., Herrera, G., & Guries, R. (2007). R. (1999). Nutrient cycling in tropical agroforestry Spatial ecology and conservation of two sloth species systems: Myths and science. In L. E. Buck, J. P. in a cacao landscape in Limón, Costa Rica. Biodiver- Lassoie, & E. C. M. Fernandes (Eds.), Agroforestry sity and Conservation, 16, 2293-2310. in Sustainable Agricultural Systems (pp.1-31). Boca Young, A. (1989). Agroforestry for Soil Conservation. Raton, FL, USA: CRC. Wallingford, UK: CAB International.

Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 887-897, September 2014 897

Use of Trichoderma fungi in spray solutions to reduce Moniliophthora roreri infection of Theobroma cacao fruits in Northeastern Costa Rica

John Seng1, Geovanny Herrera2, Christopher S. Vaughan3 & Michael B. McCoy3 1. Grinnell College, Grinnell, IA, USA; [email protected] 2. Milwaukee Public Museum, Milwaukee, WI 53233-1478, USA; [email protected] 3. Associated Colleges of the Midwest, San Pedro de Montes de Oca, Costa Rica; [email protected], [email protected]

Received 20-iii-2014. Corrected 14-iv-2014. Accepted 15-v-2014.

Abstract: Cacao (Theobroma cacao) is an important cash crop in tropical climates such as that of Latin Rica program Costa Special section (ACM) America. Over the past several decades, the infection of cultivated cacao by Moniliophthora roreri, known commonly as “monilia”, has significantly hindered cacao production in Latin America. Studies have proposed the use of Trichoderma sp. fungi in biocontrol treatments to prevent and reduce monilia infection, yet tests of Trichoderma-containing spray treatments on cacao agroforests have produced mixed results. Researchers and agricultural workers have suggested that addition of soil, fly ash, or other carbon sources to a Trichoderma spray may improve its efficacy in fighting monilia. To test these suggestions, we designed a series of spray mixtures including Trichoderma cultures, soil, and all necessary controls. We applied the spray mixtures to 80 cacao trees (20 trees for each of four resistant-selected clones to monilia) at the FINMAC organic cacao plantation in Pueblo Nuevo de Guacimo, Limón Province, in northeastern Costa Rica in March-April 2013. Five treatments were applied (control, water, water plus sterilized soil, water plus Trichoderma, and water plus sterilized soil plus Trichoderma). Each treatment was applied to four trees of each clone. We monitored the incidence of monilia infection under each spray treatment over the course of 35d. We found that spraying entire cacao trees two times with a mixture containing Trichoderma and sterilized soil significantly reduced the incidence of monilia infection by 11% (p<0.05) in only 35d, as compared to the control. This reduction in loss of cacao pods translates into an increase of plantation mean productivity of 1 500kg dried beans/ha by 198kg/ha up to 1 698kg/ha or by a total increase over the whole 110ha plantation by 21 780kg. We propose that using such an antifungal spray over the whole course of a crop cycle (120 days) would decrease infection incidence even more. Application of this fungal control measure has the potential of revitalizing the production of cacao in the region. Rev. Biol. Trop. 62 (3): 899-907. Epub 2014 September 01.

Key words: Theobroma cacao, biological control, cacao, Costa Rica, monilia, organic fungicide, Trichoderma.

Cultivation of the cacao tree, Theobroma Ramirez, Vaughan, Herrera, & Guries, 2011), cacao (L.), is very important to the economies howler monkeys (Muñoz, Estrada, Naranjo, & and ecosystems of Latin American countries Ochoa, 2006), numerous birds (Reitsma, Par- such as Costa Rica. Cacao seeds, the pri- rish, & McLarney, 2001), foraging ants (Roth, mary ingredient in chocolate, are a valuable Perfecto, & Rathcke, 1994), terrestrial amphib- cash crop which historically has comprised a ians and reptiles (Whitfield et al., 2007) and large amount of Latin American agricultural dung beetles (Harvey, González, & Somarriba, exports (Bamba & Reed, 2004). In addition, 2006). Yet the Latin American cacao industry cacao agroforest systems are thought to provide today is struggling with persistent crop disease. habitat and migration corridors between forest The infection of cacao fruits through- fragments for such diverse species as sloths out Central and South America by the fun- (Vaughan, Ramirez, Herrera, & Guries, 2007; gus Moniliophthora roreri, known commonly

Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 899-907, September 2014 899 as “monilia”, has decimated much of Latin spray: a) on reduction of monilia incidence American cacao production over the past and degree of infection, b) with and without four decades (Hoopen, Rees, Aisa, Stirrup, & addition of sterilized soil as a fungal nutrient or Krauss, 2003; Aime & Phillips-Mora, 2005; adherent, and c) on four farm-produced clonal Phillips-Mora & Wilkinson, 2007). Monilia cacao varieties. infection is extremely concerning to chocolate producers. Throughout Latin America, monilia MATERIALS AND METHODS infection claims some 30 000MT of potential cacao production annually (Bowers, Bailey, Study site: The study was conducted on the Hebbar, Sanogo, & Lumsden, 2001). Addition- organic cacao (Theobroma cacao) plantation at ally, loss of this agroforestry crop as an eco- FINMAC, in Pueblo Nuevo de Villa Franca de nomic option has aided the recent expansion Guacimo, Limón Province, Costa Rica (10°20” of banana and pineapple monocultures, very N - 83°20” W). The FINMAC plantation sits destructive ecologically, as well as, socially. at 40m above sea level (Wieme, 2011) and Recent studies demonstrated the potential is located within the Premontane Wet Forest of the fungus, Trichoderma sp., which thrives Life Zone (Holdridge, 1964). Annual rainfall in the soil of cacao agrosystems and similar averages 6000mm and mean annual tempera- environments (Mpika, Kebe, & N’Guessan, ture is 26ºC (Holdridge, 1964; Vaughan et al., 2011), as a natural antagonist to M. roreri (Ste- 2007; Wieme, 2011; Cornwell, 2012). Canals fanova, Leiva, Larrinaga, & Coronado, 1999; throughout the plantation provide irrigation and Benítez, Rincón, Limón, & Codón, 2004, Nali- drainage. Additional plant species intermixed mova, 2007). Some plantations, including FIN- with the cacao, including Eucalyptus deglupta, MAC, have proposed a passive management Cocos nucifera, Leucaena leucocephala and strategy where diseased pods are cut from trees Musa acuminata, ensure that cacao trees rarely and buried under the leaf litter near the topsoil receive more than 3hrs of direct sunlight per under the assumption that the soil population day. The cacao (Theobroma cacao) agroforest of Trichoderma will kill off the monilia spores within the plantation is at times home to par- as described in Krauss & Soberanis (2002). Yet rots, macaws, two-and-three-toed sloths, ants, the current rate of loss attributed to monilia at termites and wasps. FINMAC (25-30% pod loss/ yr) indicates that simply burying the pods and using semi-resis- Study system: Theobroma cacao at FIN- tant on-farm-produced clones is not enough to MAC is planted with 3m spacing between all combat infection effectively. trees. Six distinct clones, referred to here as Previous research also suggested the pos- clones A, B, C, D, E and F, are cultivated at the sibility of spraying cacao pods, branches, and/ plantation. The land was converted from live- or trunks with Trichoderma cultures as a pre- stock pasture to cacao plantation circa 1982, ventative management strategy to control the and has been organically managed since 1997 spread of monilia. However, Trichoderma nor- (Cornwell, 2012). The FINMAC farm is situ- mally thrives in the soil, meaning that the fruits ated on mostly flat to slightly rolling ground on and branches of a cacao tree do not provide moist soil with moderate-to-high amounts of ideal growth conditions or nutrients. To com- leaf litter. Cornwell (2012) worked at FINMAC pensate for this environmental mismatch in to characterize and quantify soil quality and agricultural applications, some studies have composition. She explained that soil moisture proposed and tested (Krauss & Soberanis, is maintained by subterranean drainage tubes, 2002) the addition of nutrient sources when soil is fertilized with composted coffee husks, applying a Trichoderma spray. and organic herbicides have been applied inter- The objectives of this study were to test mittently. During the Cornwell (2012) study, the temporal impact of Trichoderma fungal the soil on the organic cacao plantation had

900 Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 899-907, September 2014 a mean pH of 5.8. Mean cation exchange a spigot on the plantation. Trich-Aid is a pow- capacity in the agroforest soil is 22.6cmol+/L dered Trichoderma inoculum, manufactured (Cornwell, 2012). For more detailed informa- by Bio-Tech (Guapiles, Limón, Costa Rica). tion regarding soil composition and quality, see Treatments were applied twice throughout the Cornwell (2012). study period: 22 March and 9 April 2013, using a Carpi 18-l manually-operated back- Test subject selection: Twenty trees each pack sprayer. Sprays were delivered at a rate of clones B, C, D and E were selected within a of 1l•min-1 for 60 ±20s per tree, wetting the 100 x 260m block at FINMAC. Clones A and F entire trunk, branches, leaves and fruits from were not included since their fruits had recently the ground up to 2.5m. been harvested at the time of study. Trees were chosen by walking inwards from the begin- Monitoring: Fruits in each experimental ning of each clone-row (from the main farm tree were observed for monilia infection at road), marking every other tree, and skipping twelve time points between 23 March and 26 those which did not have at least four healthy April 2013: March 23 and 25, and April 4, 8, fruits longer than 15cm. For clones B and C, 10, 12, 15, 17, 19, 22, 24 and 26. At each time 20 trees meeting the guidelines were available point the number of fruits greater than 15cm Rica program Costa Special section (ACM) within single clone-rows. For clones D and E, in length was first counted. The number of 10 trees were selected from each of two clone- fruits longer than 15cm with monilia infection rows because fewer trees meeting the guide- was then recorded using a four-stage infection lines were available within each row. Marked scale (Fig. 1): (0) no infection, (1) infection trees were encircled with bright orange twine entering, indicated by irregular discoloration wrapped around adjacent trees to the selected and/or irregular bump, (2) infection consum- tree. Plantation workers were asked not to cut ing fruit, indicated by irregular brown to black or touch fruits in marked trees. patch, and (3) production of fine, thickly layered cream colored spores. Infected fruits Treatments: Five treatments were ran- were marked with brightly colored flagging domly assigned in equal frequency to trees tape. Fruits with stage 3 monilia infections from each clone using the random integer were cut from their trees and thoroughly buried generator in the “R” base package (R Develop- under a layer of leaf litter, and were notated ment Core Team, 2012) so that each treatment as grade three infected fruits in all subse- was applied to four trees per clone. In cases quent observations. Fruits with any degree of where clones were split between two rows, monilia infection at the first time point were treatments were assigned at equal frequencies excluded from analysis. between the rows. Eighteen liters of each treat- Mean incidence of monilia infection was ment were prepared 24hrs prior to application calculated using the cumulative percentage of at room temperature in 18-l plastic screw-top fruits which became infected throughout the tanks as follows: (1) nothing applied; (2) water course of the study. Mean degree of monilia with 700g soil (3) water (4) water with 10g infection was calculated by taking the arithme- Trich-Aid (a Trichoderma inoculum powder); tic mean of degrees of monilia infection, zero to and, (5) water, 10g Trich-Aid, and 700g soil. three, throughout time points, so as to provide Treatments 2, 3, 4 and 5 contained 50ml of a hypothetical mean degree of infection for any Limonoil, an agricultural treatment adhesion time point. Differences between mean inci- solution (www.grupocolono.com). dence and mean degree of infection between Soil was obtained from a nearby cow pas- treatments, clones and time points were esti- ture, finely sifted, and sterilized in the planta- mated by parametric 2-way ANOVA using R tion cacao-drying oven at 250ºC for 25min. software (R Development Core Team, 2012). Water used was non-chlorinated, obtained from A posteriori comparisons between means were

Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 899-907, September 2014 901 performed using Tukey “honestly significant difference” (HSD) tests with the “agricolae” package for R (De Mendiburu, 2013). Graphs were generated with the base R software pack- age (R Development Core Team, 2012).

RESULTS

Impact of spray treatments: Treatment five, “Trichoderma plus soil” significantly reduced mean incidence of monilia infection by 11% and mean infection degree by 0.121 (p<0.05) (Table 1) as compared to the control (treatment one).

Differences between clones: When results are split by clone, treatment five continued to reduce mean incidence (Fig. 2A) and mean degree (Fig. 2B) of monilia infection in clones C, D and E, though the change is only statisti- cally significant in the cases of clone C, for infection incidence, and clone E, for infection degree. In clone B, treatment four signifi- cantly reduced mean incidence and degree of monilia infection.

Infections over time: Examination of mean incidence of monilia infection over time, grouped by treatment (Fig. 3A) revealed that Fig. 1. Guide to monilia infection grades. Top to bottom: fruits under treatment five were significantly characteristic examples of infections of grades one, two, less infected than the control (treatment one) and three. at every time point after 12 March 2013 and

TABLE 1 Monilia infection under each treatment

Treatment1 Mean incidence2,4 Mean degree3,4 1 16.74a 0.206a 2 13.62ab 0.178ab 3 12.66ab 0.158ab 4 9.15ab 0.144bc 5 5.78b 0.085d

1. 1=hard control (nothing applied), 2=water, soil, and Limonoil, 3=water and Limonoil, 4=water, limonoil and Trichoderma, 5=water, soil, Limonoil and Trichoderma. 2. percent of fruits observed displaying any degree of monilia infection. 3. on the 0-3 scale of infection, as described in methods. 4. means with the same letter(s) are not significantly different (Tukey HSD test, p<0.05).

902 Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 899-907, September 2014 Special section (ACM) Costa Rica program Costa Special section (ACM)

Fig. 2A. Mean incidence of monilia infection, grouped by clone and treatment. Vertical lines are +1 Tukey 95% confidence intervals. FINMAC plantation, Pueblo Nuevo de Guácimo, Costa Rica. March-April 2013. Fig. 2B. Mean degree of monilia infection, grouped by clone and treatment. Vertical lines are +1 Tukey 95% confidence intervals. FINMAC plantation, Pueblo Nuevo de Guácimo, Costa Rica. March-April 2013. fruits under treatment four were significantly more likely to become infected by monilia than less infected than the control (treatment one) at those of clones B and E. every time point after 19 April 2013. DISCUSSION Differences between clones over time: Grouping mean incidence of monilia infection Trichoderma sprays reduced monilia by clone over time (Fig. 3B) revealed that fruits infection so greatly that its use to combat of clones C and D were generally significantly monilia pod rot in cacao within the study area

Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 899-907, September 2014 903 Fig. 3A. Mean incidence of monilia infection at each time point, grouping treatments. Verticle lines are +1 Tukey 95% confidence intervals. FINMAC plantation, Pueblo Nuevo de Guácimo, Costa Rica. March-April 2013. Fig. 3B. Mean incidence of monilia infection at each time point, grouping clones. Verticle lines are +1 Tukey 95% confidence intervals. FINMAC plantation, Pueblo Nuevo de Guácimo, Costa Rica. March-April 2013. is supported. Krauss & Soberanis (2002) found provided a preliminary justification for the a 5.2% reduction in mean incidence of monilia addition of soil to Trichoderma sprays as a nat- infection in eastern Peru with numerous peri- ural aid to antifungal activity. In this study we odical Trichoderma sp. sprays over a 4-mo propose that the soil added to some treatment period. Our reduction in monilia mean inci- sprays provided a nutrient source for Tricho- dence of infection by twice (11%) that amount derma, aiding in survival and thus improving was surprising, since our study was only 35d in its ability to fight monilia infection. However, length with only two Trichoderma applications. there are also a number of alternate possibili- Regarding viability of soil as an additive, ties which remain to be tested. For instance, differences between treatments four and five the soil may simply help the fungi stick to the

904 Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 899-907, September 2014 tree and fruit surfaces. It is also possible that tested clones could be used to identify com- additional soil microbes, which were not killed mon genes associated with monilia resistance in the heat-sterilization process, may have had and Trichoderma response, thus aiding in the an effect on monilia infection. process of clone breeding and selection. Dif- In addition to soil, it would be useful to ferences between clones in treatment response test other nutrient sources, such as fly ash as should be used to inform selection of clones in an organic carbon source. Use of ash in sprays order to maximize crop yield. would certainly be cheap but may be diffi- Regarding management implications, cult to prepare and deliver using a backpack though the addition of sterilized soil to a Trich- sprayer. Many classic studies of Trichoderma oderma spray per treatment five did appear to sp. have grown cultures in dedicated fungal improve the antifungal activity of this treat- media such as potato dextrose broth (Hermosa ment, it was not efficient enough for use as a et al., 2000) or malt extract broth (Castle et al., larger scale treatment, simply because the soil 1998). A high-volume solution of Trichoderma tends to clog the backpack sprayer. Further- culture in pre-sterilized potato dextrose or malt more, heating and sifting enough soil to spray extract broth could easily be inoculated, incu- an entire plantation would take a lot of time. bated to grow the fungus to a high concentra- One solution to the backpack sprayer problem Rica program Costa Special section (ACM) tion, and stored indefinitely. The concentrated might be to use motorized backpack sprayers stock could then be diluted and efficiently with better filters. Yet the more viable solu- delivered as needed via manual or motorized tion would be to develop an organic nutrient backpack sprayers. additive which is more easily prepared and Given the variation in treatment response delivered, such as the broths proposed above. between clones, we propose that the genetic The 11% reduction in monilia incidence, variation between different clones has the in only 35d, by treatment five as observed potential to affect their response to spray through the course of the study period is in treatments. Clone B exhibited an unusually itself economically intriguing. Given such an strong response to treatment four and a rela- tive lack of response to treatment five in the improvement, the plantation mean productiv- reduction of monilia infection incidence and ity of 1 500kg dried beans/ha (Cornwell, 2012) degree. The unique response of Clone B to would rise by nearly 200kg/ha to 1 698kg/ treatment as compared to other clones sug- ha. Treating the entire 110ha plantation with gests that genetic variation in cacao can affect Trichoderma and soil would thus increase total its response to Trichoderma-based antifungal mean productivity of dried beans by 21 780kg. spray treatments. It is important to note that current monilia Furthermore, clones B and E tended to losses of cacao pods of 25-30% or more per experience significantly lower incidences of year are common on FINMAC farm. However, monilia infection without Trichoderma appli- even more interesting is to remember this study cation, implying that genetic variation in cacao was only 35d in length, where infection inci- also can influence its resistance to monilia dence decreased at a positive linear rate (Fig. infection. However, genetically resistant clones 3A, 3B). Would infection incidence decrease may not have desirable cacao bean produc- even more over a 4-mo growing period with tion or quality. Studies involving more clones periodical Trichoderma applications? Could should be performed in order to learn more the normal 30% loss be eliminated over this about the variable response of cacao to monilia time period? If so, the economic, ecological and Trichoderma throughout various clones. and social benefits of Trichoderma control of Furthermore, molecular and genetic work monilia, promoting expanded cacao produc- incorporating sequence comparisons between tion, would be nothing short of phenomenal.

Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 899-907, September 2014 905 ACKNOWLEDGMENTS testigo. Esta reducción en la pérdida de frutos de cacao representa un aumento de 198kg/ha de semillas secas sobre Some of the practices carried out in this la producción media de 1 500kg/ha, o un aumento total de study have been tested by Geovanny Herrera 21 780kg en toda la plantación de 110ha. Proponemos que el uso de tal aspersión fungicida a lo largo de todo el ciclo at FINMAC, including using a spray consist- de cultivo (120d) produciría una disminución aún mayor ing of a mixture of Trichoderma with fly ash de la tasa de incidencia de infección. La aplicación de este and soil as additional nutrients. Thanks also to método de control fungicida tiene el potencial de revitalizar Hugo and Hubertien Hermelink and all of FIN- la producción de cacao en la región. MAC for providing us with necessary cacao trees and material resources. Many thanks to Palabras clave: Theobroma cacao, cacao, control biológi- co, Costa Rica, fungicida orgánico, monilia, Trichoderma. Skye Greenler for her help with field work and data collection. Thanks to the ACM for their financial, linguistic, and academic sup- REFERENCES port. Finally, thanks to all of the people of Aime, M. C. & Phillips-Mora, W. (2005). The causal Pueblo Nuevo de Guácimo for welcoming us agents of witches’ broom and frosty pod rot of cacao into their community. (chocolate, Theobroma cacao) form a new lineage of Marasmiaceae. Mycologia, 97, 1012-1022. RESUMEN Bamba, I. & Reed, M. (2004). Monetary policy impacts on cash crop coffee and cocoa using structural vec- Uso de aspersión fungicida de Trichoderma para tor error correction model. Paper presented at the reducir infección de frutos de Theobroma cacao por Meeting of American Agricultural Economists Asso- Moniliophthora roreri en el noreste de Costa Rica. El ciation, Denver, CO, USA. Retrieved from http://age- cacao (Theobroma cacao) es un cultivo comercial impor- consearch.umn.edu/bitstream/20056/1/sp04ba09.pdf tante en los climas tropicales como los de América Latina. Benítez, T., Rincón, A. M., Limón, M. C., & Codón, A. C. A lo largo de las últimas décadas la infección de cacao (2004). Biocontrol mechanisms of Trichoderma stra- cultivado con Moniliophthora roreri, conocida común- ins. International Microbiology, 7, 249-260. mente como “monilia”, ha dificultado la producción del Bowers, J. H., Bailey, B. A., Hebbar, P. K., Sanogo, S. R., cacao en América Latina de manera significativa. Algunos & Lumsden, D.. (2001). The impact of plant diseases estudios han propuesto el uso del hongo Trichoderma sp. on world chocolate production. Plant Health Pro- en tratamientos de control biológico para prevenir y reducir gress. Retrieved from http://www.plantmanagement- la infección por monilia. No obstante, pruebas realizadas network.org/pub/php/review/cacao/ con tratamientos por aspersión que contenían Trichoderma en cultivos de cacao agroforestales produjeron resultados Castle, A., Speranzini, D., Rghei, N., Alm, G., Rinker, D., diversos. Investigadores y trabajadores agrícolas han suge- & Bissett, J. (1998). Morphological and molecular rido que la adición de tierra, cenizas volantes u otras fuentes identification of Trichoderma isolates on North Ame- de carbón a la aspersión de Trichoderma podría mejorar su rican mushroom farms. Applied and Environmental eficacia en la lucha contra la monilia. Para probar la validez Microbiology, 64(1), 133-137. de estas sugerencias, diseñamos una serie de mezclas para Cornwell, E. (2012). Effects of different agricultural la aspersión que incluían cultivos de Trichoderma, tierra systems on soil quality in northern Limón provin- y todos los testigos necesarios. Aplicamos aspersiones a ce, Costa Rica. Semester research project. Field 80 árboles de cacao (20 árboles para cada uno de cuatro Research in the Environment, Social Sciences, and clones seleccionados anteriormente por su resistencia a la Humanities. San Pedro de Montes de Oca, Costa monilia) en la finca de cacao orgánico FINMAC en Pueblo Rica: Associated Colleges of the Midwest. Rev. Biol. Nuevo de Guácimo, provincia de Limón, noreste de Costa Trop. 62(3): 887-897. Rica durante marzo y abril de 2013. Se aplicaron cinco De Mendiburu, F. (2013). Agricolae: Statistical proce- tratamientos (testigo, agua, agua con tierra esterilizada, agua con Trichoderma, y agua con tierra esterilizada y Tri- dures for agricultural research. R package version choderma). Se aplicó cada tratamiento a cuatro árboles de 1.1-4. Retrieved from http://CRAN.R-project.org/ cada clon. Medimos la tasa de incidencia de infección por package=agricolae monilia bajo cada tratamiento por aspersión durante 35d. Harvey, C. A., González J., & Somarriba, E. (2006). Dung La aplicación de dos aspersiones a los árboles completos beetle and terrestrial mammal diversity in forests, con una mezcla de Trichoderma y tierra esterilizada redujo indigenous agroforestry systems and plantain mono- la tasa de incidencia de infección por monilia en 11% cultures in Talamanca, Costa Rica. Biodiversity and (p<0.05) en solo 35d, en comparación con el tratamiento Conservation, 15, 555-585.

906 Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 899-907, September 2014 Hermosa, M. R., Grondona, I., Iturriaga, E. A., Diaz-Min- for Statistical Computing, Vienna, Austria. Retrieved guez, J. M., Castro, C., Monte, E., & Garcia-Acha, I. from http://www.R-project.org (2000). Molecular characterization and identification Ramirez, O., Vaughan, C., Herrera, G., & Guries, R. of biocontrol isolates of Trichoderma spp. Applied (2011). Temporal and spatial resource use by female and Environmental Microbiology, 66, 1890-1898. three-toed sloths and their young in an agricultural Holdridge, L. (1964). Life zone ecology. San José, Costa landscape in Costa Rica. Revista de Biologia Tropi- Rica: Tropical Science Center. cal, 59, 1743-1755. Hoopen, G. M., Rees, R., Aisa, P., Stirrup, T., & Krauss, Reitsma, R., Parrish, J. D., & McLarney, W. (2001). The U. (2003). Population dynamics of epiphytic myco- role of cacao plantations in maintaining forest avian parasites of the genera Clonostachys and Fusarium diversity in southeastern Costa Rica. Agroforestry for the biocontrol of black pod (Phytophthora pal- Systems, 53, 185-193. mivora) and moniliasis (Moniliophthora roreri) on cocoa (Theobroma cacao). Mycological Research, Roth, D. S., Perfecto, I., & Rathcke, B. (1994). The effects 107(5), 587-96. of management systems on ground-foraging ant diversity in Costa Rica. Ecological Applications, 4, Krauss, U. & Soberanis, W. (2002). Effect of fertilization 423-436. and biocontrol application frequency on cocoa pod diseases. Biological Control, 24, 82-89. Stefanova, M., Leiva, A., Larrinaga, L., & Coronado, M. F. Mpika, J., Kebe, I. B., & N’Guessan, F. K. (2011). Isolation (1999). Actividad metabólica de cepas de Trichoder- ma spp para el control de hongos fitopatógenos del and identification of indigenous microorganisms of Rica program Costa Special section (ACM) cocoa farms. In O. Grillo (Ed.), Biodiversity loss in suelo. Revista Facultad de Agronomía, 16, 509-516. a changing planet (pp. 303-318). Rijeka, Croatia: Vaughan, C., Ramirez, O., Herrera, G. & Guries, R. (2007). Intech. Retreived from www.intechopen.com Spatial ecology and conservation of two sloth species Muñoz, D., Estrada, A., Naranjo, E., & Ochoa, S. (2006). in a cacao landscape in Limón, Costa Rica. Biodiver- Foraging ecology of howler monkeys in a cacao sity and Conservation, 16, 2293-2310. (Theobroma cacao) plantation in Comalcalco, Mexi- Whitfield, S. M., Bell, K. E., Philippi, T., Sasa, M., Bola- co. American Journal of Primatology, 68(2), 127-142. ños, F., Chaves, G., Savage, J. M, & Donelly, M. A. Nalimova, M. S. (2007). Introducción y eficacia técnica del (2007). Amphibian and reptile declines over 35 years biocontrol de fitopatógenos con Trichoderma spp. en at La Selva, Costa Rica. Proceedings of the National Cuba. Fitosanidad, 11(3), 75-79. Academy of Sciences, USA, 104, 8352-8356. Phillips-Mora, W. & Wilkinson, M. J. (2007). Frosty pod Wieme, R. (2011). Efficacy of organic herbicides in an of cacao: A disease with a limited geographic range organic cacao (Theobroma cacao) plantation in but unlimited potential for damage. Phytopathology, Costa Rica (Unpublished). Semester research project. 97(12), 1644-1647. Field Research in the Environment, Social Sciences, R Development Core Team. (2012). R: A language and and Humanities. San Pedro de Montes de Oca, Costa environment for statistical computing. R Foundation Rica: Associated Colleges of the Midwest.

Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 899-907, September 2014 907

Ecotourism and primate habituation: Behavioral variation in two groups of white-faced capuchins (Cebus capucinus) from Costa Rica

Shasta E. Webb1 & Michael B. McCoy2 1. Macalester College, 1600 Grand Avenue, St. Paul, MN 55105-1899; [email protected] 2. Associated Colleges of the Midwest, Costa Rican Office, San Pedro de Montes de Oca, Costa Rica; [email protected]

Received 20-iii-2014. Corrected 20-iv-2014. Accepted 15-v-2014.

Abstract: The increase of ecotourism operations within Costa Rica during the last 20yrs has brought more and

more humans into close, direct contact with several wildlife species. One of these species is the white-faced Rica program Costa Special section (ACM) capuchin (Cebus capucinus), highly gregarious, and with exposure over time, willing to come into close vicin- ity of humans and their developments. Such contact has its advantages and disadvantages for the ecotourism industry. We observed white-faced monkeys in order to assess the impact of human presence and development on monkey behavior, with a focus on aggressive, affiliative, and foraging behaviors in Curú Wildlife Refuge (CWR), located in Puntarenas, Costa Rica, and to ascertain the degree of over-habituation of capuchin popula- tions at CWR. Though there exists no discrete behavioral parameters that measure over-habituation, it can be defined as an extreme state of habituation in which non-human primates not only lose fear of humans, but also actively include humans in social interactions or treat them as a food resource. We used instantaneous focal animal and group scan sampling during 8wks in March and April 2012. Two groups (approximately 20-30 individuals each) of capuchins were observed; the first near the tourist development at the Southwestern area of CWR, representing a habituated population that regularly foraged, rested, and groomed in the presence of humans. The second, was observed in the Northeastern area of CWR, did not visit the center of human activity and exhibited fear of humans. The habituated group exhibited significantly fewer instances of threatened behav- ior in response to human presence (p<0.0001) than the non-habituated group, and spent significantly more time eating and foraging (p<0.0001). While the habituated monkeys at CWR may not be over-habituated, they could become that way as development, especially ecotourism, increases. Over-habituation is a problem that affects capuchins in certain ecotourism sites in Costa Rica. It is critical that the consequences of habituation be studied more carefully, primarily in areas where ecotourism operations draw visitors to wildlife habitats. Rev. Biol. Trop. 62 (3): 909-918. Epub 2014 September 01.

Key words: Cebus capucinus, Costa Rica, ecotourism, habituation, over-habituation, white-faced capuchin.

Historically, primatologists have largely already assessed as threatened, endangered, or ignored anthropogenic influences on nonhu- critically endangered. man primate communities. Leading primatolo- To determine potential anthropogenic gists in the field have admitted to previously threats to primates, this paper focuses on viewing humans as “pesky” and “insignificant” habituation—the gradually lessened response aspects of the environment. In recent years to an outside stimulus—and over-habituation, however, primatologists and biological anthro- defined here as the behavioral state in which pologists have shifted their thinking: humans a primate has not only lost fear of humans but are a critical factor in any primate environment also includes humans in social interactions or and must be assessed for potential influence on interacts with humans as food sources. The pri- primate health, behavior, and long-term viabil- mary examination of this topic is framed within ity, especially in cases where the primates are ecotourism, which theoretically attempts to

Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 909-918, September 2014 909 attract visitors to and educate them about pris- interact with its abundant wildlife. This paper tine natural environments while simultaneously examines the potential impacts of the interac- creating low-impact development, providing tions between humans and primates on the economic benefits to the community that lives behavior patterns of white-faced capuchins in with and protects wildlife, as well as promoting an ecotourism setting. peaceful coexistence. Ecotourism in practice is often flawed because it is largely unregulated, MATERIALS AND METHODS but conceptually has the potential for success in wildlife conservation (Buchsbaum, 2004). Study site: Privately owned Curú National This paper represents a case study examin- Wildlife Refuge and Hacienda (CWR) locat- ing two groups of wild white-faced capuchins ed in Puntarenas, Costa Rica (9°47′36.24″N, (Cebus capucinus) that inhabit Curú Wildlife 84°55′21.36″W), is considered a successful Refuge and Hacienda, a small, private eco- endeavor in ecotourism, conservation, and tourist site in Costa Rica. Behavioral differenc- wildlife and land management (Schutt & es between groups of capuchins are expected Vaughan, 1995). Curú Wildlife Refuge covers due to differential contact with humans and approximately 14.94km2 and receives about human development. In particular, behavioral 1 600mm of precipitation annually. Ninety differences regarding fear-related activity (e.g. percent of precipitation falls during the wet monkey fear of humans) and foraging behavior season, from May to November (IMN, 2012). (e.g. if monkeys spend more time occupied by In 1995, approximately 30% of CWR was fear-related behavior, they will spend less time altered environment, including pastures, forest foraging for food) are expected. plantations, living fence rows, and fruit planta- Despite the worldwide reputation of Costa tions (Schutt & Vaughan, 1995). The unaltered Rica as a leader in ecotourism (Honey, 2003; 70% was made up of natural forests, including Buchsbaum, 2004; Gall & Hobby, 2007) recent upland dry, lowland evergreen, mangrove and studies have revealed that ecotourism has been beach-marine forest. Curu Wildlife Refuge far less effective in protecting significant eco- attracts both domestic and foreign tourists and logical resources or benefitting local commu- researchers to its immense biodiversity and nities than originally thought. Stem, Lassoie, relatively accessible location. For the current Lee, Deshler, and Schelhas (2003) focused study, the previously agrarian region of CWR is on Corcovado National Park in Costa Rica referred to as the ‘Finca Side’ and the forested and reported “mixed” findings regarding the portion of CWR as the ‘Ceiba Side’ due to effectiveness of ecotourism as a conservation the presence of ceiba (Ceiba pentandra) trees. and community development tool. The study Additionally, in the agricultural areas, natural resulted in several recommendations regard- riparian forests, palm-lined canals, and living ing ecotourism, including that “ecotourism fencerows formed vegetation corridors (Schutt, would be most effective as a component of a & Vaughan, 1995). For more detailed informa- broader conservation strategy,” referring to the tion about CWR, see Schutt and Vaughan, 1995 need for firmer and more transparent planning and www.curuwildliferefuge.com. and policies on the national level (Stem et al., 2003). It should be noted that in comparison Study species: Our study species, known to other ways in which Costa Ricans finan- commonly as the white-faced capuchin, white- cially sustain themselves (e.g. forest plan- faced monkey, or white-headed monkey, tations/logging, cattle ranching, agriculture, belongs to the taxonomic subfamily Cebinae. etc.) ecotourism remains the most environmen- Cebines, which also include squirrel monkeys tally sustainable option. Despite the abundant (genus Saimiri), are easily recognizable due to issues with ecotourism, tourists continue to their long-time use in the film industry, presence visit Costa Rica for opportunities to see and in the pet trade, and use in street performances

910 Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 909-918, September 2014 as sources of income (Jack, 2010). Because of development and activity at CWR, and rarely their large range and relatively dense popula- entered into close proximity to humans other tions, white-faced capuchins are a well-studied than myself. The Ceiba group visited the center species (e.g. Hall, 2000; Rose, 2000; Williams, of human development two to three times per & Vaughan, 2001; Perry et al., 2003; Vogel, day to eat fruit provided to them by the one 2005; Crofoot et al., 2009). of the owners of CWR. The center of human White-faced capuchins are a “least con- development was the primary area frequented cern” species as determined by the IUCN, by visitors who wanted to see monkeys. I meaning they have been evaluated for threats recorded distance of each monkey group from to long-term survival of the species and appear the farm headquarters (main buildings) at 5min not to be at risk (IUCN, 2012). They have a intervals during observation of each group broad range from Honduras to Ecuador and (Rose, 2000). can survive in a wide variety of habitats—from Observations were daily alternated primary forests in the Neotropics, to highly between the Finca or Ceiba groups over a fragmented secondary or successional forests, period of 8wks in the field (Table 1). Whenever to semi-developed areas of human habitation possible observations took place between 0600

(Hall, 2000; Van Hulle & Vaughan, 2009). and 1600hrs, with breaks for breakfast and Rica program Costa Special section (ACM) Male capuchins range in size from 1.3-4.8kg lunch. Instantaneous sampling (focal and group ( =3.0kg) and females range in size from scans) was employed based on the conclusions 1.4-3.4kg ( =2.3kg). Males typically weigh of Rose (2000). 19.5%-27% more than females. Social struc- ture is multi-male multi-female with one alpha Instantaneous focal animal and group male and alpha female. Social relationships scan sampling: For focal animal sampling, the within troops, which average 16.4 members, adult monkey in clearest view was chosen and are maintained primarily via affiliative behav- observed for 30min, recording his/her specific iors such as grooming, continued proximity, behavior at 60s intervals. A prepared ethogram and coalition formation between preferred part- developed by Rose (2000) was utilized which ners (Jack, 2010). identifies specific behaviors, as well as groups of more general “Activity Classes” (Table 2). Data collection: Two groups of white- Alterations to the ethogram prepared by Rose faced monkeys were observed for this study, (2000) were also made, placing “Alarm Call” each consisting of 20-30 individuals. The Finca under “Activity Class 4: Aggression toward group foraged and spent diurnal hours in forest Other Species.” “Activity Class 5: Eating” patches and corridors in the formerly agri- was also combined with “Activity Class 6: cultural part of CWR. Finca monkeys were Foraging” for analysis because they are closely located farther away from the center of human linked in overall activity budgets, and eating

TABLE 1 Sampling intensity for each monkey group by number of days, and sampling method. Curú Wildlife Refuge and Hacienda. March-April 2012

Number of Days Total Hours Total Hours Total Hours Group of Observationa of Focal Data of Scan Data Collected Finca 22 26.7 20.27 46.97 Ceiba 26 41.2 31.5 72.7 Total 42a 67.9 51.77 119.67 a For the first 8d of observation, behavior was observed in both sides of CWR because the alternate-day system had not been initiated.

Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 909-918, September 2014 911 TABLE 2 Summary of the 10 activity classes used according to an ethogram prepared by Rose (2000)

Class Number Activity Class Title Example Behaviors 1 Affiliation Grooming, playing, huddling, hugging 2 Agonism Chasing, biting, threatening, staring 3 Sexual Behavior Soliciting, mounting, courtship, mating 4 Aggression toward other species Alarm call, threatening, alliance, mob 5 Eat Pick (food), process (food), catch and eat, drink 6 Forage Visual forage, manual forage, explore, sample 7 Movement Move, travel 8 Inactive Resting 9 Scan Vigilance, scan and call, watch group 10 Other Fur rub, nurse, share food, self-groom and foraging behaviors may be easily confused that were typically unvisited by humans other during observation. than the occasional visit by a refuge employee. In the event of losing sight of the initial Binoculars were used to view monkeys high up focal animal, the next adult monkey closest to in the trees and a wristwatch was used to time the researcher was identified and observation observation intervals. continued with that individual for the duration Statistical analyses were performed to of the interval. If the focal animal was moving examine the relationship of monkey behav- but did not appear to be fleeing, it would be ior towards human behavior or proximity to followed. If the focal animal attempted to flee, human activity. We estimated the difference in observations would be discontinued so as not mean distances of each monkey group to farm/ to appear as a predator and potentially increase refuge headquarters with a parametric one- stress on the animal. If the entire troop traveled way ANOVA. We tested the independence of away, observation was stopped, the time of behavioral activity classes with monkey groups troop departure was recorded and behavioral with a contingency table (Sokal & Rohlf, observations would cease. 1995). Additionally, I estimated differences in For group scan collection, rather than mean frequency of occurrence of behavioral recording specific behaviors, the number of activity “Class 4” between monkey groups with individuals exhibiting a behavior from a certain a parametric, one-way ANOVA. Variables were Activity Class were tallied, also at 60s inter- transformed with Log10 if not homogeneous vals. For example, Activity Class 1 is “Affilia- between groups (Sokal & Rohlf, 1995). Statis- tive Behavior,” meaning that if three monkeys tical tests were run with Statgraphics Centurion were observed grooming one another, and XVI (Statpoint Technologies, Inc., 2011). two monkeys playing, “5” would be recorded under “Class 1,” rather than distinguishing RESULTS specific behaviors. A daily mean of 2.84hrs of observations Equipment: A GPS device (GARMIN, were collected. Monkeys in the Ceiba group Etrex Venture, www.garmin.com) was used were observed for more hours because they to record coordinates of the 13 data collection were easier to find (Table 1). sites (five Ceiba and eight Finca sites). The five Ceiba sites were located on tourist trails and Monkey group distribution: Mean the center of human development. The eight distance from the farm/refuge headquarters Finca sites were located on maintenance trails was greater for the Finca group monkeys

912 Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 909-918, September 2014 ( =0.99km; Log10 =-0.025) than for the Ceiba Class 4: Aggressive behavior toward group ( =0.71km; Log10 =-0.056) (F=10.9; humans: Mean log-transformed number of df=1, 130; p=0.0013) (Fig. 1). aggressive behaviors toward other species (e.g. physical displays, alarm calls etc.) per 30min Focal Animal: Frequencies of occurrence interval was significantly greater for the Finca of the nine activity classes (Class 3 [Sexual group (log10 =0.83) than for the Ceiba group Behavior] was omitted for lack of data) were (log =0.07) (F=39.04; df=1, 58; p<0.0001) dependent on monkey group (X2=1665, df=8, 10 (Fig. 3). Additionally, in the Finca group p<0.0001) (Fig. 2). Classes 1, 4, 5 & 6, and 7 478 instances of aggression were recorded, were most relevant to my study objectives. while only 15 instances were recorded in the Class 1: Affiliative behavior between Ceiba group. individuals: Mean number of affiliative behav- iors (e.g. grooming, playing etc.) per 30-min Classes 5 and 6: Eating and Forag- interval was larger for the Ceiba group ( =6.3) ing: Mean frequency of occurrence of eating/ than for the Finca group ( =4.5), but not sig- foraging behaviors (e.g. eat, drink, search in nificantly so (F=0.44; df=1, 50; p=0.51). substrate etc.) per 30min interval was greater Rica program Costa Special section (ACM)

(X 0.001) 11

-9

-29

-49

LOG10 (Distance to Headquarters) to (Distance (km) LOG10 -69 Ceiba Finca Monkey group

Fig. 1. Mean distance to center of human activity (farm/refuge headquarters) for each monkey group for instantaneous focal animal observations. Vertical intervals are +1 least significant 95% confidence intervals. Un-transformed means: (Ceiba 0.705km, Finca 0.990km). Curú Wildlife Refuge and Hacienda, Puntarenas, Costa Rica. March-April 2012.

Class 1 Finca group Class 2 Class 4 Class 5 Class 6 Class 7 Class 8 Ceiba group Class 9 Class 10

Fig. 2. Mosaic plot of frequency of occurrence of each activity class for each monkey group for instantaneous focal animal observations. Behavioral class definitions follow Table 1. Curú Wildlife Refuge and Hacienda, Puntarenas, Costa Rica. March-April 2012.

Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 909-918, September 2014 913 0.91

0.71

0.51

0.31

0.11 LOG10 (Activity Class 4) (Activity LOG10

-0.09 Ceiba side Finca side Location (General)

Fig. 3. Mean frequency of occurrence per 30-min interval of Class 4 (Aggression Toward Other Species) (Log10) between monkey groups for instantaneous focal animal observations. Vertical lines +1 LSD 95% confidence intervals. Curú Wildlife Refuge and Hacienda, Puntarenas, Costa Rica. March-April 2012. for the Ceiba group ( =11.1) than for the Finca behaviors. I (SEW) imagine that with more group ( =4.27) (F=40.4; df=1, 149; p<0.0001). hours, the Ceiba group data would more closely resemble the data collected by Rose (2000). Activity Class 7: Movement: Mean The Finca group, on the other hand, clearly number of movement-related behaviors per maintained fear of humans, which is why 30min interval was greater for the Finca group the monkeys in the Finca acted aggressively ( =6.7) than the Ceiba group ( =5.9), but not toward me in essentially every interval I (SEW) significantly so (F=1.08; df=1, 149; p=0.30). completed. The Finca group is clearly not habituated as evidenced by their high rates of Group scan: The activity budgets of the aggression toward other species, and their gen- Finca group and Ceiba group differed signifi- eral attention to humans. The Finca group data cantly, according to scan data (X2=5798, df=8, greatly differed from data obtained by Rose p<0.0001). (2000), on fully habituated monkeys, showing Finca monkeys not as accustomed to human DISCUSSION presence. This is to be expected as the Ceiba monkeys spent most of their day on commonly The Ceiba monkey group was clearly followed tourist trails or in the center of human more habituated to human presence than the activity. Because the Finca monkeys rarely Finca group. The relative lack of aggression come into contact with humans, it logically and fear response toward other species, specifi- follows that they would view any human as a cally humans, in the Ceiba group suggests that threat. Indeed, initial reaction to habituation in Ceiba monkeys do not fear humans as potential primates includes increased rates of vocaliza- predators nor view them as dangerous competi- tion and increased aggression (Turner, 2005). tors for resources. The activity budget I (SEW) It is unclear what the consequences are gathered from the Ceiba group more closely of increased habituation of the Ceiba monkey matched behavioral data of Rose (2000) than group, as far as monkey health, visitor safety, did the activity budget of the Finca group. The and future conservation. Though little data slight discrepancy between the Ceiba group is available concerning long-term effects of behavior and that of the monkeys Rose (2000) habituating New World primates, some studies studied is most likely a result of personal (Altmann & Muruthi, 1988) of the Great Apes methodological errors such as misidentifying suggest that habituation can have long-term

914 Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 909-918, September 2014 impacts on behavior, particularly feeding and Resort, located just North of Jacó in Puntarenas social structure (Turner, 2005). Because the province, dozens of white-faced capuchins can owner of CWR routinely feeds the habituated be seen from the primary eatery of the hotel monkeys, it is possible that they will become (Van Hulle, & Vaughan, 2009). The monkeys over-habituated in the future. It is also dif- at Punta Leona have become habituated to ficult to control the number of individuals the point that they have absolutely no fear that become habituated, and it is possible for of humans, which has resulted in their steal- habituation to spread from a few individuals to ing of food directly from restaurant tables an entire population (Turner, 2005). Because and closer interactions between humans and white-faced capuchin troops often overlap in monkeys (Van Hulle, & Vaughan, 2009). In home range, the spread of habituation is a dis- Manuel Antonio National Park, white-faced tinct possibility and thus a subject of concern. capuchins are, similarly, totally unafraid of First, overly habituated monkeys can humans, which has resulted in physical aggres- become a threat to human safety and human sion towards humans, as well as theft of health. They can serve as disease vectors objects humans leave unattended (Hall, 2000). (Crofoot et al., 2009) and they can be physi- At CRW, the white-faced capuchins are not as Special section (ACM) Costa Rica program Costa Special section (ACM) cally aggressive (Hall, 2000). While the Ceiba aggressive and fearless as those at Punta Leona monkeys did not necessarily express evidence or Manuel Antonio, but it is possible that with of over-habituation, two accounts of personal continued exposure to humans and human attack by white-faced capuchins were observed. activity, in addition to receiving provisions, In both of these instances, all humans involved they could reach that point. were acting neutrally and not attempting to Westin (2007) suggested with mantled excite the monkeys in any way. Both accounts howlers in Suriname, that “although [howl- involved large male chasing researchers. How- ers] may have been altering their behavioral ever, no humans or monkeys were harmed in patterns in response to tourist presence, they either case. While these cases are only anec- appeared not to be suffering behaviorally… dotal, and therefore not appropriate for statisti- this general monkey ability to remain flexible cal analysis, they are still relevant. Secondly, over-habituated monkeys can and adaptable in the face of habitat change or jeopardize their own health and safety by disturbance (in this case through tourism) mir- increasing exposure to objects associated with rors that found in other populations of howlers, humans and development, such as moving and at other sites” (Naveda et al., 2008). This vehicles, dangerous machines, and power lines. may indeed be the case for white-faced capu- Additionally, if habituated monkeys enter areas chins in Costa Rica, but with very little long- of human development, they can be exposed term, multi-generational data investigating the to human litter such as medical waste and bro- effects of habituation, we cannot know for cer- ken glass, which I (SEW) witnessed at CWR. tain. Research into the topic of human-primate Third, they can alter their behavior, which interaction, particularly in a tourist context, might have negative effects in the long-term should be proactive, rather than reactive. Thus, regarding reproduction and foraging patterns while fully habituating white-faced capuchins (Altmann, & Muruthi, 1988). Finally, they can is not wholly negative (it does, after all, allow become bothersome to the point of causing primatologists an up-close look at social hierar- economic loss. They can cause damage, and chy and social learning), it does pose potential steal food or valuable objects. problems, especially if habituated monkeys Current consequences of over-habituated are not monitored closely. Furthermore, as so white-faced capuchins can already be observed little research is published on the process of at numerous tourist attractions at sites all habituation, because it is typically viewed as around Costa Rica. At Punta Leona Hotel and a “means to an end,” it is becoming critically

Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 909-918, September 2014 915 important to investigate the effects of habitua- Though the white-faced capuchins in tion (Turner, 2005). Costa Rica are not as dangerous and threaten- Over-habituation poses a problem for con- ing as the rhesus monkeys in Northern India, servation. Ideally, both monkeys and visitors translocation of certain populations is a viable on protected lands remain safe. At CWR, the option. With extensive protected land, adapt- monkeys and humans appear to be able to able white-faced capuchins could be relocated inhabit the same area safely, which may be to more isolated areas. While this would not a result of the owner providing them with a undo more subtle consequences of over-habitu- consistent food source. However, while in the ation, it could prevent more serious, long-term present moment CWR seems to handle habitu- consequences. Additionally, stricter rules limit- ated monkeys safely and well, it is important to ing accidental or deliberate human provision- remain vigilant in achieving conservation goals ing would help prevent the negative effects of that primarily benefit the capuchin populations. over-habituation. Perhaps the best example of primate man- In future studies in CWR, it would be agement in relation to exposure to humans informative to focus on the Ceiba monkey comes from India, where large populations group in order to more clearly assess potential over-habituation. Because CWR is an ecotour- of rhesus macaques (Macaca mulatta) have ist site, it draws humans and primates into the become a pressing issue in many urban areas. same physical space. This overlap leads to The sprawl of human development has over- habituation of the non-human primates, and in lapped with macaque populations for an extend- the long term could lead to over-habituation. ed period of time, becoming not only pests, but Currently, CWR is seen as a successful site also actual dangerous threats. Rhesus monkeys of conservation, but data suggests that in the in certain urban areas of India have become future, primates could become over-habituat- hyper-aggressive, biting and injuring dozens ed without stricter mitigation of human con- of people each day (Ekwal, Yahya, & Malik, tact. The application of ethnoprimatology—a 2002). In order to lessen the negative effects of field that combines quantitative assessment monkey-human interactions, 1 300 individual of primate behavior with qualitative assess- monkeys were translocated to a protected wild- ment of how humans perceive the animals and life area far away from human development. their conservation—is critical for CWR. Even A post-translocated study showed that the though white-faced capuchins are recognized monkeys had adapted to the new environment as “non-threatened” this is not necessarily and seemed to exhibit no negative responses accurate. Anthropogenic influences can lead to (Ekwal et al., 2002). change in primates over time, even in species In Kenya, where semi-provisioned that are officially described as “least concern” baboons near tourist lodges have become over- by the IUCN. habituated to the point of being viewed as pests, removal of animals has been less suc- ACKNOWLEDGMENTS cessful (Altmann, & Muruthi, 1988). When part of a population had been eradicated from Many thanks to Scott Legge for advice, an area of human development, another popu- support, and editing. Thanks go to Chris lation typically relocated to gain access to the Vaughan for providing useful feedback on my semi-provisioned habitat. Thus, the only sure proposal for the project and to ACM for assis- way to limit the conflict between the humans tance in facilitating field research. We thank and primates is to limit the offending human the Schutt family at Curu Wildlife Refuge behavior, rather than eradicate the primates and Hacienda for providing lodging, and this (Altmann, & Muruthi, 1988). research site.

916 Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 909-918, September 2014 RESUMEN Buchsbaum, B. D. (2004). Ecotourism and Sustainable Development in Costa Rica. (Unpublished major Ecoturismo y habituación de primates: variación paper). Virginia Polytechnic Institute and State Uni- en comportamiento de dos grupos de monos carablanca versity, Blacksburg, VA., USA. Retrieved from http:// (Cebus capucinus) en Costa Rica. El aumento de activida- scholar.lib.vt.edu/theses/available/etd-05052004- des ecoturísticas en Costa Rica durante los últimos 20 años 171907/unrestricted/EcotourismCostRica.pdf ha ocasionado que más y más personas entren en contacto Crofoot, M. C., Norton, T. M., Lessnau, R. G., Viner, T. C., directo y cercano con varias especies de vida silvestre. Una Chen, T. C., Mazzaro, L. M., & Yabsley, M. (2009). de estas especies es el mono carablanca (Cebus capuci- Field anesthesia and health assessment of free- nus), que es muy gregario y, al pasar el tiempo, tiende a ranging Cebus capucinus in Panama. International acercarse a los humanos y sus instalaciones. Tal contacto Journal of Primatology, 30(1), 125-141. tiene ventajas y desventajas para la industria del ecoturis- mo. Observamos monos carablanca con el fin de evaluar el Ekwal, I., Yahya, H. S. A., & Malik, I. (2002). A successful impacto de la presencia humana y la infraestructura en el mass translocation of commensal rhesus monkeys comportamiento de estos animales, enfocando los compor- Macaca mulatta in Vrindaban, India. Oryx, 36, 87-93. tamientos de agresión, acicalamiento y forrajeo para deter- Gall, T. L. & Hobby, J. M. (Eds.). (2007). Worldmark minar el grado de habituación excesiva de las poblaciones encyclopedia of the nations (pp. 173-360). Detroit, de monos capuchinos en el Refugio de Vida Silvestre Curú, MI, USA: Gale. Puntarenas, Costa Rica. Aunque no existen parámetros Hall, T. (2000). The effects of tourism on white-faced específicos de comportamiento para medir la habituación capuchins (Cebus capucinus) in Manuel Antonio Rica program Costa Special section (ACM) excesiva, se puede definir esta condición como un estado National Park, Puntarenas, Costa Rica. (Unpublished en que los primates no humanos no solamente pierdan undergraduate thesis). Colorado College, Colorado su temor a los humanos, sino que también incluyen a las Springs, Colorado, USA. personas activamente en sus interacciones sociales o las Honey, M. (2003). Giving a grade to Costa Rica’s green consideran como una fuente de alimentación. Los monos tourism. NACLA Report on the Americas, 36(6), 39. carablanca fueron observados a través del muestreo ins- tantáneo focal y por escaneo grupal durante ocho semanas IMN (Instituto Meterológico Nacional de Costa Rica). de marzo y abril de 2012. Dos grupos (aproximadamente (2012). Datos climáticos de Puntarenas. Retrieved 20-30 individuos cada uno) fueron observados; uno cerca from http://www.imn.ac.cr/imn/MainAdmin.aspx?__ del desarrollo turístico en la parte suroeste del refugio, que EVENTTARGET =ClimaCiudad&CIUDAD=6 representó una población habituada que de manera regular IUCN (International Union for Conservation of Nature). comía, descansaba y se acicalaba en presencia de huma- (2012). IUCN Red List of Threatened Species. Ver- nos; el segundo grupo, que se observó en el noreste del sion 2012.2. Gland, Switzerland: IUCN. refugio, no visitaba el centro de actividad humana y mos- Jack, K. M. (2010). The Cebines: Toward an explanation of traba temor hacia los humanos. El grupo habituado mostró variable social structure. In C. Campbell, A. Fuentes, significativamente menos ocasiones de comportamiento K. MacKinnon, S. Bearder, & R. Stumpf (Eds.), Pri- amenazante ante la presencia humana (p<0.0001) y empleó mates in perspective (pp. 107-123). Cary, NC, USA: más tiempo comiendo y buscando alimento (p<0.0001). Oxford University. Aunque el grupo de habituados en Curú probablemente no tenga habituación excesiva, esta condición podría surgir Naveda, A., de Thoisy, B., Richard-Hansen, C., Torres, D. más adelante, especialmente si aumenta el desarrollo, y A., Salas, L., Wallance, R., Chalukian, S., & de Bus- en particular el ecoturismo. La habituación excesiva es tos, S. (2008). Tapirus terrestris. In IUCN, IUCN Red un problema que afecta a los monos carablanca en ciertos List of Threatened Species. Version 2011.2. Retrieved sitios secoturísticos de Costa Rica. Es de suma importancia from http://www.iucnredlist.org/details/21474/0 que las consecuencias de la habituación excesiva se estu- Perry, S., Baker, M., Fedigan, L., Gros-Louis, J., Jack, K., dien más a fondo, principalmente en las áreas donde las MacKinnon, K. C., Manson, J. H., Panger, M., Pyle, actividades ecoturísticas atraen visitantes a los hábitats de K., & Rose, L. (2003). Social conventions in wild la vida silvestre. white-faced capuchin monkeys: Evidence for tradi- tions in a Neotropical primate. Current Anthropology, Palabras clave: Cebus capucinus, Costa Rica, ecoturismo, 44(2), 241-268. habituación, habituación excesiva, mono carablanca. Rose, L. (2000). Behavioral sampling in the field: Con- tinuous focal versus focal interval sampling. Beha- REFERENCES viour, 137(2), 153-180. Schutt, A. & Vaughan, C. (1995). Incorporating wildlife Altmann, J. & Muruthi, P. (1988). Differences in daily life into development: The case of the Curú Wildlife between semiprovisioned and wild-feeding baboons. Refuge and Farm, Costa Rica. In J. Bissonette & P. American Journal of Primatology, 15, 213-221. Krausmann (Eds.), Integrating people and wildlife

Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 909-918, September 2014 917 for a sustainable future (pp. 250-254). Bethesda, MD, Leona private wildlife refuge, Costa Rica. Revista de USA: Wildlife Society. Biologia Tropical, 57(1-2), 441-449. Sokal, R. R. & Rohlf, F. J. (1995). Biometry: The principles Vogel, E. R. (2005). Rank differences in energy intake rates and practice of statistics in biological research. NY, in white-faced capuchin monkeys, Cebus capucinus: USA: W. H. Freeman and Co. The effects of contest competition. Behavioral Ecolo- Statpoint Technologies, Inc. (2011). Statgraphics Centurion gy and Sociobiology, 58(4), 333-344. [computer software]. Available from http://www. Westin, J. L. (2007). Effects of ecotourism on the beha- statgraphics.com vior and health of red howler monkeys (Alouatta Stem, C. J., Lassoie, J. P., Lee, D. R., Deshler, D. D., & seniculus) in Suriname. (Unpublished Ph.D. disserta- Schelhas, J. W. (2003). Community participation in tion). University of Michigan, Ann Arbor, MI, USA. ecotourism benefits: The link to conservation prac- tices and perspectives. Society Natural Resources, Retrieved from http://deepblue.lib.umich.edu/handle 16(5), 387-413. /2027.42/57667?show=full Turner, T. R. (2005). Biological anthropology and ethics: Williams, H. E. & Vaughan, C. (2001). White-faced From repatriation to genetic identity. Albany, NY, monkey (Cebus capucinus) ecology and management USA: State University of New York. in Neotropical agricultural landscapes during the Van Hulle, M. & Vaughan, C. (2009). The effect of human dry season. Revista de Biología Tropical, 49(3-4), development on mammal populations of the Punta 1199-1206.

918 Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 909-918, September 2014 Species richness and relative species abundance of Nymphalidae (Lepidoptera) in three forests with different perturbations in the North-Central Caribbean of Costa Rica

Carolyn Stephen1 & Ragde Sánchez2 1. Knox College, Galesburg, IL 61401, USA; [email protected] 2. Asociación Theria para la Investigación y Conservación, San José, Costa Rica; [email protected]

Received 20-iii-2014. Corrected 30-iv-2014. Accepted 26-v-2014.

Abstract: Measurements of species richness and species abundance can have important implications for regulations and conservation. This study investigated species richness and abundance of butterflies in the fam- Rica program Costa Special section (ACM) ily Nymphalidae at undisturbed, and disturbed habitats in Tirimbina Biological Reserve and Nogal Private Reserve, Sarapiquí, Costa Rica. Traps baited with rotten banana were placed in the canopy and the understory of three habitats: within mature forest, at a river/forest border, and at a banana plantation/forest border. In total, 71 species and 487 individuals were caught and identified during May and June 2011 and May 2013. Species richness and species abundance were found to increase significantly at perturbed habitats (p<0.0001, p<0.0001, respectively). The edge effect, in which species richness and abundance increase due to greater complementary resources from different habitats, could be one possible explanation for increased species richness and abun- dance. Rev. Biol. Trop. 62 (3): 919-928. Epub 2014 September 01.

Key words: butterflies, disturbance effects, edge effect, nymphalidae, species abundance, species diversity, spe- cies richness, Chiquita Nature and Community Project, Nogal Private Reserve.

Biological diversity has become increas- disturbance to ecosystems (Erhardt, 1985; ingly important in the realm of conservation Fleishman, Thomson, Nally, Murphy, & Fay, biology. Changes in species richness and spe- 2005; Bobo, Waltert, Fermon, Njokagbor, & cies relative abundance due to disruptions in Mühlengber, 2006; Pöyry et al., 2006; Leidner, continuous habitats have been used in multiple Haddad, & Lovejoy, 2010). Habitat loss is studies to suggest or evaluate management thought to be the greatest cause for insect decisions on which conservation efforts are extinction, and their populations can be easily needed to maintain the survival of a species or influenced by anthropogenic forest fragmenta- a community of species (Horner-Devine, Daily, tion (Hogue, 1993; Leidner et al., 2010). Bar- Ehrlich, & Boggs, 2003; Uehara-Prado, Brown, low, Overal, Araujo, Gardner, & Peres (2007) & Freitas, 2007; Sjödin, Bengtsson, & Ekbom, found that butterfly species richness in Brazil 2008; Hjältén et al., 2012). tended to be greater in primary forest than on Insects, and in particular butterflies, have Eucalyptus plantations, but that relative species contributed greatly to understanding tropical abundance of the species existing there tended biodiversity (DeVries, Alexander, Chacon, & to be greater on the Eucalyptus plantations. Fordyce, 2012). Although plentiful, insects Uehara-Pradoet al. (2007) compared primary are still incredibly prone to great popula- forest and fragmented forest in the Brazilian tion increases and decreases, and have been Atlantic, finding that butterfly species richness used frequently as important indicators of increased at fragmented sites while relative

Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 919-928, September 2014 919 abundance did not change, suggesting that and 84°7’ W) is at an elevation of 180-220m forest fragmentation and disturbance does not above sea level and is made up of 345ha of negatively affect species diversity. Horner- lowland rainforest. Mean annual temperature Devine et al. (2003) found that coffee farms is 25.3°C with a mean high of 30.2°C and a within 1-2.5km of a forest reserve had greater mean low of 20.2°C. Mean annual precipitation butterfly species richness and relative species is 3777mm. Landscape inclines generally fall abundance than both, farms 6-9km away and between 10-25% with rolling hills, although within the forest reserve itself, indicating that closer to creeks and the river, slopes can reach nearby forest may have a positive impact on 60%. Soils are of volcanic origin. The Sara- butterfly species richness and relative species piquí River forms the Western border of the abundance. It also suggests that forest edges reserve. Tirimbina falls within two Holdridge may experience increases in species richness life zones, Humid Tropical Pre-Montane For- and abundance. est and transition to Basal and Humid Tropical This study focuses on the butterfly family Forest. Primary forest makes up 85% of TBR Nymphalidae, which contains more species while the rest is secondary forest (Tirimbina than any other butterfly family in the world, Biological Reserve, 2010b). except the family Lycaenidae. Nymphalidae The data of the banana plantation/forest species can be found on every continent except border was part of a long term study executed Antarctica, although they are most diverse in by Chiquita Nature and Community Project at Neotropical regions (DeVries, 1987). More Nogal Private Reserve and forest fragments specifically, this study looked at fruit-feeding around it. This fragment is located close to nymphalids, a feeding guild in Nymphalidae, Puerto Viejo, Sarapiquí at 40-60m elevation. which feed on rotten fruit juices (DeVries et Mean annual precipitation falls between 3 al., 2012). The objectives of this study were to 712-4 000mm. The mean annual temperature compare species richness and relative species is 25.8°C, and the vegetation corresponds to abundance in the butterfly family Nymphalidae Holdridge Life Zone of Humid Tropical Forest in both the canopy and the understory between (Barquero Villalobos, 2009). Although close (a) mature forest habitat and two disturbance to Puerto Viejo, the surrounding area of this habitats: (b) river/forest (a natural disturbance) plantation is rural with fragmented forest. In and (c) banana plantation/forest (an anthropo- general, banana and pineapple farms are espe- genic disturbance). cially common in Sarapiquí (Municipalidad de Sarapiquí-Secretaria Técnica de Gobierno MATERIALS AND METHODS Digital, 2012a).

Study site: This study took place primar- Data collection: In TBR, data was taken ily in Tirimbina Biological Reserve (referred from a river/forest border along the Río Sara- to as TBR in this study), located in the sec- piquí and from within the forest itself. In TBR, ond district of La Virgen, Sarapiquí, Heredia data was taken 5d/wk for 3weeks and 4d the Province in the North central-Caribbean area 4th week between 2 to 31 May. Data used from of Costa Rica. La Virgen has a population of the banana plantation/forest border were taken 8 715 people, and is a rural area (Municipalidad 3-7 May and 31 May-5 June, 2011 by staff of de Sarapiquí-Secretaria Técnica de Gobierno Chiquita Nature and Community Projects as Digital, 2012b). The TBR (345ha) is located part of their long-term butterfly survey. in the Sarapiquí River Basin of approximately Traps were made out of mosquito netting 1 923km2 in size. It is bordered by other pri- held in a cylindrical shape with an open bottom vate properties that altogether create 600ha and closed top approximately 30cm in diam- of continuous forest cover (Tirimbina Bio- eter and 1m in height as described in DeVries logical Reserve, 2010a). The TBR (10°24’ N (1987). Between 50 and 100mL of rotten

920 Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 919-928, September 2014 banana were placed in a small plastic cup on a A species accumulation curve was created platform hanging about 5-8cm from the open to estimate the proportion of butterflies identi- bottom of the trap. Banana was mashed and left fied to the actual number of species in the popu- to sit about 1d before baiting. Traps were re- lation. Relationship between daily precipitation baited with banana only as necessary through- data versus daily captures of number of species out the week. Traps were checked once/d and number of individuals was estimated with (trap-day). All butterflies captured during a simple regression carried out with Statgraphics trap-day were counted by number of individu- XV.I (Statpoint Technologies, Inc. 2011). A als of each species and identified using DeVries regression was not done for the banana planta- (1987). Collected specimens were given to the tion/forest habitat due to differences in data Instituto Nacional de Biodiversidad (INBio), collection on precipitation A “square-root-Y” Santo Domingo de Heredia, Costa Rica. Data model (best fit) was used in the mature forest on precipitation levels each day were obtained habitat and a “double-squared” model (best fit) from TBR. Sampling from the banana planta- was used in the river/forest border habitat for tion study was conducted in a similar manner. number of species. A “square root-Y” model Twenty traps were set up at the river/forest (best fit) was used in the forest habitat and a “squared-Y” model (best fit) in the river/for- Rica program Costa Special section (ACM) border throughout the entire study. For the first est border habitat for number of individuals. week of the study, 30 traps within the forest Simper Analysis was used to evaluate changes were used from another ongoing study. For the in species composition with the PAST program next 3wks, 20 different traps were set up and (Hammer, Harper, & Ryan, 2001). used to collect forest data. In each habitat, traps were located at 10 or 15 paired points with one trap in the canopy and one in the understory. RESULTS Traps in both perturbed habitats were set up inside the forest near the border, though not In total, 71 species and 487 individuals were caught and identified (Table 1). In general necessarily right on the tree line. the three most abundant species were Cithae- Differences in: a) mean number of spe- rias pireta pireta, Hamadryas laodamia sau- cies captured per trap-day and b) mean num- rites, and Caligo brasiliensis sulanus. When ber of individuals per species per trap-day comparing the five most abundant species in were estimated between canopy and under- each habitat, only two species were found in story both within and between habitats with more than one habitat: Catonephele orites and two-way parametric ANOVA. A posteriori Colobura dirce dirce (Table 2). At the genus comparisons between means were tested with level, only three were found in two habitats: Scheffe intervals. Parametric test assumptions Hamadryas (forest and river), Catonephele were carried out before executing ANOVAS (forest and banana plantation), and Colobura and procedures followed Sokal and Rohlf (forest and banana plantation). (1995). Statistical analyses were carried out A total of 36 species were only found in with Statgraphics XV.I software (Statpoint one habitat: 8 in mature forest habitat, 21 in the Technologies, Inc., 2011). river/forest border habitat and 7 in the banana Brillouin diversity indexes and Brillouin plantation/forest border habitat (Table 3). evenness indexes were calculated using Micro- For 20 of those species, only one individual soft Excel 2007 Diversity Addin. Uneven sam- was found. ple sizes were accounted for by calculating the diversity index for each trap-day and check- Comparison of number of species: Mean ing for significance through an ANOVA test number of species captured per trap-day was with Statgraphics XVI (Statpoint Technolo- greatest in the banana plantation/forest border gies, Inc., 2011). habitat ( =0.74) and least in mature forest

Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 919-928, September 2014 921 TABLE 1 Total account of number of individuals and species identified in each habitat by canopy and understory. Note that these numbers do not account for unequal sample sizes

No. Individuals No. Species Habitat Canopy Understory Total Canopy Understory Total Banana/Forest 76 99 175 23 15 32 River/Forest 163 38 201 42 20 52 Forest 69 42 111 30 16 40 Total 308 179 487 56 31 71

TABLE 2 Comparison of top five most abundant individuals in each habitat

Most Common Species by Habitat Species name No of ind* No of ind/trap-day % of total no of ind* Mature forest Hamadryas arinome ariensis 13 0.0306 11.712 Tigridia acesta acesta 10 0.0235 9.009 Nessaea aglaura aglaura 8 0.0188 7.207 Catonephele orites 8 0.0188 7.207 Dulcedo polita 6 0.0141 5.405 Colobura dirce dirce 6 0.0141 5.405 River/forest Hamadryas laodamia saurites 22 0.0579 10.945 Memphis xenocles xenocles 17 0.0447 8.458 Prepona laertes demodice 15 0.0395 7.463 Fountainea halice chrysophana 15 0.0395 7.463 Historis odius odius 13 0.0342 6.468 Banana/forest Cithaerias pireta pireta 43 0.215 24.571 Caligo brasiliensis sulanus 18 0.085 10.286 Catonephele orites 10 0.05 5.714 Caligo atreus dionysos 10 0.05 5.714 Colobura dirce dirce 8 0.04 4.571

*Number of Individuals. habitat ( =0.254) with the river/forest border habitat. Mean species number was greater habitat as an intermediary ( =0.511) (F=29.1; in canopy than understory in mature forest df=2, 997; p<0.0001). Overall, mean species habitat, but not significantly so. The lack of number captured per trap-day was greater parallel trends in figure 1 indicates that habitat in the canopy ( =0.626) compared to the plays a strong impact on changing differences understory ( =0.379) when observations of between mean number of species in the canopy all habitats were combined (F=24.8; df=1, and understory. 994; p<0.0001). However, this trend was not the same for each individual habitat (F=16.07; Species accumulation curve: The num- df=2, 994; p<0.0001, Fig. 1).The river/for- ber of species accumulated in mature forest est border habitat had the greatest difference appeared to begin to plateau towards the end between mean number of species between the of the last sampling period, indicating that canopy and the understory and no difference this study may have identified the majority of existed for the banana plantation/forest border Nymphalidae species in that habitat. However,

922 Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 919-928, September 2014 TABLE 3 Number of individuals of each species found in only one habitat.

Species found in only one habitat Species Name No of Ind* Forest Callicore astarte patelina 2 Consul panariste jansoni 1 Dulcedo polita 6 Hamadryas amphinome mexicana 1 Morpho deidamia polybaptus 1 Morpho menelaus amathonte 1 Opsiphanes invirae cuspidatus 1 Opsiphanes quiteria panamensis 1 Total Number of Species 8 River/forest border Caligo ilioneus oberon 5

Callicore texa titania 1 Rica program Costa Special section (ACM) Chloreuptychia arnaca 1 Cissia terrestris 1 Consul fabius cecrops 2 Ectima erycinoides erycinoides 6 Fountainea halice chrysophana 15 Fountainea ryphea ryphea 1 Hamadryas feronia farinulenta 2 Memphis arginussa eubaena 2 Memphis elara 1 Memphis glauce centralis 5 Memphis lyceus 1 Memphis morvus boisduvali 1 Memphis oenomais 4 Opsiphanes tamarindi tamarindi 1 Pareuptychia metaleuca 1 Pareuptychia ocirrhoe ocirrhoe 1 Siderone galanthis galanthis 1 Temenis laothoe agatha 6 Total Number of Species 21 Banana plantation/forest border Agrias amydon philatelica 1 Cissia pseudoconfusa 3 Eryphanis lycomedon 1 Megeuptychia antonoe 3 Memphis proserpina proserpina 2 Myscelia leucocyana smalli 2 Zaretis itys itys 1 Total Number of Species 7

*Number of Individuals.

Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 919-928, September 2014 923 1 Canopy 0.8 Understory

0.6

0.4

0.2 Mean number of species

0 Forest River / Forest Banana / Forest

Habitat

Fig. 1. Mean number of species/trap day between the canopy and the understory across habitats. Vertical lines are ±1 Scheffe 95.0% confidence intervals. Two-way ANOVA with significant interaction (F=16.1; df=2, 994; p<0.0001). in both the river/forest habitat and the banana although the banana plantation/forest border plantation/forest habitat, the number of iden- habitat strayed from this pattern when looking tified species appeared to continue sloping at each habitat (F=19.2; df=2, 994; p<0.0001, upwards, indicating that there may still have Fig. 3). The river/forest border was the only been several unidentified species, leaving the habitat with a significant difference between final data incomplete. the mean number of individuals in canopy and understory. The lack of parallel trends in figure Comparison of number of individuals: 3 indicates that habitat plays a strong impact on Mean number of individuals was greatest in changing differences between mean number of the banana plantation/forest border habitat individuals in the canopy and the understory. ( =0.875) and least in the mature forest habitat ( =0.263) with the river/forest border habitat Brillouin and Brillouin Evenness Diver- as an intermediary ( =0.534) (F=32.3; df=2, sity indices: The Brillouin Diversity Index 997; p<0.0001, Fig. 2). Overall, mean number indicated significantly different mean indices of individuals captured per trap-day was great- between habitats. The banana plantation/forest er in the canopy ( =0.650) than the understory habitat had the greatest diversity ( =0.082), ( =0.467) when observations of all habitats the mature forest habitat had the smallest diver- were combined (F=9.93; df=1, 994; p=0.0016), sity ( =0.013), and the river/forest border

1

0.8

0.6

0.4

0.2

Mean number of individuals 0 Forest River / Forest Banana / Forest Habitat Fig. 2. Mean number of individuals/trap day between each habitat. Vertical lines are ±1 Scheffe 95.0% confidence intervals.

924 Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 919-928, September 2014 1.2 Canopy 1 Understory

0.8

0.6

0.4

0.2 Mean number of individuals 0 Forest River / Forest Banana / Forest

Habitat

Fig. 3. Mean number of individuals/trap day between the canopy and the understory across habitats. Vertical lines are ±1 Scheffe 95.0% confidence intervals. Two-way ANOVA with significant interaction (F=19.2; df=2, 994; p<0.0001). Special section (ACM) Costa Rica program Costa Special section (ACM)

DISCUSSION was intermediary ( =0.055). The Brillouin Evenness Index had similar results; the banana Species richness and species abundance plantation/forest habitat had the greatest index increased in the habitats of disturbance, with ( =0.18), the river/forest border as interme- the greatest numbers in the banana plantation/ diary ( =0.12), and the mature forest habitat forest border habitat. Brillouin Diversity Index with the smallest index ( =0.35). and Brillouin Evenness Index indicated greater Simper Analysis revealed that Cithaerias butterfly diversity in habitats of disturbance as pireta pireta contributed the most to dissimilar- well. Species richness and abundance tended to ity between groups (% Contribution=7.769%) be greater in the canopy than in the understory. with a relatively great abundance in the banana Previous studies have indicated that borders plantation/forest border habitat (mean abun- between two different kinds of habitat can dance=0.215) and relatively smaller abundanc- attract or repel different butterfly species, creat- es in the river/forest border habitat (mean ing something called “edge effects”. According abundance=0) and in the mature forest habi- to Ries, Sisk, and Karban (2008), butterfly tat (mean abundance=0.00941). Hamadryas density tends to increase at the border between laodamia saurites (Contribution %=3.094%), two habitats when there is a strong difference Caligo brasiliensis sulanus (Contribution between habitat quality and distribution of nec- %=2.87%), and Nessaea aglaura aglaura essary or complementary resources. DeVries (Contribution %=2.858%) also had greater (1987) attributed this increased species richness contributions to dissimilarity. and relative species abundance at forest edges to similarities between canopies and forest Comparison with captures versus pre- edges, such as more light and a more dynamic cipitation: In the mature forest habitat, the and diverse physical environment. Tree gaps number of species captured/day decreased sig- within continuous rain forest may also experi- nificantly with increased daily precipitation ence canopy butterfly species near the ground (F=8.63; df=1, 17; p=0.0092; R2=33.68%). due to changes in vegetative structure (Pardon- Number of species captured/day in the river/ net, Beck, Milberg, & Bergman, 2013), reaf- forest border habitat also decreased with firming that areas of disturbance may contain increased precipitation, but not significantly similar characteristics to the canopy, and thus (F=1.17; df=1, 17; p=0.2955; R2=6.414%). attract more of those species. Horner-Devine et

Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 919-928, September 2014 925 al. (2003) hypothesized that increases near bor- to flight (Raupp, 2006). Precipitation levels ders could be due to increased microclimates could account for high levels of variance in that include necessary host plants and food, as samples. It also suggests that species abundance well as increased space for movement. is affected by more than just habitat. Of the 71 species identified, 36 (50.7%) of Limitations to my study may include error those species were found in only one habitat. generated from temporal and location differ- Of the five most abundant species found in ences in data collection. Data from the banana each habitat, only two in total were found pres- plantation/border habitat came from a study ent in more than one habitat. The top contribu- site 20km from TBR and was collected 2yrs tors to dissimilarity according to the Simper before the data collected for the other two habi- Analysis were also all species within the five tats. In the first week of data collection in the most abundant in each habitat. This suggests mature forest habitat, traps were set up in dif- that each habitat provides key services to spe- ferent locations than in weeks 2-4 due to a lack cific species and has strong effects on species of available traps, creating more chance for composition, even though each of these habi- error. This study was also conducted in May, tats includes primary forest and all are found in during which the dry season usually transitions the same region. Pardonnet et al. (2013) found into the rainy season, which may have affected that tree gaps in Peruvian rain forests caused changes in butterfly species and abundance. changes in butterfly assemblages when com- Future research may want to account for pared to adjacent nearby understory, indicating changes in precipitation levels and changes in that species composition may change within species composition over increased time peri- the same continuous forest should vegetative ods. May tends to be the transition between the structure vary, as it does at habitat borders. rainy season and the dry season in Sarapiquí, Increased precipitation negatively affected and there might be differences between rainy the number of species and individuals sig- season data and dry season that are worth nificantly in mature forest habitat. Although not comparing, especially because commonality of significantly affected, number of species and some species seemed to change from the begin- individuals did tend to decrease with increased ning of the month to the end. Therefore, results rainfall in the river/forest border habitat as well. may vary depending on the time of year. As This could be because rainfall is not conducive indicated by the increasing curves in figure 4, a

60

50

40

30

20 Forest Number of species River / Forest 10 Banana / Forest

0 3 May 5 May 7 May 2 June 4 June 11 May 13 May 15 May 17 May 19 May 21 May 23 May 25 May 27 May 29 May 31 May

Days

Fig. 4. Number of species/trap-day accumulated over time.

926 Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 919-928, September 2014 longer sampling period may also be needed to riqueza de especies, Proyecto Chiquita Naturaleza y Comu- obtain a complete picture of species composi- nidad, Reserva Privada Nogal. tion in each habitat. REFERENCES

ACKNOWLEDGMENTS Barlow, J., Overal, W. L., Araujo, I. S., Gardner, T. A., & Peres, C. A. (2007). The value of primary, secondary Special thanks to Christian Miranda for and plantation forests for fruit-feeding butterflies in teaching me all about butterfly identification the Brazilian Amazon. Journal of Applied Ecology, and accompanying me so many times in the 44, 1001-1012. field. We thank the Chiquita Nature and Com- Barquero Villalobos, K. (2009). Corredor Biológico munity Project for allowing us to use data from Local Nogal-La Selva. Chiquita Nature and Com- their own research project. We also thank Isidro munity Project. Retrieved from http://www.nogalna- Chacón for helping us with species identifica- tureand community.com /ndex.php/en/conservation/ tion and everyone at the ACM office, most biological-corridors especially Michael McCoy and Chris Vaughan Bobo, K. S., Waltert, M., Fermon, H., Njokagbor, J., & Mühlengber, M. (2006). From forest to farmland: for guidance on our project and review of this Butterfly diversity and habitat associations along a Special section (ACM) Costa Rica program Costa Special section (ACM) paper, as well as Mario Morera for helping the gradient of forest conversion in southwestern Came- first author in Spanish translations. roon. Journal of Insect Conservation, 10, 29-42. DeVries, P. J., Alexander, L. G., Chacon, I. A., & Fordyce, J. A. (2012). Similarity and difference among rain- RESUMEN forest fruit-feeding butterfly communities in Central Riqueza y abundancia relativa de especies de and South America. Journal of Animal Ecology, Nymphalidae (Lepidoptera) en tres bosques con diver- 81(2), 472-482. sos grados de alteración en el centro norte del Caribe DeVries, P. J. (1987). The butterflies of Costa Rica and de Costa Rica. La medición de la riqueza y abundancia their natural history: Papilionidae, Pieridae, Nympha- de especies puede indicar la salud de un ecosistema y tener lidae. Princeton, NJ, USA: Princeton University. implicaciones importantes para la conservación y su regu- Erhardt, A. (1985). Diurnal Lepidoptera: Sensitive indica- lación. Este estudio investigó la riqueza y abundancia de tors of cultivated and abandoned grasslands. Journal mariposas de la familia Nymphalidae en hábitats alterados of Applied Ecology, 22(3), 849-861. y no alterados, primordialmente en la Reserva Biológica Tirimbina y en la Reserva Privada Nogal, Sarapiquí, Costa Fleishman, E., Thomson, J. R., Nally, R. M., Murphy, D. Rica. Se colocaron trampas cebadas con banano podrido D., & Fay, J. P. (2005). Using indicator species to pre- en el dosel y sotobosque de tres hábitats: en un bosque dict species richness of multiple taxonomic groups. primario, en el lindero entre un río y un bosque y en el Conservation Biology, 19(4), 1125-1137. lindero entre una plantación de bananos y un bosque. Se Hammer, Ø., Harper, D. A. T., & Ryan, P. D. (2001). atrapó e identificó un total de 71 especies y 487 individuos PAST: Paleontological statistics software package for durante mayo y junio 2011 y en mayo 2013. Se encontró education and data analysis. Palaeontologia Electro- que la riqueza y la abundancia de especies aumentaron sig- nica 4(1), 9. nificativamente en los hábitats de los linderos (p<0.0001, Hjältén, J., Stenbacka, F., Pettersson, R. B., Gibb, H., p<0.0001 respectivamente). El efecto de borde, según el Johansson, T., Danell, K., & Hilszczański, J. (2012). cual la riqueza y la abundancia de especies aumentan a Micro and macro-habitat associations in saproxylic causa de la existencia de más recursos complementarios beetles: Implications for biodiversity. PLoS ONE, provenientes de dos hábitats distintos, podría ser una posi- 7(7), 1-11. ble explicación del aumento en la riqueza y la abundancia de especies en hábitats alterados. Horner-Devine, M. C., Daily, G. C., Ehrlich, P. R., & Boggs, C. L. (2003). Countryside biogeography of Palabras clave: abundancia de especies, diversidad de tropical butterflies. Conservation Biology, 17(1), especies, efecto de alteración, mariposas, Nymphalidae, 168-177.

Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 919-928, September 2014 927 Hogue, C. L. (1993). Latin American Insects and Ento- http://www.scientificamerican.com/article. mology. Berkeley, California, USA: University of cfm?id=what-do-butterflies-do-wh California. Ries, L., Sisk, T. D., & Karban, R. (2008). Butterfly edge Leidner, A. K., Haddad, N. M., & Lovejoy, T. E. (2010). effects are predicted by a simple model in a complex Does tropical forest fragmentation increase long-term landscape. Oecologia, 156(1), 75-86. variability of butterfly communities? PLoS ONE, Sjödin, N. E., Bengtsson, J., & Ekbom, B. (2008). The 5(3), 1-8. influence of grazing intensity and landscape com- Municipalidad de Sarapiquí-Secretaria Técnica de Gobierno position on the diversity and abundance of flower- Digital. (2012a). Datos socioeconómicos y ambienta- visiting insects. Journal of Applied Ecology, 45(3), les. Municipalidad de Sarapiquí. Retrieved from 763-772. http://www.sarapiqui.go.cr/index.php?option=com_ Sokal, R. R. & Rohlf, F. J. (1995). Biometry: The princi- content&view=article&id=137 :datos-socioeconomi- ples and practice of statistics in biological research. cos-y-ambientales&catid=12&Itemid=254 New York, NY, USA: Wiley. Municipalidad de Sarapiquí-Secretaria Técnica de Gobierno Digital. (2012b). Información pobla- Statpoint Technologies, Inc. (2011). Statgraphics XV.I cional. Municipalidad de Sarapiquí. Retrie- Statgraphics Centurion. Retrieved from http://www. ved from http://www.sarapiqui.go.cr /index. statgraphics.com/downloads.htm php?option=com_content&view=article&id=136: Tirimbina Biological Reserve. (2010a). Conservation informacionoblacional&catid =12&Itemid=254 Efforts. Tirimbina. Retrieved from Tirimbina website Pardonnet, S., Beck, H., Milberg, P., & Bergman, K. http://www.tirimbina.org/conservation.html (2013). Effect of Tree-Fall Gaps on Fruit-Feeding Tirimbina Biological Reserve. (2010b). Physical Descrip- Nymphalid Butterfly Assemblages in a Peruvian Rain tion. Tirimbina. Retrieved from Tirimbina website Forest. Biotropica, 45(5), 612-619. http://www.tirimbina.org/what-is-tirimbina/physical- Pöyry, J., Luoto, M., Paukkunen, J., Pykälä, J., Raatikai- description.html nen, K., & Kuussaari, M. (2006). Different responses Uehara-Prado, M., Brown, K. S. Jr., & Freitas, A. V. L. of plants and herbivore insects to a gradient of vege- (2007). Species richness, composition and abundance tation height: An indicator of the vertebrate grazing of fruit-feeding butterflies in the Brazilian Atlantic intensity and successional age. OIKOS, 115, 401-412. forest: Comparison between a fragmented and a Raupp, M. (2006). What do butterflies do when it continuous landscape. Global Ecology and Biogeo- rains? Scientific American. Retrieved from graphy, 16(1), 43-54.

928 Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 919-928, September 2014 Beach almond (Terminalia catappa, Combretaceae) seed production and predation by scarlet macaws (Ara macao) and variegated squirrels (Sciurus variegatoides)

Jonathan J. Henn1, Michael B. McCoy2 & Christopher S. Vaughan2 1. Biology Department, St. Olaf College, 1500 St. Olaf Ave, Northfield, MN, USA; [email protected] 2. Associated Colleges of the Midwest, Costa Rica Program, San Jose, Costa Rica; [email protected], [email protected]

Received 20-iii-2014. Corrected 24-iv-2014. Accepted 26-v-2014.

Abstract: Knowledge of ecological impacts of exotic beach almond (Terminalia catappa) in the central Pacific Rica program Costa Special section (ACM) of Costa Rica are little known, but studies have found this species to be a potentially important food source for endangered scarlet macaws (Ara macao). In this study, reproductive phenology and seed predation by variegated squirrels (Sciurus variegatoides) and scarlet macaws were measured during March and April 2011 on beaches of central Pacific coastal Costa Rica. Seed productivity and predation levels were quantified on a weekly basis for 111 beach almond trees to assess the importance of beach almond as a food source for scarlet macaws and the extent of resource partitioning between seed predators. Seed production of the trees was great (about 194 272 seeds) and approximately 67% of seeds were predated by seed predators. Macaws consumed an estimated 49% of seeds while squirrels consumed 18%. Additionally, evidence of resource partitioning between squirrels and macaws was found. Scarlet macaws preferred to feed on the northern side and edge of the canopy while squir- rels preferred to feed on the southern and inside parts of the canopy. Both species ate most seeds on the ocean side of the tree. Despite the status of this tree as an exotic species, the beach almond appears to be an important resource for scarlet macaw population recovery. The resource produced by this tree should be taken into account as reforestation efforts continue in Costa Rica. Rev. Biol. Trop. 62 (3): 929-938. Epub 2014 September 01.

Key words: Ara macao, beach almond, Costa Rica, phenology, scarlet macaw, seed predation, Sciurus variega- toides, Terminalia catappa, variegated squirrel.

The role of, and response to, exotic spe- can result in benefits to native species, espe- cies has recently become a hotly debated topic cially in environments where climate and land (Davis et al., 2011). While exotic species are use is changing rapidly, resulting in modified often viewed as detrimental to native com- native communities (Rodriguez, 2006; Sch- munities, it is important to consider what the laepfer, Sax, & Olden, 2010). actual effects of an exotic species are in this age In this case, the scarlet macaw (Ara macao of rapid environmental change. In many cases L.) in Costa Rica, suffered severe population introduced species do indeed result in negative declines in the 1980s and early 90´s due pri- environmental changes (Vitousek, D’Antonio, marily to deforestation and poaching (Wieden- Loope, & Westbrooks, 1996), but in some feld, 1994; Vaughan, Nemeth, Cary, & Temple, cases, there is evidence that an introduced spe- 2005). This population decline resulted in the cies can improve conditions for some native listing of scarlet macaws as threatened on the species. Changes generated by introduced spe- Appendix I of the Convention on International cies such as pest control, pollination, competi- Trade in Endangered Species (CITES). How- tive and predator release, and trophic subsidies ever, over the last 20 years, habitat restoration,

Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 929-938, September 2014 929 local education, and poaching control have expected that these predators would interact in resulted in considerable growth of the Costa some way, either competing for or partitioning Rican scarlet macaw population (Vaughan et the resource (MacArthur, 1958). al., 2005). As the macaw population has been The objective of this study was to inves- recovering, much about their behavior and tigate the role and the effects of an exotic ecology has been investigated, including diet species as a food source for threatened scarlet (Vaughan, Nemeth, & Marineros, 2006). Most macaws in Costa Rica and how interaction with notably, recent macaw diet studies have indi- variegated squirrels affects macaw use of the cated a large reliance on exotic species, with resource. Specific objectives were to investi- 52% of their diets consisting of exotic spe- gate 1) production and predation levels of T. cies while one specific exotic tree, Terminalia catappa seeds during the dry season and 2) catappa (L.) makes up about 36% of the diet seed predator feeding patterns and interactions (Matuzak, Bezy, & Brightsmith, 2008). on T. catappa trees. Tropical forest, especially dry forest in Central America has been has been severely MATERIALS AND METHODS reduced primarily by deforestation for cattle grazing (Quesada & Stoner, 2003). When graz- Study site: This study was performed on ing occurs, forests have less regeneration of the central Pacific coast of Costa Rica between many native trees that macaws feed on (Stern, the Tárcoles River and the resort Punta Leona Quesada, & Stoner, 2002) and as this reduction (9°45’ N - 84°40’ W). The study area is char- in native food sources has occurred, T. cata- acterized by the transition between dry and ppa has been widely planted in reforestation humid tropical forest (Tosi, 1969), but is now efforts, especially along beaches (Vaughan et dominated by agriculture and small forest al., 2006). patches. The climate is warm and humid, with a Terminalia catappa is native to Southeast mean annual temperature of 25-30°C and mean Asia but, due to ocean currents and humans, it annual precipitation of 2.5-3.3m and distinct has spread to almost all tropical areas. While wet (May-December) and dry (January-April) T. catappa is common throughout Costa Rica, seasons (Coen, 1983). high tolerance to wind and salt stress makes Seven beaches were studied; Playas Azul, it most prevalent on beaches. On beaches, T. Tárcoles, Pita, Agujas, Limoncito, Manta, and catappa is known to prevent erosion of beach Blanca (Fig. 1). Playas Azul, Tárcoles, and Pita sand, provide food and shelter for humans are primarily residential, while playa Agujas and animals, and has potential for commercial has a section developed for tourism while most cultivation for its seeds (Thomson & Evans, is unoccupied by humans. Playas Limoncito, 2006). Based on observations, the seed produc- Manta, and Blanca are part of the resort Punta tion peak of T. catappa coincides with that of Leona and playas Manta and Blanca are fre- many other dry forest species, occurring during quented by resort guests, while Limoncito is the dry season (January-May) (Frankie, Baker, small and isolated, with little human presence. & Opler, 1974). Scarlet macaws, as seed predators, are also Tree selection and characteristics: Judg- important in the reproduction of T. catappa. ment sampling was used to get a wide array of In addition to macaws, variegated squirrels trees of different sizes, locations, and surround- (Sciurus variegatoides Ogilby, 1839), ants, ing characteristics. Number of trees selected and various bats are known to feed on T. cata- per beach varied by beach length and was as ppa seeds (Zuchowski, 2007). Of these other follows; Azul (n=14), Tárcoles (n=20), Pita seed predators, squirrels are the most com- (n=20), Agujas (n=24), Limoncito (n=5), Manta mon (personal observation). In utilizing the (n=16), and Blanca (n=12). Both the total tree same resource as scarlet macaws, it would be height and the height of the leaved canopy

930 Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 929-938, September 2014 Beaches Seed predation data: Seed predation was measured on a weekly basis in 0.5m2 plots. Eight plots per tree were organized with four Azul plots on the interior, approximately 0.5m from the center of the crown and four plots around Costa Rica the edge of the crown. Plots were oriented Tárcoles North, South, East, and West, with two plots in each direction (Fig. 1). Plots 1, 3, 5, and 7 were considered interior plots while plots 2, 4, 6, and 8 were considered exterior plots. Plots 1 and Pita 2 were considered North plots, plots 3 and 4 were considered East plots, plots 5 and 6 were considered South plots, and plots 7 and 8 were Sampling Plot Setup considered West plots. Preliminary trials indi- cated that there was negligible post-dispersal Agujas 2 predation of entire seeds or empty shells, and that the exocarp of the seeds remained green Rica program Costa Special section (ACM) 1 for approximately one week. Thus, only seeds Limoncito with some green exocarp left on their outer 8 7 3 4 shell were considered during weekly sampling. Manta 5 Seeds from each plot were categorized by 5 seed predation classes based on shell markings. These classes included; macaw pre- Blanca 6 dation, macaw rejection, squirrel predation, squirrel rejection, and no predation. Predated Fig. 1. Map of study area and sampling plot setup. seeds were defined as opened seeds where the embryo had been eaten (Fig. 2, Fig. 3, Fig. 4 and Fig. 5). Rejected seeds were defined as (total height - height of trunk without leaves) seeds showing evidence of being picked by a were measured with a clinometer. Additionally, predator, but not opened to expose the inner a tape measure was used to measure the crown part (Fig. 6, Fig. 7). No predation seeds were diameter (distance from edge to edge of the defined as those that showed no sign of pre- crown). A combination of crown height and dation. Macaws and squirrels each had two diameter was used to calculate crown volume. methods of opening seeds (Fig. 2, Fig. 3, Fig. 4 and Fig. 5). Seed production data: Seed production Total seed production and predation were was monitored on a weekly basis on all trees calculated for each tree based on the seed, studied. To quantify the reproductive phenol- flower, seed predation, and seed rejection den- ogy, the number of mature seeds, immature sity obtained during sampling, combined with seeds, and flower heads were counted in three the total volume of the crown and area under randomly selected sections of approximately the crown for each tree. For analytical pur- 1m3 per tree. A mature seed was defined as poses, the no predation category consisted of a seed the size where predators were likely seeds that were rejected by predators and seeds to consume it, determined by observations of that showed no signs of predation. seeds already eaten. Immature seeds were seeds Contingency tables and Chi-squared tests smaller than the seeds frequently consumed by (Sokal & Rohlf, 1995) were used to analyze predators. Flower heads were defined as any trends in preferred feeding location of seed flowering head containing flowers with petals. predators along with the association between

Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 929-938, September 2014 931 Fig. 2. Scarlet macaw (Ara macao) predated fruits (type 1). The fruits have large, smooth chunks taken out of the top half in a messy and imprecise way.

Fig. 3. Scarlet macaw (Ara macao) predated fruits (type 2). Fig. 5. Variegated squirrel (Sciurus variegatoides) predated The fruits are split in half and often most of the exocarp has fruits (type 2). The fruits have been carefully opened on the been removed. Method of opening is messy and imprecise. top. Small grooves made by the squirrel’s teeth can be seen.

Fig. 4. Variegated squirrel (Sciurus variegatoides) predated fruits (type 1). The fruits are mostly if not completely free of exocarp. There are many small grooves which were made by the teeth of the squirrel. Opening to seed is often not directly on top.

932 Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 929-938, September 2014 RESULTS

Seed production and predation: Across 111 trees, approximately 194 272 mature seeds were produced during the 2-mo study period ( =1 854 seeds/tree). A total 24 294 seeds were predated by scarlet macaws, 9 857 seeds rejected by macaws, 8 814 seeds predated by squirrels, 2 448 seeds rejected by squirrels, and 5 419 seeds that fell without signs of predation during the study period. This makes for a total of 50 832 seeds that fell from trees during the study period, 31% of the total mature seeds produced. Of those that fell, 33% had not been Fig. 6. Seeds rejected by scarlet macaws (Ara macao). consumed by predators. The proportion of The exocarp of the seeds has crescent shaped imprints or seeds that fell varied by beach, with Northern chunks taken out by the curved beak of the macaw. beaches (Azul and Tárcoles) having a higher Rica program Costa Special section (ACM) proportion of seeds produced that fell com- pared to other beaches. However, Azul differed from Tárcoles in that a higher proportion of fallen seeds were predated (Table 1).

Predator preference and interaction, beach: Both the frequency and amount of pre- dation varied significantly by beach for macaws and squirrels. Macaw predation was more fre- quent than expected on Azul, Pita, Limoncito, and Manta and less frequent than expected on Tárcoles and Agujas (X2=48.7, df=6, p<0.001) (Fig. 8A). Squirrels, on the other hand, pre- Fig. 7. Seeds rejected by variegated squirrels (Sciurus ferred to feed on Azul and Tárcoles beaches variegatoides). The exocarp of the seeds has long and while feeding less frequently than expected narrow gashes that are often divided by the pair of front on the more Southern beaches (X2=15.8, df=6, teeth. Most damage to exocarp is often located near the p=0.01) (Fig. 2A). Notably, macaws and squir- top on one side. rels differed in feeding preferences on Tár- coles (squirrel preferred), Pita, Limoncito, and seed predators. Observed predation frequency, Manta (macaw preferred) beaches and were split by plot under each tree and by beach, similar on Azul, Agujas, and Blanca. Seeds that was compared to the expected frequency. were not predated were found more often than When predation occurred, differences in mean expected on Tárcoles, Limoncito, and Manta number of seeds predated between plots and and less frequently on Agujas (X2=21.9, df=6, beaches for both macaws and squirrels were p=0.001) (Fig. 8A). analyzed by one-way parametric ANOVAs. When predation did occur, the amount For these ANOVAs, the data were log-trans- varied by beach for each seed predator. For formed and a posteriori comparisons between macaws, significantly more seeds were con- means were analyzed using the Tukey meth- sumed on Southern beaches (Agujas, Limon- od. For interpretability, means are reported cito, Manta, and Blanca) while fewer seeds without transformation. were consumed on the Northern beaches (Azul,

Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 929-938, September 2014 933 TABLE 1 Terminalia catappa mature seed production and predation during study period

Macaw Squirrel Fallen Seeds No Predation Total Mature Mature Seeds Total Fallen Predation Predation Beach Seeds per Tree Seeds % of total % of fallen seeds seeds Azul 9 976 713 6 148 62 50 25 25 Tárcoles 22 649 1 132 9 888 44 39 18 43 Pita 45 961 2 298 12 757 28 49 17 34 Agujas 46 404 1 933 8 061 17 47 10 44 Limoncito 16 622 3 324 3 613 22 62 4 34 Manta 38 887 2 430 6 493 17 50 21 29 Blanca 13 774 1 148 3 873 28 48 28 24 All Beaches 194 272 1 854 50 832 31 49 18 33

Tárcoles, and Pita) (Table 2). Squirrel preda- A 3 tion was consistently less than macaw preda- 2 tion and significantly more predation occurred on Manta compared to Azul and Pita beaches. 1 Significantly more intact seeds fell on Agu- jas and Blanca compared to Azul and Pita 0 beaches (Table 2). -1 Tárcoles was the only beach where

Pearson’s Residual Pearson’s macaws didn’t consume approximately 50% of -2 the predated seeds. Squirrels, on the other hand, -3 consumed a very low proportion of predated seeds on Agujas and Limoncito compared to Pita Azul other beaches (Table 1). Manta Blanca Agujas Tárcoles Predator preference and interaction, Limoncito plot: There were fewer significant differenc- Beach es in predation frequency by plot. Observed B 1.5 macaw predation frequency was not signifi- cantly different than expected (X2=10.5, df=7, p=0.16), although there was a tendency to 0.5 feed more often above plots 2, 7, and 8, and feed less frequently above plots 3, 4, and 6. Squirrel predation frequency also did not vary -0.5 significantly from expected (X2=10.5, df=7, Pearson’s Residual Pearson’s -1.5 Fig. 8. Chi-squared test Pearson’s residuals for observed presence/absence of fallen seeds predated by macaws (black bar) and squirrels (gray bar) or showing no sign -2.5 of predation (white bar) against expected frequency by 1 2 3 4 5 6 7 8 A) beach and B) plot. Positive values indicate predation Plot occurred more frequently than expected.

934 Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 929-938, September 2014 TABLE 2 Mean number of fallen seeds (number/m2/week) predated by macaws and squirrels, or showing no signs of predation by plot and beach (see Fig. 1)

Plot Macaw Squirrel No Predation Beach Macaw Squirrel No Predation 1 1.65ab 0.93 1.16 Azul 1.21a 0.72a 1.06ab 2 1.98b 0.79 1.58 Tárcoles 1.32ab 0.98ab 1.11abc 3 1.62ab 0.66 0.99 Pita 1.77abc 0.73a 0.99a 4 1.14a 0.83 1.11 Agujas 2.64c 1.00ab 1.52bc 5 1.80ab 1.04 1.62 Limoncito 2.10c 0.95ab 1.59abc 6 1.94ab 0.91 1.08 Manta 2.17c 1.16b 1.22abc 7 1.73ab 0.92 1.26 Blanca 1.68bc 0.85ab 1.70c 8 2.76b 1.19 1.40 – – – – F 3.74 0.95 1.97 – 7.15 3.15 4.38 (p) (<0.001) (0.47) (0.058) – (<0.001) (0.005) (<0.001)

Different letters (a,b,c) indicate significant differences between areas using Tukey tests (p<0.05). Special section (ACM) Costa Rica program Costa Special section (ACM) p=0.16) but there was a tendency to feed more Astronium graveolens and a study by Coates- frequently in plots 1 and 6-8 and less frequently Estrada, Estrada, & Meritt (1993) that found than expected in plots 2-5. The observed fre- a mean predation rate of 51% by Amazona quency of intact seeds did differ significantly autumnalis on Stemmadenia donnell-smithii. from expected (X2=21.9, df=6, p=0.001), with These examples are, however, much greater the most notable difference being a much lower than other studies of seed predation by par- frequency observed in plot 3 (Fig. 8B). rots. A study of scarlet macaw predation of When macaws did feed, the amount of Brazil nut (Bertholletia excelsa) revealed that seeds eaten above each plot differed signifi- macaws consumed only 10% of the seed crop cantly, with higher amount of predation in plots (Trivedi, Cornejo, & Watkinson, 2004). The 2 and 8 compared to plot 4. While not sig- predation rate reported in this study, however, nificantly different, there was also a tendency only includes seeds that had fallen and there to eat more seeds in plot 6. Overall, macaws were many seeds left to fall in most trees at the tended to consume more seeds around the edge end of the study. of the canopy on the ocean side. Squirrel preda- Macaws preferred to feed on three beach- tion did not differ significantly by plot although es; Blanca, Manta, and Limoncito. All three there were slightly more seeds consumed above of these beaches are located within a resort/ plots 1, 5, and 8. Plot was marginally sig- club called Punta Leona, but are the furthest nificant in in number of seeds that fell without beaches from the macaw roosting site in the predation (Table 2). Guacalillo Mangrove Reserve (Vaughan et al., 2005) to the north of the study area. This DISCUSSION would not be expect based on optimal foraging theory (MacArthur & Pianka, 1966), but could Regarding scarlet macaw predation on be a result of the conservation efforts made beach almond, the high proportion of T. cata- by Punta Leona. The resort has planted many ppa seeds in scarlet macaw diets (Matuzak et T. catappa trees on its grounds and through- al., 2008) resulted in the consumption of almost out the whole central Pacific coast of Costa half of the seeds produced by T. catappa. Rica. Additionally, the beaches of Punta Leona This is very similar to a study by Villaseñor- are part of the ecological blue flag program Sanchez, Dirzo, & Renton (2010) that found a in Costa Rica, which is only awarded to the 43% seed predation rate by Amazona finschi on healthiest beaches of Costa Rica.

Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 929-938, September 2014 935 Squirrels, on the other hand, had prefer- This study suggests that T. catappa seeds ences for the most human-dominated beaches. are a very important food resource for scar- Azul, Tárcoles, and Pita beaches are all resi- let macaws during the study period (March- dential, with houses adjacent to the beach. April, 2011). As macaw populations have been Squirrels are often associated with humans, increasing in the central Pacific of Costa Rica so they are better adapted to the disturbances (Vaughan et al., 2005), T. catappa has become common on developed beaches. Interestingly, an important part of their diet (Matuzak et al., the number of macaw feeding events was not 2008). Additionally, this study reveals high necessarily lower on residential beaches, but T. catappa seed predation by macaws along the number of seeds eaten was. This indicates beaches. As a tree that grows quickly and pro- that macaws spend less time feeding on resi- duces many seeds, T. catappa planting projects, dential beaches, even though they visit them which are occurring throughout the Central about the same amount as other beaches. This Pacific of Costa Rica, appear to be important in was probably observed because macaws were macaw recovery. Additionally, T. catappa seeds more likely to be disturbed by human presence. are important for other seed predators, and cur- Regarding preferred feeding locations, rent predation rates suggest that enough food is both seed predators fed more often over the produced to support both squirrels and macaws. ocean side (plot 8) and less often away from the This system is an example of how exotic ocean (plot 4). This is likely because the main species can interact with and influence native method of natural propagation of T. catappa is populations. In this case, this exotic species is through the ocean currents and tides (Thomson likely integral in supporting an endangered spe- & Evans, 2006), making it advantageous to cies, and facilitating their recovery, while possi- produce a large number of seeds on the ocean bly beneficial to beach communities (Thomson side. This larger concentration of seeds would & Evans, 2006). This does not, however, mean draw more seed predators. Plot 4, on the other that the overall effects of T. catappa on the hand, faces away from the ocean and is often surrounding community should not be investi- engulfed by other trees, decreasing the light gated to assess whether the species is harming that reaches that side. other elements of the community. The preferred feeding locations in trees The importance of Punta Leona resort as provide some evidence of resource partition- an example to other beaches, is shown by its ing between seed predators in this study. There success and, thus should be a model for other was a weak division in predation between the beaches. Despite being very populated, Punta north and south sides of the tree canopy, where Leona had the greatest macaw presence, which macaws tended to feed on the northern side indicates that beach health and cleanliness is while squirrels preferred to feed on the south- likely more important than human presence. ern side of the tree. This indicates that macaws Not only should reforestation of native trees may prefer to feed more often in the morning as along with T. catappa take place, but measures they fly from their roosting site to the north of should be taken to clean up beaches and reduce the study area. There was also a slight tendency pollution in the area. for squirrels to prefer feeding on the inside During March and April of 2011, the fol- of the tree canopy, while macaws tended to lowing conclusions were reached: a) T. catappa feed on the edge of the canopy. This resource trees on the central Pacific coast of Costa Rica partitioning is likely a result of morphological produced a large number of seeds where natu- differences between species. Macaws fly from ral food sources may have been scarce; b) there tree to tree landing on outer branches. Alter- was evidence of resource partitioning of the natively, squirrels tend to climb the trunk to seed resource between squirrels and macaws; access the crown, making it easier to consume and c) T. catappa trees were a very important seeds along the interior of the tree. component of scarlet macaw diets and the

936 Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 929-938, September 2014 resource size allowed squirrels and macaws to Palabras clave: Ara macao, almendro de playa, Costa persist on the same resource. Rica, fenología, lapa roja, depredación de semillas, Sciurus variegatoides, Terminalia catappa, ardilla chiza.

ACKNOWLEDGMENTS REFERENCES

Thanks goes to the Associated Colleges Coates-Estrada, R., Estrada, A., & Meritt, D. (1993). Fora- of the Midwest, Costa Rica program staff for ging by parrots (Amazona autumnalis) on fruits of coordinating this study abroad program. Kath- Stemmadenia donnell-smithii (Apocynaceae) in the leen Shea provided important feedback on this tropical rainforest of Los Tuxtlas, Mexico. Journal of Tropical Ecology, 9, 121-124. project. Also, thanks to my (JJH) host families Coen, E. (1983). Climate. In D. Janzen (Ed.). Costa Rican during my stay in Costa Rica. Finally, thanks to Natural History (pp. 35-46). Chicago, IL, USA: Uni- Punta Leona resort, especially Arnulfo Villalo- versity of Chicago. bos Arias for his assistance in the field. I would Davis, M., Chew, M. K., Hobbs, R. J., Lugo, A. E., Ewel, J. also like to thank The Parrot Society UK for J., Vermeij, G. J., & Briggs, J. C. (2011). Don’t judge species on their origins. Nature, 474(1), 53-154. funding different aspects of the project. Frankie, G. W., Baker, H. G., & Opler, P. A. (1974). Com- parative phenological studies of trees in tropical wet Rica program Costa Special section (ACM) RESUMEN and dry forests in lowlands of Costa Rica. Journal of Ecology, 62, 881-919. Producción de semillas y depredación del almen- MacArthur, R. H. (1958). Population ecology of some dro de playa (Terminalia catappa, Combretaceae) por warblers of northeastern coniferous forests. Ecology, lapas rojas (Ara macao) y ardillas chizas (Sciurus varie- 39, 599-619. gatoides). El conocimiento de los impactos ecológicos del MacArthur, R. H. & Pianka, E. R. (1966). On the optimal almendro de playa exótico (Terminalia catappa) en el Pací- use of a patchy environment. American Naturalist, fico Central de Costa Rica son poco conocidos, pero los 100, 603-609. estudios han encontrado que esta especie es una fuente de Matuzak, G. D., Bezy, M. B., & Brightsmith, D. J. (2008). alimento potencialmente importante para la lapa roja (Ara Foraging ecology of parrots in a modified landscape: macao), en peligro de extinción. En este estudio, se midie- Seasonal trends and introduced species. The Wilson ron la fenología reproductiva y la depredación de semillas Journal of Ornithology, 120, 353-365. por las ardillas (Sciurus variegatoides) y lapas rojas duran- Quesada, M. & Stoner, K. E. (2003). Threats to the conser- te marzo y abril 2011 en las playas de la costa Pacífica vation of tropical dry forest in Costa Rica. In G. W. Central de Costa Rica. Los niveles de productividad y Frankie, A. Mata, & S. B. Vinso, (Eds.). Biodiversity depredación de semillas se cuantificaron semanalmente Conservation in Costa Rica: Learning the lessons in para 111 almendros de playa, para evaluar la importancia the seasonal dry forest (pp. 266-280). Berkeley, CA, del almendro de playa como fuente de alimento para la lapa USA: University of California. roja y el grado de repartición de recursos entre los depreda- Rodriguez, L. F. (2006). Can invasive species facilitate dores de semillas. La producción de semillas de los árboles native species? Evidence of how, when, and why fue alta (cerca de 194 272 semillas) y aproximadamente these impacts occur. Biological Invasions, 8, 927-939. el 67% de las semillas fueron comidas por los depreda- Schlaepfer, M. A., Sax, D. F., & Olden, J. D. (2010). The dores, las lapas rojas consumen un estimado de 49% de potential conservation value of non-native species. las semillas, mientras que las ardillas consumen el 18%. Conservation Biology, 25, 428-437. Adicionalmente, se encontró evidencia de la repartición Sokal, R. R. & Rohlf, F. J. (1995). Biometry: The principles de recursos entre las ardillas y las lapas. Las lapas rojas and practice of statistics in biological research. New prefieren alimentarse en el lado norte y el borde de la copa, York, USA: W. H. Freeman and Co. mientras que las ardillas prefieren las regiones del sur y el Stern, M., Quesada, M., & Stoner, K. E. (2002). Changes interior de la copa. Ambas especies se comieron la mayoría in composition and structure of a tropical dry forest de las semillas en la parte del árbol con lado al mar. A pesar following intermittent cattle grazing. Revista de Bio- de la situación de este árbol como una especie exótica, la logia Tropical, 50(3/4), 1021-1034. playa de almendras parece ser un recurso importante para la Thomson, L. A. J. & Evans, B. (2006). Terminalia catappa recuperación de la población de lapas rojas. El recurso que (tropical almond), ver. 2.2. Species Profiles for produce este árbol debe tomarse en cuenta para continuar Pacific Island Agroforestry. Permanent Agriculture con los esfuerzos de reforestación en Costa Rica. Resources (PAR). Hōlualoa, HI, USA. Retrieved

Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 929-938, September 2014 937 from http://agroforestry.net/tti/T.catappa-tropical- Villaseñor-Sanchez, E., Dirzo, I. R., & Renton, K. (2010). almond.pdf Importance of the lilac-crowned parrot in pre-dis- Tosi, J. (1969). Ecological map of Costa Rica. San José, persal seed predation of Astronium graveolens in Costa Rica: Tropical Science Center. a Mexican tropical dry forest. Journal of Tropical Trivedi, M. R., Cornejo, F. H., & Watkinson, A. R. (2004). Ecology, 26, 227-236. Seed predation on Brazil nuts (Bertholletia excelsa) Vitousek, P. M., D’Antonio, C. M., Loope, L. L., & by macaws (Psittacidae) in Madre de Dios, Peru. Westbrooks, R. (1996). Biological invasions as glo- Biotropica, 36, 118-122. bal environmental change. American Scientist, 84, Vaughan, C., Nemeth, N. M., Cary, J., & Temple, S. (2005). 468-478. Response of a scarlet macaw (Ara macao) population Wiedenfeld, D. (1994). A new subspecies of scarlet macaw to conservation practices in Costa Rica. Bird Conser- vation International, 15, 119-130. (Ara macao) and its status and conservation. Ornito- logia Neotropical, 5, 99-104. Vaughan, C., Nemeth. N., & Marineros. L. (2006). Scarlet Macaw, (Ara macao), diet in Central Pacific Costa Zuchowski, W. (2007). Tropical plants of Costa Rica. Itha- Rica. Revista de Biología Tropical, 54, 919-926. ca, New York, USA: Cornell University.

938 Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 929-938, September 2014 Species diversity and activity of insectivorous bats in three habitats in La Virgen de Sarapiquí, Costa Rica

Amanda Cormier Colorado College 902 North Cascade Avenue, Colorado Springs, CO, 80946, USA; [email protected]

Received 20-III-2014. Corrected 20-IV-2014. Accepted 26-V-2014.

Abstract: Pineapple farms make up 45 000ha of Costa Rican landscape and are the second most exported crop. This is economically beneficial for the Costa Ricans, but greatly affects the natural flora and fauna because it is such a low growing crop. This study examined the differences in insectivorous bat species diversity and

activity in the habitat gradient between the forest in Tirimbina Biological Reserve in La Virgen de Sarapiquí, Rica program Costa Special section (ACM) Heredia, Costa Rica and the nearby pineapple farm called Finca Corsicana. Over a four week period in March and April 2013, ultrasonic recorders were placed at different sites to pick up the bats’ calls. Then the recordings were analyzed to identify the species. There were four families present and 19 different species. There was a significant decrease in the number of bat passes (the number of times a bat passes the recorder) in the pineapple farm ( =22.6), in comparison to the border ( =39.9), and the forest ( =44.2) (p=0.0028). Agricultural environ- ments affected and lowered bat presence. Also, a greater mean number of bats recorded between 1900-1930hrs compared to 1730-1800hrs, coincided with the setting of the sun and beginning of bat activity. More research is need throughout the night and the year to establish clearer patterns of bat use and activity in different habitats. Rev. Biol. Trop. 62 (3): 939-946. Epub 2014 September 01.

Key words: bat activity, Costa Rica, habitat gradient, insectivorous bats, tropical rain forest, ultrasonic recorders.

Conservation efforts are being made all However, because it is a low-growing plant it over the world to help protect the environment drastically changes the landscape and thus the from human destruction. In order for these habitat of many species. Can bat species adapt conservation efforts to be implemented, there to these new environments? must be an understanding of: a) how changes Miller, Arnett, and Lacki (2003) exami- to the landscape affect the flora and the fauna, ned 56 bat studies on habitat management of and b) how to correct the problem or at least forest-roosting bats and concluded that without create a balance between the needs of humans roosting areas, bats will leave. Thus conser- and the natural life. This study focuses on how vation efforts should be made to maintain the gradation in diversity between a biological roosting areas. It is highly beneficial to farmers reserve (protected forest), the border of the to keep bats in an area because insectivorous forest, and a pineapple farm affect bat species bats greatly decrease pests and frugivorous diversity and activity. and nectivorous bats help with pollination. It Pineapple is the second most exported crop has been estimated that a colony of big brown in Costa Rica. Of the pineapples consumed in bats (Eptesicus fuscus) can eat 1.3 million the US and Europe, 85% and 71% are from insects a year, or each bat can eat 4 to 8g per Costa Rica, respectively (Ingwersen, 2012). night. In monetary terms this is around $74/ This makes pineapple economically important acre that farmers in south-central Texas, USA to Costa Rica. As of 2011, there were 45 000ha save in pesticides (Boyles, Cryan, McCracken, of land cultivated in pineapple in Costa Rica. & Kunz, 2011).

Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 939-946, September 2014 939 Of 1 116 species of the bats in the world, eco-image includes information about the form 113 are found in Costa Rica and about 60 and the texture of an object. The frequency are known in the rural town of La Virgen de of these calls is higher than the human range, Sarapiquí, where this study was conducted which stops at 20kHz. One bat can call as many (Laval & Rodríguez-Herrera, 2002). Around as 400 000 times per night. Bats use two types 75% of all bat species are insectivores, the of calls, the frequency-modulated (fm), which rest eat fruit, nectar, fish, birds, or blood (Bat is in a wide frequency band for a short period Conservation International, Inc., 2011). Even of time, and the quasi-constant frequency (qcf), though insectivorous bats make up the majo- which is in a narrow frequency band for a lon- rity of bat species, there are fewer studies on ger period of time (Schnitzler & Kalko, 1998). insectivorous bats in comparison to other types Each species’ fm and qcf calls are different, because insectivorous bats usually fly higher which is how recording devices can be used to than mist nets, which are typically used in bat identify different bat species. studies. The solution to this problem is recor- The main objectives of this study were ding devices. to determine species richness in a gradient of Within the past 15 years there has been disturbance in La Virgen de Sarapiquí. Through an increase in sonic recorder use to identify the use of sonic recorders, this study aimed to bat species difficult to catch in nets. Initially, identify bat species in order to determine their research groups worked to identify all sounds activity and diversity within: a) the forest at made by each bat species by capturing them Tirimbina Biological Reserve, b) the edge of and recording calls of each species in open this forest, and c) the edge of the pineapple rooms. With species recordings, they were farm, Finca Corsicana. able to identify bats in the field with recording This will help build a better understanding equipment (O’Farrell & Miller, 1999; Rydell, of the interaction between bats and humans Arita, Santos, & Granados, 2002; Jung, Kalko, based on the effects of agriculture and create & von Helversen, 2007). Evidence of the a better understanding of how best to protect usefulness of recording calls was shown in a bat habitats. study (Sampaio, Kalko, Bernard, Rodríguez- Herrera, & Handley, 2003) that focused on bat MATERIALS AND METHODS species adaptability to new surroundings in the Amazon Basin in Brazil with mist net captures Study area: This study took place in La at ground and canopy levels. However, sonic Virgen de Sarapiquí, Heredia, in northcentral recordings detected 15 bat species not caught Costa Rica (10°24’41” - 10°27’05” N, and during the 2.5yrs of the study. MacSwiney, 84°06’52” - 84°07’43” W) during March and Clarke, and Racey (2008) specifically tested April 2013. La Virgen is a small town that runs how effective the recorders are in the identifi- along the Sarapiquí River. About 7km to the cation of bat species and found that mist nets north of the center of La Virgen is Tirimbina missed 30% of bat fauna and aerial insectivores Biological Reserve and about 2km further (insectivorous bats that do not land to catch North, along Route 126, is the Finca Corsicana, insects) were not caught at all; they were only a pineapple farm. The majority of the study was identified by recorders. done in Tirimbina Biological Reserve, which is The use of recording devices is possi- 345ha of protected forest (Tirimbina Biological ble because the majority of bat species use Reserve, 2010) (Table 1). echolocation to navigate and find food during the night (Rodríguez-Herrera & Montero, Habitat sites: There were three main 2009). The bats send out the call and use the habitat sites: 1) the forest, 2) forest border, and echo that bounces back to form an image of the 3) pineapple farm. The forest was part of their surroundings, called an eco-image. This Tirimbina Biological Reserve, which is 98.55%

940 Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 939-946, September 2014 TABLE 1 on the road to Tirimbina’s field station (C2), The 12 recorder-collection sites and their GPS coordinates and the third at La Quinta hotel (C3). During week D (April 15, 2013 to April 19, 2013) the Site GPS Coordinates first recorder was placed within the forest off A1 Second Bridge 10°24’51” N, 84°07’07” W the Botarrama trail (D1), the second on an area A2 Open Building 10°24’58” N, 84°07’23” W known as the island, which divides the river A3 House 1 10°25’37” N, 84°07’43” W in two (D2), and the third at house 2 near the B1 Corteza Trail 10°24’44” N, 84°07’03” W B2 Cacao Farm 10°24’41” N, 84°07’18” W pineapple farm (D3). B3 School 10°25’37” N, 84°07’43” W C1 Hunter Trail 10°24’34” N, 84°06’52” W Bat call collection: Each week one recor- C2 Road to Field Station 10°25’12” N, 84°07’17” W der (Song Meter SM2BAT+, Wildlife Acous- C3 La Quinta 10°27’05” N, 84°07’17” W tics, Concord, Massachusetts, USA) was placed D1 Botarrama Trail 10°25’04” N, 84°06’52” W in each of the three habitat sites and this process D2 Island 10°25’08” N, 84°07’07” W was repeated for 4wks. They were placed out D3 House 2 10°26’52” N, 84°07’38” W on Monday or Tuesday and collected Saturday or the following Monday to ensure five whole nights of data collection. Table 2 illustrates the Rica program Costa Special section (ACM) forest, 1.15% open area; mostly comprised nights data was collected. Recorders ran from of walking trails, and 0.30% wetlands. The 1730hrs to 0530hrs. Each week, after retrieving forested area is made up of 85% primary forest the recorders, the data chips from the recorders and 15% secondary forest and these include were inserted into my computer (MacBook tropical pre-montane forest that transitions into Pro) and the calls were transferred to an exter- basal and tropical forest. The west border of nal hard drive. For each night there were eight the reserve is made up of the Sarapiquí River files in wav format, each 1.5hrs in length. Then and Route 126, so the recorder-collection sites before moving the recorders to new recorder- along the border have forest on one side and collection sites the batteries were changed open space on the other. Finally the 1 200ha (only alkaline batteries will last the full length pineapple farm is less than 2km NW of Tirim- of time) and the data was deleted off the data bina and is found on both sides of Route 126. chips in the recorders from the previous week to make space for new data. Recorder-collection sites: At each of the three habitat sites there were four recorder- Bat call identification: During daylight collection sites (one used per week). During hours I analyzed the recordings using Raven week A (March 11, 2013 to March 15, 2013) Pro 1.5, (The Cornell Lab of Ornithology, Itha- one recorder was placed on the tower of ca, NY, USA), a computer program designed the second bridge within the forest (A1), the to analyze sounds. I opened each file within second within an open building at the edge of this program, so that I was able to see 3min the forest (A2), and the third at house 1 along at a time. Passes were defined as the number the edge of the pineapple farm (A3). During of times a bat flew next to the recorder. Based week B (March 18, 2013 to March 22, 2013) on the definition by Estrada-Villegas, Meyer, one recorder was placed just off the Corteza & Kalko (2010), a pass ends when the time trail, which is past the bridge (B1), the second between two calls is three times longer than in the cacao plantation, which is along the river the time between the first two calls. I recor- (B2), and the third on the side of a building part ded the number of passes/species/ min in an of Escuela Las Palmitas next to the pineapple excel spreadsheet. farm (B3). During week C (April 8, 2013 to To identify bat species, I compared the April 12, 2013) one recorder was placed on peak frequency, the length of the call, and the the Hunter trail (C1), the second was placed shape of the call to known calls (O’Farrell &

Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 939-946, September 2014 941 TABLE 2 Data collection dates. Data not collected during the first two weeks for the full five nights at each site was because batteries did not last long enough

Week A Site 11-Mar 12-Mar 13-Mar 14-Mar 15-Mar 16-Mar A1 Forest X A2 Edge X X X X X A3 Pineapple X X X X X Week B Site 18-Mar 19-Mar 20-Mar 21-Mar 22-Mar 23-Mar B1 Forest X X B2 Edge X X B3 Pineapple X X Week C Site 8-Apr 9-Apr 10-Apr 11-Apr 12-Apr 13-Apr C1 Forest X X X X X C2 Edge X X X X X C3 Pineapple X X X X X Week D Site 15-Apr 16-Apr 17-Apr 18-Apr 19-Apr 20-Apr D1 Forest X X X X X D2 Edge X X X X X D3 Pineapple X X X X X

Miller, 1997; Fenton, Rydell, Vonhof, Eklof, & c) bat families and three habitats; and d) four Lancaster, 1999; Kossl, Mora, Coro, & Vater, weeks and three habitats, with parametric two- 1999; O’Farrell & Miller, 1999; O’Farrell, way ANOVA. Differences between means of Miller, & Gannon, 1999; Siemers, Kalko, & the variable, “number of species”, were esti- Schnitzler, 2001; Mora, Macías, Vater, Coro, mated between: a) three habitats and two time & Kossl, 2004; Jung et al., 2007; MacSwiney periods with parametric two-way ANOVA. et al., 2008; Pio, Clarke, MacKie, & Racey, Assumptions were tested prior to execution of 2010; Jung & Kalko, 2011). Peak frequency parametric ANOVAS (Sokal & Rohlf, 1995). is the frequency at the maximum power of the Tests were executed with Statgraphics statisti- call. These researchers have captured different cal software (www.statgraphics.com). species and recorded their calls in order to accurately identify the species. Many of the RESULTS articles mentioned above contained pictures of the sounds that I compared to the images in In 94 30-min segments over 4wks, several Raven Pro 1.5. trends appeared. There were fewer bat passes Time constraints hindered the analysis of for the pineapple farm ( =22.6) than for the all 12h from each night. Instead I analyzed forest ( =44.2) or border ( =39.9) (F=5.69, the calls only from 1730-1800hrs and 1900- d.f.=2, 312, p=0.0028, Fig. 1). There was a 1930hrs for each night. significant increase in the mean number of bat Differences between means of the varia- passes between 1730-1800hrs ( =25.4) and ble, “number of passes”, were estimated bet- 1900-1930hrs ( =44.3) (F=11.71, d.f.=1, 313, ween: a) the three habitats and two time p=0.0001). This trend was consistent for all periods; b) bat families and two time periods; habitats (F=0.65, d.f.=2, 309, p=0.5241, Fig. 1).

942 Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 939-946, September 2014 80 19 different species were identified (Cormura brevirostris, Cynomops greenhalli, Cyttarops 70 17:30-18:00 hrs 60 19:00-19:30 hrs alecto, Diclidurus albus, Eptesicus brasilien- 50 sis, Eptesicus furinalis, Euderma maculatum , 40 Eumops spp, Lasiurus ega, Molossus curren- 30 tium, Molossus molossus, Myotis albescens, 20 Myotis nigricans, Peropteryx kappleri, Perop- Mean Number of Passes 10 teryx macrotis, Pteronotus parnellii, Rhogeessa 0 io, Saccopteryx bilineata, Saccopteryx leptu- Forest Border Pineapple ra). There was a significant difference in mean number of bat passes per family with Emba- Habitat llonuridae having the least ( =24.0), then Ves- Fig. 1. Mean number of passes recorded for each habitat pertillionidae ( =34.3), and Molossidae with for each 30-minute period. Vertical lines are +1 Tukey the greatest mean number of passes ( =45.9) HSD confidence intervals (95.0%). Two-way ANOVA (F=4.32, d.f.=2, 308, p=0.0141). When families with insignificant interaction (F=0.65; df=2, 304; p=0.52). were compared to the three habitats there was March-April 2013, La Virgen de Sarapiquí, Costa Rica. a significant difference (F=3.33, d.f.=4, 302, Rica program Costa Special section (ACM) p=0.0109). The mean number of bat passes for There was also a significant difference in Emballonuridae was greatest at the border sites the mean number of species per habitat and ( =40.5) and the smallest in the forest ( =11.5, 30-min period (F=6.98, d.f.=2, 88, p=0.0015, Fig. 3). Mean number of passes of Molossidae Fig. 2). At 1730-1800hrs, there were more spe- was greatest at the border as well ( =54.4) and cies in the forest ( =3.69) and border ( =3.59) smallest in the pineapple farm ( =16.4, Fig. 3). than during 1900-1930hrs (forest =3.0, bor- In contrast, the greatest mean number of passes der =3.47). The opposite occurred at the of Vespertillionidae was in the forest ( =53.7) pineapple site where there were fewer species and the least in the border habitat ( =28.3). during 1730-1800hrs ( =2.18) than during There was no significant difference in 1900-1930hrs ( =4.06). mean number of passes at each habitat between Four families of bats were found: Emba- weeks (F=1.45, d.f.=6, 303, p=0.939, Fig. 4). llonuridae, Molossidae, Vespertillionidae, and For all four weeks the mean number of passes only two passes of Moormopidae (removed for statistical testing). Of the four families, 80 70 Emballonuridae 5 60 Molossidae 17:30-18:00 hrs 4.5 Vespertilionidae 19:00-19:30 hrs 50 4 40 3.5 30 3 20 2.5 Mean Number of Passes 10 2

Mean Number of Species 0 1.5 Forest Border Pineapple Forest Border Pineapple Habitat Habitat

Fig. 2. Mean number of species recorded per habitat for Fig. 3. Mean number of passes recorded for each family in each 30-minute period. Vertical lines are +1 Tukey HSD each habitat. Vertical lines are +1 Tukey HSD confidence confidence intervals (95.0%). Two-way ANOVA showing intervals (95.0%). Two-way ANOVA showing significant significant interaction (F=6.98; df=2, 88; p=0.0015). interaction (F=3.33; df=4, 302; p=0.011). March-April March-April 2013, La Virgen de Sarapiquí, Costa Rica. 2013, La Virgen de Sarapiquí, Costa Rica.

Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 939-946, September 2014 943 90 two families, more passes occurred during 80 Forest 1730-1800hrs ( =36.14) than during1900- 70 Border 1930hrs ( =32.64). 60 Pineapple 50 40 DISCUSSION 30 20 The main objective of this study was to Mean Number of Passes 10 examine the effect of habitat gradation on bats 0 and results showed a difference in bat activity 11-15 Mar. 18-22 Mar. 8-12 Apr. 15-19 Apr. with the greatest mean number of passes in Week the forest and decreasing to the smallest mean number of passes in the pineapple farm. No Fig. 4. Mean number of passes recorded per week for each habitat. Vertical lines are +1 Tukey HSD confidence significant difference existed in mean number intervals (95.0%). Two-way ANOVA with insignificant of species between habitats, but a significant interaction (F=1.45; df=6, 303; p=0.2). March-April 2013, difference existed in mean number of passes La Virgen de Sarapiquí, Costa Rica. per family in each habitat. Based on different calls, different families per habitat increased, decreased and, except for seemed better suited to different habitats. For the pineapple farm, increased again in week D. example, Vespertilionidae use fm calls with Between the two 30-min periods was an a long frequency and short duration, making hour of time. This hour created a significant them suited for forest habitats, where the difference in the activity of the different fami- greatest number of passes occurred (Schnitzler lies (F=6.61, d.f.=2, 305, p=0.0015, Fig. 5). & Kalko, 1998). Voigt and Holderied (2012) There were fewer passes for the Emballonuri- found Molossidaes better suited for open spa- dae ( =11.57) and the Molossidae ( =23.06) ces partly because of their physical conforma- during 1730-1800hrs than between 1900- tion (body size and wing length) and their calls 1930hrs (Emballonuridae =35.53, Molossidae are qcf, almost cf, which means they have a =64.40). Mean number of passes of Vesperti- small change in frequency and a longer length. lionidae changed little and unlike the other In their study of M. currentium and M. sinaloe in open and closed spaces, the M. currentium performed better in the dense forest-like areas 100 than the M. sinaloe. In my study there were a 17:30-18:00 hrs number of M. currentium and no M. sinaloe, 80 19:00-19:30 hrs which may explain why there was a greater number of Molossidae found in the forest and 60 the border than the pineapple farm. Alpízar, 40 Víquez, Hong, Rodríguez-Herrera, and Gonzá- lez-Maya (2012) in Tirimbina, with bat species 20 in dense and open spaces within the forest, Mean Number of Passes found greater numbers of Emballonuridae in 0 Emballonuridae Molossidae Vespertilionidae the forest than the open spaces, but overall Family there was more species richness in open spaces (Jaccard=0.308). I found more Emballonuri- Fig. 5. Mean number of passes recorded for each family dae in the border than the other two habitats. for each 30-minute period. Interactions and vertical lines Emballonuridae use gcf calls, which have both are +1 Tukey HSD confidence interval (95.0%). Two- way ANOVA showing significant interaction (F=6.61; a large change in frequency and a long length df=2, 305; p=0.0015). March-April 2013, La Virgen de of time making them ideally suited for border Sarapiquí, Costa Rica. spaces (Schnitzler & Kalko, 1998).

944 Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 939-946, September 2014 While there was no significant difference my equipment. Additionally, Priscilla Alpízar in number of passes per habitat between weeks, helped in teaching me how to use the computer there were some interesting trends that can be programs and how to identify the species based explained by the individual recorder-collection on the images. Finally, I would like to thank sites. In week B, the mean number of pas- all of Tirimbina’s staff for their encouragement ses in the forest and the border were almost throughout the data collection process, especia- the same (forest =48.07, border =48.14). lly Emmanuel Pequeño for helping me put out During week B the forest recorder was placed and take down the recorders every week. low to the ground, which could have prevented the microphone from picking up all of the calls and the border site was in the cocoa plantation, RESUMEN which is along the Sarapiquí River, but the Diversidad y actividad de especies de murciélagos recorder was more in the plantation than on the insectívoros en tres hábitats de La Vírgen, de Sarapi- edge of the forest. In week C the forest recorder quí, Costa Rica. Hay 45 000ha de piñeras en Costa Rica. was placed in a tree on the side of a slope that La piña es la segunda fruta de mayor exportación en el was above a small river, thus creating more of país, por lo cual es muy importante para la economía. Este cultivo beneficia al país en términos económicos; no an open space than forest space, which could Rica program Costa Special section (ACM) obstante, tiene efectos negativos en la flora y la fauna de have lead to a decrease in passes ( =27.0). áreas naturales al ser una planta herbácea de baja altura. The forest recorder in week D was the furthest Esta investigación examinó la diversidad y actividad de los from the trail than in the other weeks creating a murciélagos insectívoros en el gradiente entre un bosque denser environment which may be why there is (Reserva Biológica Tirimbina) y una piñera (Finca Corsi- the highest mean number of passes at this site cana) en la Virgen, Sarapiquí, Costa Rica. Durante cuatro ( =65.60). semanas, puse grabadoras en tres sitios (bosque, borde de bosque y piñera) para grabar las llamadas de los murcié- lagos. Después analicé las llamadas para identificar las The last major trend was the increase of especies. Encontré cuatro familias y 19 especies. Hubo una mean number of passes between 1900-1930hrs reducción en el número de pases de murciélagos (número when compared to 1730-1800hrs (p=0.0001). de veces que un murciélago vuela al lado de la grabadora) Bats are nocturnal species and the sun sets en la piñera ( =22.6), en comparación con el borde de around 1730hrs, so by 1900hrs there has been bosque ( =39.9) y el bosque ( =44.2) (p=0.0028). El ambiente agrícola afectó y redujo la presencia de murciéla- more time for the bats to wake up and head out gos. Familias diferentes parecen estar adaptadas a hábitats to feed. Jung & Kalko (2010) noticed a decrea- diferentes. Finalmente, un mayor número de murciélagos se in bat activity after 2030hrs in their human fueron registrados entre las 19:00 y 19:30 hrs, comparados populated site, but no change throughout the con las 17:30-18:00 hrs, lo cual coincide con el atardecer night in their forest site. More analysis would y el inicio de la actividad nocturna de los murciélagos. Se have to be done of the bat calls from my study necesita más investigación, tanto en horas nocturnas, como del resto del año, para entender mejor los patrones de acti- in order to clearly notice a trend in bat activity vidad y el uso de diferentes hábitats. throughout the night. Palabras clave: murciélagos insectívoros, actividad de murciélagos, grabadoras ultrasónicas, gradiente de hábi- ACKNOWLEDGMENTS tats, bosque tropical lluvioso, Costa Rica. I would like to thank Bernal Rodríguez- Herrera for helping to guide me through this REFERENCES project with his vast knowledge of bats. Also, I would like to thank Christopher Vaughan and Alpízar, P., Víquez, L. R., Hong, F., Rodríguez-Herrera, Michael McCoy for their help in editing and B., & González-Maya, J. F. (2012). Efecto de los claros de bosque en la composición de murciélagos advising during the study. Judy Magnan arran- insectívoros en la Reserva Biológica La Tirimbina, ged my home stay for 2mo and Alejandra Loai- Sarapiquí, Costa Rica. Revista Biodiversidad Neotro- za was instrumental in helping to collect all of pical, 2(2), 138-142.

Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 939-946, September 2014 945 Bat Conservation International, Inc. (2013). Bat Conserva- O’Farrell, M. J. & Miller, B. W. (1997). A new examination tion International, Inc. Retrieved from Bat Conser- of calls of some Neotropical bats (Emballonuridae vation International website. Retrieved from http:// and Mormoopidae). Journal of Mammalogy, 78(3), www.batcon.org 954-963. Boyles, J. G., Cryan, P. M., McCracken, G. F., & Kunz, T. O’Farrell, M. J. & Miller, B. W. (1999). Use of vocal signa- H. (2011). Economic importance of bats in agricultu- tures for the inventory of free-flying Neotropical bats. re. Science, 332, 41-42. Biotropica, 31(3), 507-516. Estrada-Villegas, S., Meyer, C. F. J., & Kalko, E. K. V. O’Farrell, M. J., Miller, B. W., & Gannon, W. L. (1999). (2010). Effects of tropical forest fragmentation on Qualitative identification of free-flying bats using aerial insectivorous bats in a land-bridge island sys- the Anabat detector. Journal of Mammalogy, 80(1), tem. Biological Conservation, 143, 597-608. 11-23. Fenton, M. B., Rydell, J., Vonhof, M. J., Eklof, J., & Pio, D. V. V., Clarke, F. M., MacKie, I., & Racey, P. A. Lancaster, W. C. (1999). Constant-frequency and (2010). Echolocation calls of the bats of Trinidad, frequency-modulated components in the echolocation West Indies: Is guild membership reflected in echo- calls of three species of small bats (Emballonuri- location signal design? Acta Chiropterologica, 12(1), dae, Thyropteridae, and Vespertilionidae). Canadian 217-229. Journal of Zoology, 77(12), 1891-1900. Rodríguez-Herrera, B. & Montero, K. (2009). El ultra- Ingwersen, W. W. (2012). Life cycle assessment of fresh sonido de los murciélagos: Historia y retos. Boletín pineapple from Costa Rica. Journal of Cleaner Pro- Oficial del Proyecto de Conservación de Aguas y duction, 35, 152-163. Tierras ProCAT, 2(2), 19-24. Jung, K. & Kalko, E. K. V. (2010). Where forest meets Rydell, J., Arita, H. T., Santos, M., & Granados, J. (2002). urbanization: Foraging plasticity of aerial insectivo- Acoustic identification of insectivorous bats (order rous bats in an anthropogenically altered environ- Chiroptera) of Yucatan, Mexico. The Journal of the ment. Journal of Mammalogy, 91(1), 144-153. Zoological Society of London, 257(1), 27-36. Jung, K. & Kalko, E. K. V. (2011). Adaptability and vul- Sampaio, E. M., Kalko, E. K. V., Bernard, E., Rodríguez- nerability of high flying Neotropical aerial insectivo- Herrera, B., & Handley, C. O., Jr. (2003). A biodi- rous bats to urbanization. Diversity and Distributions, versity assessment of bats (Chiroptera) in a tropical 17(2), 262-274. lowland rainforest of central Amazonia, including Jung, K., Kalko, E. K. V., & von Helversen, O. (2007). methodological and conservation considerations. Stu- Echolocation calls in Central America emballonurid dies on Neotropical Fauna and Environment, 38(1), bats: Signal design and call frequency alternation. 17-31. Journal of Zoology, 272, 125-137. Schnitzler, H. U. & Kalko, E. K. V. (1998). How echolo- Kossl, M., Mora, E., Coro, F., & Vater, M. (1999). Two- cating bats search and find food. In T. H. Kunz & toned echolocation calls from Molossus molossus in P. A. Racey, (Eds.), Bat biology and conservation Cuba. Journal of Mammalogy, 80(3), 929-932. (pp. 183-196). Washington, D.C., USA: Smithsonian Institution. Laval, R. K. & Rodríguez-Herrera, B. (2002). Murciélagos de Costa Rica. San José, Costa Rica: Editorial INBio. Siemers, B. M, Kalko, E. K. V., & Schnitzler, H. U. (2001). MacSwiney, M. C., Clarke, F. M., & Racey, P. A. (2008). Echolocation behavior and signal plasticity in the What you see is not what you get: The role of ultra- Neotropical bat Myotis nigricans (Schinz, 1821) sonic detectors in increasing inventory completeness (Vespertilionidae): A convergent case with European in Neotropical bat assemblages. Journal of Applied species of Pipistrellus? Behavioral Ecology and Ecology, 45, 1364-1371. Sociobiology, 50, 317-328. Miller, D. A., Arnett, E. B., & Lacki, M. J. (2003). Habitat Sokal, R. R. & Rohlf, F. J. (1995). Biometry: The principles management for forest-roosting bats of North Ame- and practice of statistics in biological research. New rica: A critical review of habitat studies. Wildlife York, NY, USA: W. H. Freeman and Co. Society Bulletin, 31(1), 30-44. Tirimbina Biological Reserve. (2010). Retrieved from Mora, E. C., Macías, S., Vater, M., Coro, F., & Kossl, M. Tirimbina website: http://www .tirimbina.org). (2004). Specializations for aerial hawking in the Voigt, C. C. & Holderied, M. W. (2012). High maneuvering echolocation system of Molossus molossus (Molossi- costs force narrow-winged molossid bats to forage dae, Chiroptera). Journal of Comparative Physiology, in open space. Journal of Comparative Physiology, 190, 561-574. 182, 415-424.

946 Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 62 (3): 939-946, September 2014