1

Azteca ants inhabiting Cecropia trees are moving up in elevation in Monteverde

Meghan R. Garfink Department of Environmental Science and Policy University of California, Davis UCEAP Tropical Biodiversity and Conservation Spring 2019 7 June 2019

ABSTRACT

Global climate change impacts ecological relationships of all taxa and kingdoms, yet little has been documented of the impact on insects. This study documents the changing elevational range of four Azteca ant species in Cecropia trees in the Monteverde cloud forest. In Monteverde, Costa Rica, many species of Cecropia form a mutualism with Azteca in which ants receive nutrients in exchange for protection of the tree. Data was collected in May 2019 and compared to similar studies done in 1989 (Longino), 2003 (Mazzei), 2008 (Loope), and 2016 (Jensen). Since 1989, maximum elevational range of Azteca increased by 132 meters. Average elevation of all species of Azteca and Cecropia have increased significantly since 1989 as well. One species of Cecropia, C. polyphlebia, contained Azteca at 1500 m, despite historically not harboring Azteca. This data indicates that Azteca are moving upslope and inhabiting different species of Cecropia in Monteverde, potentially due to climate change.

Hormigas Azteca que habitan en árboles de Cecropia en Monteverde se están desplazando hacia arriba en elevación

RESUMEN

El cambio climático global afecta las relaciones ecológicas de todos los taxones y reinos, sin embargo, el impacto en los insectos ha sido poco documentado. Este estudio documenta el cambio del rango de elevación de cuatro especies de hormigas Azteca que viven en árboles de Cecropia en el bosque nuboso de Monteverde. En Monteverde, muchas especies de Cecropia tienen una relación mutualista con Azteca, en el que las hormigas reciben nutrientes y casa a cambio de proteger el árbol hospedero. Recopilé datos en mayo de 2019 y los comparé con estudios similares realizados en 1989 (Longino), 2003 (Mazzei), 2008 (Loope) y 2016 (Jensen). Desde 1989, el rango de elevación máxima de Azteca ha aumentado 132 metros. La elevación promedio de todas las especies de Azteca y Cecropiatambién ha aumentado significativamente desde 1989. Encontré hormigas Azteca en C. polyphlebia a 1500 m, a pesar de que históricamente no albergaba a Azteca. Estos datos indican que las hormigas están utilizando árboles en elevaciones más altas y habitando diferentes especies de Cecropia en Monteverde, potencialmente debido a cambios en el clima.

Elevational range of Azteca and Cecropia Garfink 2

Climate change is affecting species distributions and ecological interactions globally (Welch 2017). Often, these interactions are mutualisms, where both species benefit from the relationship. Mutualisms can be obligate, when one organism cannot survive without the other, or facultative, where both species are not dependent on each other to survive (Jorgensen et al. 2008). Many observed changes in ecological relationships have been correlated to climate change, including the mutualism between Azteca ants and Cecropia trees in Monteverde, Costa Rica (Mazzei 2003, Loope 2008, Jensen 2016). Cecropia are native to lowland, and mid- elevation cloud forests in the American tropics. They grow alongside forests, forest edges, pastures, and residential areas (Longino 1991). Eighty percent of Cecropia are myrmecophytic: plants which lives in a mutualistic relationship with ants (Berg et al. 2005). In this mutualism, ants protect against herbivores and kill vine ends that begin to attach to the Cecropia trunk. Cecropia provide shelter and food from glycogen-rich Mullerian bodies found on the stems of Cecropia leaves (Janzen 1969). Azteca in this mutualism are obligate and therefore cannot survive without nesting in Cecropia wood (Berg et al. 2005). The four species of Azteca found in Monteverde are A. constructor, A. xanthrochroa, A. alfari, and A. coeruleipennis (Longino 1989). There are four species of Cecropia found in the Monteverde area: C. peltata, C. obtusifolia, C. insignis, and C. polyphlebia (also known as C. angustifolia) (Longino 1989). C. peltata is found on the Pacific slope in lower elevations. C. insignis can be found in lowland regions of the Caribbean slope and C. obtusifolia can be found in mid-elevations on both slopes. Typically, non-myrmecophytic Cecropia (C. polyphlebia) is found at higher elevations where it can survive without the ant mutualism (Zuchowski 2007). Azteca are limited by cold, wet environments (Longino 1991). For this reason, C. polyphlebia historically do not partner with Azteca. Instead, C. polyphlebia relies on secondary compounds to defend from predators (Longino 1989, 1991, Jensen 2016). For over 30 years, non-continuous data has been collected in Monteverde comparing Cecropia and Azteca ranges at varying elevations: a primary study from 1989 (Longino on Azteca and Cecropia), and replicate studies in 2003 (Mazzei on Azteca), 2008 (Loope on Azteca and Cecropia), and 2016 (Jensen on Azteca and Cecropia). Comparing these studies, trends show that Azteca are moving up in elevation in Monteverde over the past 30 years. While there is no known cause for these changes, many studies on ecological interactions have been attributed to climate change. To track elevational shifts in Monteverde, I asked the questions (1) what is the current distribution of Azteca and Cecropia, and (2) how does this compare to previous studies? To answer this, I surveyed Azteca and Cecropia at varying altitudes in Monteverde and compared the results to previous work.

MATERIALS AND METHODS

I collected samples of Cecropia and Azteca from 1000-1800 m on the Caribbean and Pacific slopes in Monteverde, Costa Rica between 6 May 2019 and 19 May 2019. I sampled along trails near the San Luis Valley (1000-1200 m.a.s.l.), San Gerardo (Caribbean slope, 1100- 1300 m), La Calandría (1200-1300 m), Cerro Plano (1400-1500 m), Bajo del Tigre Reserve (1400 m), the Estación Biológica forest (around 1500-1600 m), and Cerro Amigos (1600-1800 Elevational range of Azteca and Cecropia Garfink 3

m). I identified Cecropia in the field with the help of plant expert and Monteverde local, Eladio Cruz, and with additional assistance from the Tropical Plants of Costa Rica (Zuchowski 2007). In the appendix of this paper there are images and notes of Cecropia that I identified for future reference. Also included are tables of my collected data, satellite imagery of the locations surveyed, and the averages, maximums, minimums, and sample sizes from past data. I measured tree height with a Nikon Forestry Pro laser range finder. I then calculated tree height accuracy as +/- 0.37 m by taking the same measurement from one tree 10 times and calculating the standard deviation. I recorded the diameter at breast height with a tape measurer for each tree. Altitude was recorded at each sample with an Oregon Scientific altimeter. I knocked on Cecropia trunks and used an extendable tree trimmer to observe ant activity. If present, 3-4 ants were collected using an aspirator and placed in ethanol for preservation. Ant species were identified in lab with the help of University of Utah professor and ant expert, John Longino. I classified ants with setae on the hind tibia as A. alfari. Ants with a flat, shelf-like metanotal groove, were identified as A. coeruleipennis. Other ants were classified as either A. xanthrocroa or A. constructor. Physical and ecological differences (such as nesting in Cecropia) between the two species are insignificant (Longino pers. comm. 2019), so A. xanthrocroa and A. constructor were combined into the group A. xanthrocroa/constructor for statistical analysis. I produced box plots of the elevational ranges of Azteca and Cecropia in Microsoft Excel. I also made scatter plots with regression lines of the average elevations of all Azteca and Cecropia species per year. A scatter plot and linear regression of maximum elevational height for all Azteca by year was also produced.

RESULTS

Sixty one Cecropia were sampled between 1000 m and 1800 m. Of the 61 trees, 37 were C. obtusifolia, four were C. peltata, 19 were C. polyphlebia, and one was C. insignis. Forty-two trees of Cecropia contained Azteca. Of the 42 samples, 36 were identified as A. xanthrocroa/constructor, 5 were A. alfari, and 1 was A. coeruleipennis. No trees contained more than one species of ant, except for six individual C. obtusifolia trees which were also being harvested by leaf cutter ants. Azteca spp. were found throughout the sampling area below 1532 m. A. xanthrocroa/constructor had the largest range: 1035-1532 m. A. alfari was found between 1037 m and 1240 m. One A. coeruleipennis was identified and found at 1087 m (Fig. 1). C. obtusifolia was most common throughout the survey. C. obtusifolia was found above 1035 m and below 1532 m. C. peltata was found between 1085-1239 m. One C. insignis was sampled along the distribution and was found at 1238 m on the Caribbean slope. C. polyphlebia was found at higher elevations, yet had the largest range of 564 meters (from 1238-1802 m) (Fig. 2). Five C. polyphlebia samples contained A. xanthrocroa/constructor.

Elevational range of Azteca and Cecropia Garfink 4

Fig. 1: Elevational ranges of Azteca ants by species in 2019. Average altitude per species is represented by the horizontal black lines. Minimum and maximum altitude per species is represented through the vertical black lines. (A. xanthrocroa N=35, A. alfari N=5, A. coeruleipennis N=1).

Fig. 2: Elevational ranges of Cecropia trees by species in 2019. Average altitude per species is represented through the horizontal black lines. Maximum and minimum altitude per species is represented through vertical black lines. (C. obtusifolia N=35, C. peltata N=4, C. polyphlebia N=19, C. insignis N=1) Elevational range of Azteca and Cecropia Garfink 5

In comparison with previous studies, the average elevation of A. xanthrocroa/constructor has increased by 172.5 m since the original 1989 study by Longino (Fig. 3). A. coeruleipennis has increased by 251 m (Fig. 4). A. alfari has increased by 271 m, the most since 1989 (Fig. 5). Average elevation of C. obtusifolia has increased 172 m since 1989 (Fig. 6) and C. peltata has increased 283.5 m (Fig. 7). All documented Cecropia and Azteca have increased average elevational range since 1989. Maximum elevation of all Azteca measured in Monteverde is 1532 m, 132 m higher than in 1989 (Fig. 8). The relationship between year and altitudinal height is strongly correlated and increasing (R2 = 0.983) (Fig. 8).

Fig. 3: Average elevational height in meters of Azteca xanthrocroa/constructor in 1989 (Longino), 2003 (Mazzei), 2008 (Loope), 2016 (Jensen), and 2019 (Garfink).

Fig 4: Average elevational height in meters of Azteca coeruleipennis in 1989 (Longino), 2003 (Mazzei), 2008 (Loope), 2016 (Jensen), and 2019 (Garfink).

Elevational range of Azteca and Cecropia Garfink 6

Fig. 5: Average elevational height in meters of Azteca alfari in 1989 (Longino), 2003 (Mazzei), 2008 (Loope), 2016 (Jensen), and 2019 (Garfink).

Fig. 6: Average elevational height in meters of C. obtusifolia in 1989 (Longino), 2003 (Mazzei), 2008 (Loope), 2016 (Jensen), and 2019 (Garfink).

Elevational range of Azteca and Cecropia Garfink 7

Fig. 7: Average elevational height in meters of C. peltata in 1989 (Longino), 2003 (Mazzei), 2008 (Loope), 2016 (Jensen), and 2019 (Garfink).

Fig. 8: Maximum altitude in meters of Cecropia-inhabiting Azteca spp. in Monteverde in 1989 (Longino), 2003 (Mazzei), 2008 (Loope), 2016 (Jensen), and 2019 (Garfink).

Elevational range of Azteca and Cecropia Garfink 8

DISCUSSION

There is a clear increase in the elevational distribution of Azteca in Cecropia in Monteverde over the past thirty years. While the elevational height difference from 2016 to 2019 is small, there is a strong overall trend in average elevation in all species of Azteca that occupy Cecropia in Monteverde. There are four possible explanations for how Azteca and Cecropia are distributed: (1) habitat characteristics determine Azteca distribution, (2) habitat characteristics determine Cecropia distribution, (3) Azteca influences Cecropia distribution, (4) Cecropia influences Azteca distribution (Longino 1991). In all possible explanations, climate is a leading contributor to distribution. Temperature, rainfall, and seasonality in explanations 1 and 2 can disturb Azteca/Cecropia habitats and thus alter their distributions. Cold, wet environments typically determine the upper limit for Azteca range (Longino 1991). Increasing annual temperature and decreasing cloud cover in high elevations of the cloud forest could create new inhabitable trees for Azteca who were previously limited by heavy rains. Explanations 3 and 4 show how ecological relationships can change due to succession and chemical defense alterations (Longino 1991). Changes in atmospheric composition (particularly higher levels of CO2) can influence plant metabolism and decrease the production of secondary compounds (Alnsour et al 2015). A lack of secondary compounds in C. polyphlebia due to changes in atmospheric composition could explain in the presence of Azteca at high elevations. It is also possible that Azteca are experiencing high levels of competition at lower elevations. If new inhabitable Cecropia are available at higher elevations due to changing temperature and weather patterns, it is possible that Azteca could shift range to relieve competitive pressures. It could be argued that Cecropia and Azteca are going up in elevational range due to urbanization and deforestation in the Santa Elena and Cerro Plano area of Monteverde (since Cecropia are commonly found on forest edges and pastures). If land in Monteverde was being deforested for road building or farming construction, it is likely that Cecropia would be shifting range (with Azteca following). This, however, is not the case. Monteverde has significantly decreased deforestation since the late 1970’s (Pounds 1991). The Monteverde Conservation League has planted hundreds of thousands of trees to serve as windbreaks, corridors, and habitat (Nadkarni et al. 2000). For this reason, it is unlikely that the Azteca are moving in elevation due to deforestation and road building. The year 2019 was drier than most in Costa Rica due to the El Niño phenomenon which brings dry, high pressure air to the Tropics. This, however did not impact the results of this study since the upwards trend of Azteca elevational range has been occurring over the past 30 years (or more), not just in 2019. Earth’s annual average temperature has increased 0.8°C since pre-industrial times (NASA 2010). The Intergovernmental Panel on Climate Change (IPCC 2014) expects global temperatures to rise an additional 2.3-2.9°C by 2100 under “business as usual” conditions (IPCC 2014). Models predict a 600 m upslope trend in temperature isotherms in the next few decades in tropical mountain ranges (Smith et al. 2014). A shift this large could be detrimental to ant species in Costa Rica, since many tropical ant species have ranges of 600 m in total (Smith et al. 2014). Research on invertebrate species could lead to a better understanding of the impact of climate change. I would suggest that future studies barcode Azteca found in Cecropia to make identification easier for future projects. Identifying Azteca is notoriously difficult and using barcoding would give a clear, definite, identification for tracking species ranges. Gathering GPS Elevational range of Azteca and Cecropia Garfink 9

information at each Cecropia would also make tracking Cecropia and Azteca colonies easier. Future studies should also take note on the presence or absence of vines/other invading plant species on Cecropia. I found that most Cecropia in the Cerro Amigos area without Azteca had dying leaves and were covered in vines. Since C. polyphlebia in the 1500 m range were found containing Azteca, further research could potentially study the presence or absence of secondary compounds in these trees which usually protect from vines and predators. Continuing with studies on the changing elevational range of Azteca and Cecropia could help solidify triggers for the range shifts and prevent further changes.

ACKNOWLEDGEMENTS

I would like to thank my primary advisory, Emilia Triana, for her endless support and encouragement during my research. Emi was always there to help me with paper editing, drive me to far away Cecropia trees, and laugh at me as I tried to suck up Azteca with my broken aspirator. I would also like to thank Frank Joyce, my secondary advisor and excellent professor, for encouraging and pushing me to produce my best work (and for wheeze laughing at my jokes). I would also like to thank Eladio Cruz for his immaculate plant knowledge, and Jack Longino for his time and patience in teaching me how to identify ants. Finally, thank you to my classmates at EAP for listening to me frantically talk about Azteca during all hours of the day (especially those who heard it at 5am in the study room). Thank you.

LITERATURE CITED

Alnsour, M.; and Ludwig-Muller, J. 2015. “Potential effects of climate change on plant primary and secondary metabolism and its influence on plant ecological interactions.” Journal of Endocytobiosis and Cell Research, vol. 26, pp. 90-99 Berg, C.C.; Rosselli, P. F.; and Davidson, D.W. 2005. “Cecropia” Flora Neotropica.” New York Botanical Garden Press, vol. 94, pp. 1–230. Haber, W.A.; Zuchowski, W.; and Bello, E. 1996. An Introduction to Cloud Forest Trees. P. 53. Heikkinen, N. “Climate Change Could Alter Interactions among Species”. Scientific American. July 5, 2016. Web. IPCC, 2014. “Climate Change 2014: Synthesis Report. Contribution of Working Groups I, II and III to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change” [Core Writing Team, R.K. Pachauri and L.A. Meyer (eds.)]. IPCC, Geneva, Switzerland, 151 pp. Janzen, D.H. “Allelopathy by Myrmecophytes: The Ant Azteca as an Allelopathic Agent of Cecropia.” Ecology, vol. 50, no. 1, 1969, pp. 147–153. Jensen, I. 2016. “Climate change and elevational range shifts by mutualistic Azteca (Hymenoptera: Formicidae) and Cecropia (Urticaceae) in Monteverde, Costa Rica.” CIEE Tropical Biodiversity and Conservation Spring 2016. Jorgensen, S.E.; Fath, B.D. 2008. “Encyclopedia of Ecology.” Web. “Key to Costa Rica Azteca Workers”. AntWiki. N.p. April 8, 2016. Web. Longino, J.T. 1989. “Geographic Variation and Community Structure in an Ant-Plant Mutualism: Azteca and Cecropia in Costa Rica.” Biotropica, Vol. 21, No. 2, pp. 126-132. Elevational range of Azteca and Cecropia Garfink 10

Longino, J.T. 1991. “Taxonomy of Cecropia-inhabiting Azteca ants”. Journal of Natural History, Vol. 25, pp. 1571-1602 Loope, G. 2008. “Range change in Cecropia and Azteca: the effects of climate change on mutualistic partners in Monteverde, Costa Rica.” CIEE Tropical Biodiversity and Conservation Spring 2008. Mazzei, P. 2003. “Effect of climate change on four species of Azteca in Monteverde, Puntarenas, Costa Rica.” CIEE Tropical Biodiversity and Conservation Fall 2003. Pounds, J.A., M.P.L. Fogden, and J.H. Campbell. 1999. Biological response to climate change on a tropical mountain. Nature 398: 611-615. Smith, A.M.; Hallwachs, W.; and Janzen, D.H. 2014 “Diversity and phylogenetic community structure of ants along a Costa Rican elevational gradient.” Ecography, Vol. 37, pp. 720- 731. Welch, C. “Half of All Species Are on the Move-And We’re Feeling It”. National Geographic. April 27, 2017. Web. “World of Change: Global Temperatures”. NASA Earth Observatory. N.P. December 2010. Web. Zuchowski, W. 2007. Tropical Plants of Costa Rica: A Guide to Native and Exotic Flora. P. 45- 48

APPENDIX

Locations of land surveyed:

Elevational range of Azteca and Cecropia Garfink 11

Elevational range of Azteca and Cecropia Garfink 12

Elevational range of Azteca and Cecropia Garfink 13

Descriptions of Cecropia identified (Eladio Cruz pers. comm. 2019, Zuchowski 2007):

Cecropia insignis

Leaves: large, large/thick lobes. Each leaf has 10 lobes or less Inflorescences: medium length (6-18cm) Location: mostly on wet Atlantic slope Other notes: fallen leaves tend not to curl up

Elevational range of Azteca and Cecropia Garfink 14

Cecropia obtusifolia

Leaves: Typically have 10 or more lobes. Secondary veins are more spread apart than polyphlebia. Inflorescences: long and skinny. Typically, more than 50cm long Location: found on both slopes. Moist, wet habitat Other notes: most likely to be confused with C. polyphlebia

Elevational range of Azteca and Cecropia Garfink 15

Cecropia peltata

Leaves: white underside. Leaf has 11 lobes or less Inflorescences: short, fat fruit spikes (4-10cm long) Location: mostly dry, low elevations Other notes: most likely to be confused with C. insignis

Elevational range of Azteca and Cecropia Garfink 16

C. polyphlebia (C. angustifolia)

Leaves: venation is closer together than C. obtusifolia. Secondary veins and new leaf bracts are reddish in color. Leaves have 10 or more lobes Inflorescences: fruit spikes are short and resemble a sea-urchin Location: on both slopes (Pacific and Caribbean) in elevations higher than 1200m. Often found without Azteca ants. Other notes: most likely to be confused with C. obtusifolia

Elevational range of Azteca and Cecropia Garfink 17

Mean, max, min altitude (meters), and total number found of Azteca and Cecropia in 2019, 2016, 2008, 2003, and 1989.

Garfink 2019 Mean altitude Max altitude Min altitude Total number (meters) (meters) (meters) found A. xanthrocroa/A. 1336.5 1532 1035 36 constructor A. coeruleipennis 1087 1087 1087 1 A. alfari 1187 1240 1037 5 C. obtusifolia 1313 1532 1035 37 C. peltata 1143.5 1239 1085 4 C. insignis 1238 1238 1238 1 C. polyphlebia 1562 1802 1238 19

Jensen 2016 Mean altitude Max altitude Min altitude Total number (meters) (meters) (meters) found A. xanthrocroa/A. 1309 1530 765 55 constructor A. coeruleipennis 873 1430 720 14 A. alfari 959 1300 725 7 C. obtusifolia/C. 1398 1770 940 35 angustifolia hybrid C. obtusifolia 1261 1530 865 44 C. peltata 821 1100 720 19

Loope 2008 Mean altitude Max altitude Min altitude Total number (meters) (meters) (meters) found A. xanthrocroa/A. 1211 1450 815 79 constructor A. coeruleipennis 793 975 720 22 A. alfari 1016 1280 725 15 C. obtusifolia/C. 1259 1610 890 34 angustifolia hybrid C. obtusifolia 1179 1450 725 60 C. peltata 835 1135 720 31

Mazzei 2003 Mean altitude Max altitude Min altitude Total number (meters) (meters) (meters) found A. xanthrocroa/A. 1235.5 1450 925 51 constructor A. coeruleipennis 840 980 745 10 A. alfari 997 1310 780 19

Longino 1989 Mean altitude Max altitude Min altitude Total number (meters) (meters) (meters) found Elevational range of Azteca and Cecropia Garfink 18

A. xanthrocroa/A. 1164 1400 860 46 constructor A. coeruleipennis 836 940 730 20 A. alfari 916 1240 750 27 C. obtusifolia 1140 1400 770 57 C. peltata 860 1250 730 40

Elevational range of Azteca and Cecropia Garfink 19

Data (Garfink 2019):