Effects of Climate on Rodent Abundance Parker 1 Possible e ffects of l ong t erm w eather c hange s and e xtreme w eather e vents on m ice p opulation a bundance s in Monteverde, Costa Rica Emily S. Parker Department of Ecology and Evolutionary Biology University of California, Santa Cruz EAP Tropical Biology and Conservation, Fall 2017 15 December 2017 ABSTRACT Mice and rat abundance is determined by a multitude of factors , including , but not limited to : climate change, predation, and food availability. In this study, I analyzed the effects of daily and long term weather on rodent populations to u nderstand a possible decrease in abundance from 2016 to 2017 in Monteverde, Puntarenas, Costa Rica. I took obs ervations of individual s from four different sites: Santa Elena , Bajo del Tigre , Dwight and R achel Crandell Memorial Reserve , and the EstaciÃ³n BiolÃ³gica Monteverde. While a considerable number of specimens were captured in Bajo del Tigre, I observed the to tal abundance of rodents to be significantly lower than that observed one year prior in the fall of 2016 in the same sites . Due to this disparity, I hypothesized that climate change and weather could be affecting the populations. I obtained rainfall data taken from the time period around each study period and monthly totals of both 2016 and 2017 . I tested this data as a control against the number of rodents caught for each respective study. For both years, the amount of daily rainfall was not statistically correlated to the number of rodents captured daily. These results indicated that low catch rates were not caused by daily rainfall in 2017; therefore, it is possible the population levels have been decreasing over the past year. I went on to explore new hypotheses of long term climate change, coati interference , and food availability as possible explanations for a decreasing rodent populati on. Posibles efectos de los cambios climÃ¡ticos a largo plazo y eventos climÃ¡ticos extremos sobre la abu ndancia de ratones en Monteverde, Costa Rica RESUMEN La abundancia de roedores est Ã¡ determinada por muchos factores, incluyendo, pero no limitados a cambio clim Ã¡ tico, depredaciÃ³n , y disponibilidad de alimentos. En esta investigaciÃ³n, analicÃ© los efectos del clima diario y del clima a largo plazo en poblaciones de roedores, para entender una posible disminuciÃ³n en su abundancia del aÃ±o 2016 al 2017 en Monteverde Puntarenas, Costa R ica. Propuse muestrear ratones en cuatro sitos: Santa Elena, Bajo del Tigre, Reserva Dwight y Rachel Crandell, y EstaciÃ³n BiolÃ³gica Monteverde. Mientras que capturÃ© una cantidad considerable de roedores en Bajo del Tigre, observÃ© una abundancia total de ro edores significativamente menor que la encontrada el a Ã± o anterior en los sitos mismos. Debido a esta disparidad hipoteticÃ© que cambios en clima y tiempo podrÃan afectar las poblaciones de roedores. Obtuve los datos de precipitaciÃ³n para el per iodo correspo ndiente a ambas investigaciones, asÃ como total de lluvia mensual de 2016 y 2017. ContrastÃ© los datos de lluvia con el nÃºmero de roedores atrapados en cada estudio. Para ambos a Ã± os, la cantidad de lluvia no se correlaciona con la cantidad de roedores captu rados diariamente. Estos resultados indicaron que la lluvia diaria no fue la causa de bajas tasas de captura en 2017; y es posible que las poblaciones han disminuido
Effects of Climate on Rodent Abundance Parker 2 en Ãºltimo a Ã± o por otros motivos . Por lo tanto, explorÃ© nuevas hipÃ³tesis considerando cambi o de clima a largo plazo, interferencia de pizotes, asÃ como la disponibilidad de comida como posibles explicaciones a la disminuciÃ³n de pequeÃ±os roedores . Wild life stability is often at the mercy of factors outside what we consider to be in ecosystems. The basis of community ecology is that populations along with abiotic factors directly affect other populations . The s easonal or yearly population abunda nce fluctuations of different groups of fauna are a common and natural phenomenon. The effects from climate, food abundance, habitat stability, hum an interaction, competition, reproduction rate , and predation are just a few of the major factors that determine the population sizes from one cohort to the next. In population ecology, predation is often used as the main model for a trophic cascade a ffecting different prey species. However, it has been hypothesized that the over a given prey population is a faÃ§ade, and the true control of both predator and prey populations is bottom up e ffects f rom abiotic factors (White 2013). Ecological theory indicates that climate and weather can positively or negatively affect habitat and food availability. Researchers have gone on to determine climate to be a str onger direct or indirect factor than predation or anthropogenic influences in determining rodent population levels (Allen et al. 2018; Sipos et al. 2 017). According to previous studies done in Northern Europe, the effects of climate change, particularly the changes in snow fall and winter weather, play role s in the oscillating pattern of rodent abundance (Kausrud 2008) . In Costa Rica, the climate is generally consistent; the tropics have relatively non fluctuating temperatures. While the north western province of Guanacaste is subject to a dry season due to the North East Trade Winds. Monteverde, Puntarenas is further inland and higher in elevation , and therefore does experience high wind speeds, but does not greatly suffer from extreme lapses of water. This allows for Monteverde to stay green year round . However, climate change is a prominent force affecting ecosystems all ove r the planet in noticeable ways. O ne of the most evident consequences is the increase in the number and intensity of hurricane s and tropical storms t hat are reaching landmasses, causin g massive amount s of destruction to both human and wildlife habitats (Trenberth 2011) . R ecently, Hurricane Nate (classified as a tropical storm while affecting Costa Rica) brought down l arge amounts of trees and caused multiple destructive landslides due to he avy rains and high wind speeds. These factors could contrib ute to an environment not conduc ive to high dens ities of rodents that depend on near tree burrows for habitat a nd large amounts of fruits . While storms like Nate are clear evidence of climatic disturbance, r esearcher J. Al a n Pounds presented evidence , during a n Education Abroad P rogram night talk , that dry days are getting drier and the rainy days are getting wetter in Costa Rica . This is also supported by other research analyzing how the warming of the atmosphere is allowing a higher water holding capacity, which can cause extrem es of both dry and wet regions (Trenberth 2011). This indicates that the long term weather in Costa Rica is most likely experiencing changes from climate change. W hilst collecting data for preliminary observations, I found capturing rodents in Monteverde to be difficult . This was both strange and interesting, as previous data indicat ed that similar sites were once home to high d rodents (Chinchilla 2009; Thoene 2016) . As we learn more about climate change and its e ffects on daily and seasonal weather, we begin to consider these factors when understanding why there appears to b e a lapse in rodent
Effects of Climate on Rodent Abundance Parker 3 populations in the Monteverde region. It raises the question: are rodent populations in Monteverde affected by localized climate change and extreme climatic events ? MATERIALS AND M ETHODS Field Data I established locations at four sites of varying distances from each other to analyze different populations of rodents . Site 1 was located in Santa Elena (SE) in the forested area near the MegaSuper market. Site 2 was in the neighborhood of Bajo del Tigre (BT) near the home of Frank Joyc e. Site 3 was in the Dwight and Rachel Crandell Memorial Reserve (CR) , and Site 4 was deep into the trails of the EstaciÃ³n BiolÃ³gica Monteverde (EB) ( Map 1 ) . I rotated between sites each night to minimize daily travel time and maximize attention to detail for each data set per site. I set 17 Sherman traps each day in the afternoon around 3:00 PM and collected them the following morning at 8:00 AM. I repeated this for th ree rotations , beginning 20 November 2017 and setting traps every night through 1 December 2017. On the nights of 20 November and 25 November, another student researcher joined me and there were respectively 36 and 35 traps set those nights. Map 1. Monteverde area indicating sites of study. I set each trap by the base of a tree in an area with leaf litter, usually nearby a burrow or hole, as the rodents like to forage and nest there. Each site contains this type of terrain , having trees and other f lora that could provide stable habitat and act as food sources . I set the traps approximately two to three meters (around 10 15 steps) apart from each other, each with a
Effects of Climate on Rodent Abundance Parker 4 mixture of rice, oat, and vanilla as bait . Lastly, I tied fluorescent flagging onto a living substrate (tree, shrub, etc.) to help keep track of every trap. In the mornings, I traveled to the respective sites and began to collect data. I firstly noted the condition of the trap s to observe if there had been any outside interference from other animals or abiotic factors. Next, I collected the traps, folding down the unsuccessful, and began to process those with a specimen. For each specimen, I placed the animal into a cloth sack. Using a measuring tape, I took me asurements of the length of the foot from base to tip of the longest toe and the tail ; I also took the weight of the specimen using a hanging scale. These physical measurements were taken to gain an understanding of the morphologies of the rodents caught a nd to use for individual differentiation . Lastly, I trimmed a section of fur on the lower back of the animal as a marker as to avoid collecting data from the same individual twice. Finally, I released the all animals to their site of capture and cleaned all equipment used. Data Analysis Along with my own data collection, I obtained data from previous studies done by Chinchilla (2009) and Thoene (2016) pertaining to rodent populations . I also obtained precipitation data for Monteverde from th e Monteverde Institute and the personal records of Frank Joyce and Kat y VanDusen for the current year and past years. I studied if the amount of rodents caught over a period of time varied with the amount of precipitation that occurred during the time of the catch. To analyze the effect of rain on this sample, I determined the amount collected from the day the traps were set and the day the traps were checked for specimens. I then graphed these values as a control variable to the number of rodents captured. Furthermore, I analyzed the total monthly rainfall for April through November of 2016 and 2017. R ESULTS Overall, I set tr aps for 241 trap nights, with a total of 14 captured rodents over the four s ites (5.81% capture rate) . My sample cons isted of three different species: Peromyscus mexicanus (Mexican Deer Mouse; MDM ; n=2 ), Oligoryzomys fulvescens (Pigmy Rice Rat; PRR ; n=4 ), and Heteromys nubicolen s (Spiny Pocket Mouse ; SPM ; n=8 ). Three individuals were caught as recaptures : t wo Mexican Deer Mice, of which are not included in this data set, and one Spiny Pocket Mouse, caught by a peer researcher on a separate catch night, which is included. I captured eight females, one male, and two individuals whose sex could not be identified. I was only successful in catching mice at two sites, Santa Elena and Bajo del Tigre (Figure 1) . However, at the Santa Elena site only one specimen w as obtained (1.96% catch rate) , and the majority were captured at Bajo del Tigre (14.77% catch rate) . For every trap night at Santa Elena , the traps appeared to have been de baited by rain or wind upon observations the next morning. The traps at the Crandell Reserve appeared to be unaffected by outside wildlife . However, at the EstaciÃ³n BiolÃ³gica Monteverde , it appeared that coatis ( Nasua narica ) possibly tampered with the traps, as they were found empty, broken, and at far distances from their origi nal placement site (Appendix Photo B) . Furthermore, fellow student researchers found coatis tampering with traps at their sites ( Appendix Photo A ). Two traps were lost during the last trap night at the EstaciÃ³n BiolÃ³gica (2 December 2017) .
Effects of Climate on Rodent Abundance Parker 5 Figure 1. Number of rodents caught per site in 2016 versus 2017 (Thoene 2016) . Note that Santa Elena was not a site for the Thoene Thoene (2016) similarly gathered data in Bajo del Tigre, Crandell Reserve, and at the EstaciÃ³n BiolÃ³gica Monteverde and obtained a larger sample size ( n=34; 16.92% capture rate) (Figure 1) . In 2009, Chinchilla condu cted a rodent based study at the EstaciÃ³n BiolÃ³gica Monteverde and simila rly found the three species of rodents that I found in Bajo del Tigre , along with multiple other species . H is study found these species to be in higher abundances and at the EstaciÃ³n BiolÃ³gica (Table 1) . Table 1. The % of catches based on the number of successful traps at the EstaciÃ³n BiolÃ³gica Monteverde in 2009 and in 2017 . Key created by Federico Chinchilla: A=Abundant (80% <); C= Common (40% 80%); UC= Uncommon (10% 40%); R= Rare (10%>) ; Not Seen (0%) (Chinchilla 2009). Species F all 2017 2009 Peromyscus mexicanus Not Seen A Oligoryzomys fulvescens Not seen R Heteromys nubicolens N ot Seen U Furthermore, Thoene also consisted of higher abundance and species diversity; it included the three species captured in this study, along with Scotinomys teguina , singing mouse ( Thoene 2016). I analyzed the rodent and rainfall data from 2016 and found a very low correlation between amount of rain and rodents caught (R 2 =0. 0 6279 ; Figure 2 ). Then, using the same analysis for the data from the fall of 2017, I again found there to be a very low correlation between the amount of rainfall and rodents caught (R 2 = 0.0335 ; Figure 3 ). 0 2 4 6 8 10 12 14 16 18 20 Santa Elena Bajo Tigre Crandell Reserve EstaciÃ³n BiolÃ³gica Amount of Rodents Captured Site 2017 2016 Does not include Santa Elena Site
Effects of Climate on Rodent Abundance Parker 6 Date # Rodents Site 06 Nov 16 3 EB M 22 Nov 16 0 CR 25 Nov 16 1 CR 26 Nov 16 4 BT 27 Nov 16 1 BT 28 Nov 16 4 BT 29 Nov 16 4 BT 30 Nov 16 2 BT 01 Dec 16 4 BT 04 Dec 16 11 EB M Figure 2. The catch data from 2016, and the correlation between the amount of rainfall on catch dates and the number of rodents caught in 2016 (Thoene 2016; Joyce et al. 2017). RÂ² = 0,0628 0 2 4 6 8 10 12 0 20 40 60 80 100 120 140 160 180 200 # Rodents Caught Amount of Rain mm
Effects of Climate on Rodent Abundance Parker 7 Date # Rodents Site 21 Nov 17 2 BT 22 Nov 17 0 CR 23 Nov 17 0 EB 24 Nov 17 1 SE 25 Nov 17 0 CR 26 Nov 17 5 BT 27 Nov 17 0 EB 28 Nov 17 0 SE 29 Nov 17 0 CR 30 Nov 17 6 BT 1 Dec 17 0 SE 2 Dec 17 0 EB Figure 3. The catch data from 2017, and the correlation between the amount of rainfall on catch dates and the number of rodents caught in 2017 (Joyce et al. 2017). Furthermore, I analyzed the monthly rainfall data for both 2016 and 2017 (Figure 4) . I then found that the average s for both years were not statistically different (p=0.47 3 ), but through the standard deviation , we can see that 2017 had higher r ain extremes than 2016 (Figure 5 ). Furthermore, it can be seen that the months of April, May, and October had large increase of total monthly rainfall from 2016 to 2017. RÂ² = 0,0335 0 1 2 3 4 5 6 7 0 2 4 6 8 10 12 14 16 18 20 # of Rodents Caught Amount of Rain (mm)
Effects of Climate on Rodent Abundance Parker 8 Figure 4. A comparison between the total monthly rainfall for April through November for 2016 and 2017 (Monteverde Institute 2017). Figure 5 . The total monthly average per year and the standard deviation. The averages are not statistic ally different (p= p=0.473 ), but the standard deviation for 2017 is larger than that of 2016. D ISCUSSION Overall, I caught 14 mice at test sites Bajo del Tigre , Dwight and Rachel Crandell Memorial Reserv e, Santa Elena, and the EstaciÃ³n BiolÃ³gica Monteverde . I n the past , these site s, excluding Santa Elena , have exhibited high rodent density. My original prediction was that an increased amount of rainfall in 2017 would cause a higher amount of daily rainfall and result in lower amounts of rodents caught . W hen graphed against each oth er, amount of nightly rainfall 0 100 200 300 400 500 600 700 800 900 1000 Total Amount of Rain (mm) Month 2016 2017 0 100 200 300 400 500 600 700 800 2016 2017 Average Monthly Rainfall (mm) Year 278.83 208.54
Effects of Climate on Rodent Abundance Parker 9 proved to have no significant correlation to how many rodents were caught on a given night for both 2016 and 2017. This means, something must have occurred between the fall of 2016 and the fall of 2017, as Thoene was able to capture specimens at sites where I was unable. These results bring up a new question: what factors, other than daily weather change, could have contributed One hypothesis still relates to weather, but in a long te rm form. By comparing the total amount of rainfall for each month , I was able to qualitatively understand how month month variation can occur drastically over time. I calculated the overall monthly r ainfall averages between 2016 and 2017 and found them to be statistically similar. However, when comparing data from April, May and October of both years , we can see that see that 2017 had a considerable increase in rainfall during these months. This bring s up the possibility that earlier in 2017 , increased rainfall could have contributed to decreasing food and habitat availability, leading to decreased rodent populations. During the summer months, the populations could have begu n to recover, but Tropical S torm Nate caused a huge increase in the rainfall (314.5 mm) compared to that of October 2016 (Figure 4). With the storm came massive amounts of destruction, seen in the many fallen trees and landslides found in the aftermath. Many rodents could have been i nitially killed during the storm by being caught in a landslide or another form of natural destruction caused by high amount of precipitation and winds speeds. Along with habitat destruction, food sources could have been drastically lost during the storm. Most rodents depend on seeds for their diet; it is possible that many were displaced and killed during the storm. When visiting the biological station at Pitilla only days after the storm, my study group qualitatively observed a huge decrease in the amount of seeds. A similar decrease occurred a fter Hu rricane Sandy hit Haiti in 2011; farmers and scientists observed that fr uit s were in much lower abundance s , as they were knocked off trees by high speed wind and heavy rainfall ( Kolbe et.al. 2012 ). While this data would need to be quantitatively verified for Monteverde after Nate , it supports the hypoth . If proven with further testing, the effects of the storm could be a logical and valid explanati on for the lack of rodents in this study. Something else to consider is the effect that other wildlife might have had on my study. At the EstaciÃ³n BiolÃ³gica Monteverde , I was unable to obtain any specimens over 51 trap nights. However, in Chinchilla and Th captured many individuals (Table 1; Table 2). This indicates that at the EstaciÃ³n there was, at one time, a high enough amount of Ot her than the storm, what else could have directly affected this study site to cause a decrease in the number of rodents caught ? One possible explanation is the high abundance of coatis at the EstaciÃ³n BiolÃ³gica Monteverde ( personal observation ) . These animals are omnivores and are known for eating food scraps left behind by humans. Every trap night done at this site resulted in a majority of the traps being found empty, broken, and/or moved (Appendix B ) . F ellow student researchers observed co ati s tampering with traps in Cu ri C ancha (Appendix A ) . According to Chinchilla, this high population of coatis is abnormal to the EstaciÃ³n BiolÃ³gica Monteverde (personal communication) . While this is a qualitative observation and there is no census data to su pport this claim, it could explain why this site was an unsuccessful site for catching rodents: the traps beco me ineffective if the coatis break and de bait the m before a rodent comes along. However, the presence of coatis does not explain the lack of rodents caught at the Crandell Reserve, as there does not appear to be any population of coatis there, or at least no evidence of
Effects of Climate on Rodent Abundance Parker 10 tampered traps. Still, this possibility of interference i s worth considering when understanding the catch rates of rodents in this study . Despite the evidence of wildlife interference, it is clear that long term weather effects and extreme climatic events, such as Tropical Storm Nate, could serve as determining factors in the hypothesis that climate change is affecting rodent populations. This study support s that hypothesis. However, ther e is not enough data to confirm why rodent populations have reduced, or at least have appear ed to be reduced. Further data mus t be collected year rou nd, regardless of the weather, and at a variety of sites. It would also be helpful for a future study to take census data on the approximate amount of fruits available to rodents throughout the year . Of course, this should all be done while keeping other community factors in mind, s uch as the effect of other wild life. Understanding how climate could directly or indirectly affect rodent populations could provide insight to how climate change will continue to affect mammal population s, faunal dispersal, and overall, entire ecosystems. ACKNOWLEDGMENTS First and foremost, I would like to thank Dr. Federico Chinchilla for being my primary advisory , helping me with my Spanish, and supplying me wit h endless ideas, knowledge, support , and laughs throughout this process. I would also like to thank Andr Ã© for his ideas and motherly support. I would also like to thank Dr. Frank Joyce and Katie Van Dusen for allowing me to use their home in Bajo del Tigre as a test site and for supplying me with rain data for 2016 2017. Furthermore, I would like to thank the folks at the Monteverde Institute for supplying me with lab space and field space, Luisa Moreno for supplying me with and M arvin Hidalgo at the EstaciÃ³n BiolÃ³gica Monteverde for allowing me to use the trails. I also would like to thank my host family, Oscar and Xeomy Fennell for housing me, feeding me, and supporting me through my research . stu dent researchers Noelle Pruett and Kathy Dao for their collaboration with s etting traps and sharing data, and Jeremy Giampaoli for his help reviewing this paper. L ITERATURE CITED Allen, B. L., Fawcett, A., Anker, A., Engeman, R. M., Lis le, A., & Leung, L. K. P. 2018 . Environmental effects are stronger than human effects on mammalian predator prey relationships in arid Australian ecosystems. Science of The Total Environment, 610 , 451 461. Combes, C. 1991 . Ethological Aspects of Parasite Transmission. Th e American Naturalist, 138 (4), 866 880. Chinchilla, F. 2009 . Seed predation by mammals in forest fragments in Monteverde, Costa Rica. Rev. Biol. Trop, 57(3), 865 877. Hillman, A. E., Lymbery, A. J., Elliot, A. D., & Andrew Thompson, R. C. 2017 . Urban environments alter parasite fauna, weight and reproductive activity in the quenda ( Isoodon obesulus ). Science of The Total Environment, 607 608 (Supplement C), 1466 1478. Kausrud, K. L., Mysterud, A., Steen, H., Vik, J. O., Ã˜stbye, E., Cazelles, B. , Stens eth, N. C. 2008 . Linking climate change to lemming cycles. Nature, 456 , 93. Kolbe, A., Puccio, M., Muggah, R. 2012 .
Effects of Climate on Rodent Abundance Parker 11 Institute: Strategic Note, 6, 1 13. McCain, C. M. 2004 . The mid domain effect applied to elevational gradients: species richness of small mammals in Costa Rica. Journal of Biogeography, 31 (1), 19 31. Mill s, J. N., & Childs, J. E. 1998 . Ecologic studies of rodent reservoirs: their relevance for human health. Emerging Infectious Diseases, 4 (4), 529 537. Reid, F. A. 1997 . A Field Guide to the Mammals of Central America and Southeast Mexico . New York: Oxford University Press, Inc. Sipos, J., Suchomel, J., Purc hart, L., & Kindlmann, P. 2017 . Main determinants of rodent population fluctuations in managed Central European temperate lowland forests. Mammal Research, 62 (3), 283 295. Thoene, D. 2016 . Immune Dimorphism in Wild Mice of Monteverde, Costa Rica. UCEAP Fall 2016. Thompson, R. C. A. 2013 . Parasite zoonoses and wildlife: One health, spillover and human activity. International Journal for Parasitology, 43 (12), 1079 1088. Trenberth, K. 2011. Changes in precipitation with climate change. Climate Research, 47 (1/2), 123 138. White, T. C. R. 2013 . Experimental and observational evidence reveals that predators in natural environments do not regulate their prey: They are passengers, not drivers. Acta Oecologica, 53 (Supplement C), 73 87.
Effects of Climate on Rodent Abundance Parker 12 APPENDIX Photo A . Evidence of Coatis tampering with Sherman traps. Photo taken by Kathy Dao taken November 2017. Photo B . Broken trap as a result of Coati tampering. Photo taken by Federico Chinchilla on 2 December 2017.