The effects of reforestation on mammal diversity and abundance Pruett 1 The effects of r eforestation on mammal d iversity and a bundance in Monteverde, Costa Rica Noelle Pruett University of California, Santa Barbara Department of Ecology, Evolution, and Marine Biology EAP Tropical Biology and Conservation Program, Fall 2017 15 December 2017 ABSTRACT Reforestation is an important mechanism in helping previously deforested landscapes recover . However, reforested areas tend to have less diversity in vegetation, due to the fact that humans are not able to replicate the same diversity or pattern of regeneration that would occur naturally. I investigated whether or not these changes in fore st types and vegetation affect the diversity and abu ndance of mammals that inhabit the se area s . I used b oth camera traps and Sherman traps to survey mammals from areas with reforestation and with natural regeneration. I observed a total of 72 mammals from five species in the camera traps. I found that while there were no significant differences in the number of species present, there was a significantly larger number of individual mammal appearances o n camera traps in locati ons with natural regeneration. I captured a total of 28 individual rodents over five species in the Sherman traps . Though I found more individual rodents in areas with reforestation, there was not a significant difference between areas with reforestation and natural regen eration. These results indicate that three of the five larger mammal species observed in the cameras ( Nasua narica , Dasyprocta punctata, and Puma concolor ) were significantly more abundant in areas with natural regeneration than reforestation. Though more research is needed to determine the specific reasons these species are less abundant in reforested areas, I speculate that it is due to dietary changes. Efectos de la r eforestaciÃ³n en la diversidad y abundancia de mamÃferos en Monteverde, Costa Rica RESUMEN La reforestaciÃ³n es un mecanismo importante en ayudar a recuperar Ã¡reas previamente deforestadas. Sin embargo, las Ã¡reas reforestadas tienden a tener menor diversidad en la vegetaciÃ³n, debido al hecho de que los humanos no pueden replicar el mismo patrÃ³n de diversidad de la regeneraciÃ³n que ocurrirÃa naturalmente. InvestiguÃ© si estos cambios de tipos de bo sque y vegetaciÃ³n afecta la diversidad y abundancia de mamÃferos que habitan esas Ã¡reas. UtilicÃ© cÃ¡maras trampa y trampas Sherman para registrar los mamÃferos en bordes reforestados y bordes regenerados naturalmente. En las cÃ¡maras trampa registrÃ© en total 72 mamÃferos de cinco especies. EncontrÃ© que, aunque no hubo diferencias en el nÃºmero de especies presentes, sÃ hubo significativamente mayor nÃºmero de individuos en los sitios de regeneraciÃ³n natural. En las trampas Sherman registrÃ© 28 roedores de cinco especies. Aunque encontrÃ© mÃ¡s ratones en Ã¡reas reforestadas, no hubo diferencia entre Ã¡reas reforestadas y regeneradas. Estos resultados indican que tres de las cinco especies de mamÃferos observados en las cÃ¡maras ( Nasua narica , Dasyprocta punctata , y Pum a concolor ) fueron significativamente mÃ¡s abundantes en
The effects of reforestation on mammal diversity and abundance Pruett 2 Ã¡reas regeneradas que en las reforestadas. Aunque hace falta investigar mÃ¡s para determinar razones especÃficas de menor abundancia en Ã¡reas reforestadas, que en este estudio especulo se debe a requer imientos de las dietas. Human development has a large impact on natural biodiversity. This is often seen in the form s of habitat destruction or modificat ion as humans develop areas for their own use. Central America, specifically Costa Rica, has been hit particularly hard by d eforestation in the past. I n , deforestation reached a rate of 3.2% per year, which was the fifth highest rate in the world (SÃ¡nchez Azofeif et al. 2001). Though trends in deforestation have been reversed in the past thirty years , the prior deforestation left many forest fragments and forest edges ( SÃ¡nchez Azofeif et al., 2001). These areas are unable to support large amounts of bio diversity and tend to lack organisms that require large amounts of space (Galletti et al ., 2009). As Costa SÃ¡nchez Azofeif et al., 2001). it significantly changes the vegetation that would grow if the forests were able to regenerate naturally. Humans are not able to plant and allow for the spread of seeds the same way t hat would occur naturally. In reforested areas , there tends to be less bio diversity and a different species composition than in naturally regenerating forest s (Oosterhoorn et al., 2000). For the purposes of this study, reforestation is defined as areas in which humans have planted some vegetation, and natural regeneration is defined as areas that have been allowed to recover without human planting. Monteverde, Costa Rica, is a largely unique place in the sheer amount of biodiversity it hosts over such a sma ll area (Martin et al., 2016). This is largely due to its status as a cloud forest, a tropical forest at a high elevation, which draws many unique species. Intact forests are extremely important in supporting the diversity it hosts, as forest fragments and edges decrease diversity and are not able to support large mammals. Because of this, there have been some reforestation efforts in previously deforested areas (Lamb et al., 2005). While r eforestation is necessary to help preserve this diversity, some chan ges in the methodologies of reforestation may be needed in order to foster the same amount of diversity that would occur in a naturally regenerating forest (Mugwedi et al., 2017). A study done in the Andes in Colombia found that areas with reforestation ha d a lower abundance of medium sized mammals than natural forests, which indicates that mammals may be negatively affected by the differences in plant diversity in reforested areas (SÃ¡nchez et al., 2008). This study seeks to examine how forest edges refores ted by humans change the diversity and abundance of mammals. Based on the differences in vegetation and plant diversity, I expect lower mammal diversity and abundance in reforested areas than in areas regenerating naturally. MATERIALS AND METHODS Study Sites I observed mammal presence along a total of four edges with recovering forest in the del Tigre, which will be referred to as the Bajo del Tigre site for the p urposes of this project. The edges in Curi Cancha and Bajo del Tigre have been recovering by means of natural regeneration,
The effects of reforestation on mammal diversity and abundance Pruett 3 while both edges in La Calandria were reforested in 2003 (Hamilton 2017) . The edge in Bajo del Tigre was along a pasture that was cu t down approximately 60 years ago but has been allowed to recover relatively recently (Joyce 2017) . The edge in Curi Cancha was created approximately 50 years ago and has been regenerating since June 2009 (Ramirez 2017) . The first edge in La Calandria , referred to as La Calandria 1, had primarily Inga punctata and Ocotea s p p. planted. La Calandria 2, the second edge in La Calandria, was reforested using Ocotea floribunda and Ocotea monteverdensis (Hamilton 2017). Camera Traps I placed three Bushnell HD camera traps at equal intervals on trees along each recovering edge to monitor medium and large mammal presence . However, one of the traps in Curi Cancha was nonfunctional, so only two traps recorded data there . I chained and loc ked each trap to the tree to prevent theft. I set each camera with medium sensitivity approximately a half meter up the tree to take three pictures every time it detected motion. I checked the c amera traps and changed their SD Cards every five to six days. I included each mamm al presence more than a minute apart even if I saw the same mammal twice, as for smaller mammals like agoutis, I could not differentiate between individuals. I did not include d omestic animals, such as Canis familiaris and Felis catus , in the results. I set all the camera traps to record data between 22 November 2017 and 2 December 2017. Sherman Traps I placed seventeen to eighteen Sherman traps at site, each spaced evenly between ten and fifteen feet apart in alcove s where I expected to find rodents . I baited each trap with vanilla, oats, and rice, and rebaited them the next morning if necessary. I placed them at each site for three consecutive days and checked them each morning for rodents. If I found rodents , I recorded their weight , tail length, foot length, species, and gender . I then returned them to the location where I captured them . I used these measurements to help figure out which rodents were recaptured. I trimmed a section of their fur in order to recognize whether or not I recaptured rodents , as I did not include these in the results. I counted a rodent presence if I found a rodent in a trap or if I found part of the tail, but not the rodent, in the trap. I photographed r odents and identified them using The Mammals of Costa Rica (Wainwright 2007). When I captured rodents , I washed their traps before rese t t ing them . I set all the Sherman t raps between 20 November 2017 and 30 November 2017. Data Analysis I analyzed c amera trap data both in terms of the number of species and the num ber of individual appearances. I analyzed S herman trap data using the number of species and the number of rodents captured, excluding recaptures. I compared data between naturally regenerating and reforested areas, and between sites. I analyzed the results using Chi S quared tests. I used this because this tests for differences between distributions, so it was effective in testing the differences between the numbers of species and abundances.
The effects of reforestation on mammal diversity and abundance Pruett 4 RESULTS Camera Traps I set 11 traps for a total of 2134 hour between 22 November 2017 and 2 December 2017. The cameras in Curi Cancha Reserve recorded a total of 435 hours, while the cameras in Bajo del Tigre recorded a total of 602 hours, for a total of 1,037 hours in the sites with natural regeneration. The first site in La Calandria recorded a total of 440 hours, while the second recorded 656 hours, for a total of 1,096 hours in the sites with reforestation. They rec orded a to tal of five species of mammals: Nasua narica (White nosed coati) , Puma concolor (Cougar) , Dasyprocta punctata (Central American agouti) , Dasypus novemcinctus (Nine banded armadillo) , and Didelphis marsupialis (Common opossum) . There was a total o f 7 2 individual mammal appearances (Fig. 1). Figure 1 : Mammal Species and Individuals in Areas with Reforestation and Natural Regeneration At the sites with natural regeneration, I observed five species of mammals, with a total of 64 individual appearances (Fig. 1 ). The site in Curi Cancha had five species with 53 appea rances, and Bajo del Tigre had three species with 11 appearances. The sites with reforestation, both in La Calandria Reserve , recorded a total of eight individual appearances across three species. The f irst site in La Calandria had six appearances across three sp ecies, and the second site had two appearances from two species (Fig. 2). The difference in number of species between reforested and naturally regenerated sites was not significant , though the difference in the number of individual appearances was significant ( Fig. 1, Chi Squared test, p<0.01 , df=1 ). Curi Cancha had significantly more individual appearances than any other site ( Fig. 2, Chi Squared test, p<0.01 , df=3 ), though when Curi Cancha was excluded from the res ults, Bajo del Tigre still had significantly more individual appearances than the reforested sites ( Fig. 2, Chi Sq uared test, p<0.05 , d=2 ), which showed there were more individual appearances in each naturally regenerating site than in the reforested sites. 0 10 20 30 40 50 60 70 Reforested Natural Regeneration Number of Mammals Number of Species Number of Individual Appearances
The effects of reforestation on mammal diversity and abundance Pruett 5 Figure 2 : Mammal Species and Appearances Across the Different Site s Overall, areas with natural regeneration had more species present, though this was not statistically significant . However, there were significantly more individual appearances of N. narica , P. concolor , and D. punctata in naturally regenerating than in ref orested areas ( Figs. 3 & 4, Chi Squared test, p<0.05 , df=1 ) . Figure 3 : Species Distribution in Areas with Natural Regeneration and Reforestation 0 10 20 30 40 50 60 Bajo del Tigre Curi Cancha La Calandria 1 La Calandria 2 Number of Mammals Number of Species Number of Individual Appearances 0 5 10 15 20 25 30 Natural Regeneration Human Reforestation Number of Mammal Appearances Nasua narica Puma concolor Dasyprocta punctata Dasypus novemcinctus Didelphis marsupialis
The effects of reforestation on mammal diversity and abundance Pruett 6 Figure 4 : Species Distribution s Across the Different Site s Sherman Traps I placed a total of 212 Sherman traps over ten days, which caught a total of 28 rodents. I identified a total of five rodent species : Heteromys nubicolens, Tylomys watsoni, Oligoryzomys fulvescens, Scotinomys teguina, and Sigmodon h ispidus. Three species were seen in both areas with natural regeneration and reforestation, which was not a significant difference . N aturally recovering areas had ten individuals captured, while reforested areas had 18 (Fig. 5), which was also not a signif icant difference . Figure 5 : Rodent Species and Individuals in Areas with Natural Regeneration and Reforestation 0 5 10 15 20 25 Bajo del Tigre Curi Cancha La Calandria 1 La Calandria 2 Number of Mammal Appearances Nasua narica Puma concolor Dasyprocta punctata Dasypus novemcinctus Didelphis marsupialis 0 2 4 6 8 10 12 14 16 18 20 Number of Species Number of Individuals Number of Rodents Natural Regeneration Human Reforestation
The effects of reforestation on mammal diversity and abundance Pruett 7 La Calandria 2 had the most captures, with 12 individuals captured over three days. However, it also had the least diversity, with only one species ( H. nubicolens ) present. La Calandria 1 had the most diversity, with three species, and six total captures. Only two rod ents were caught in Curi Cancha, though each was a different species. Bajo del Tigre had eight rodents captured between two species (Fig. 6). Though I captured five species in total, H. nubicolens was the only species to be caught more than once, and was m ost abundant in areas with reforestation, especially La Calandria 2 (Fig. 7). The difference in number of species caught and individuals caught between sites was not significant . However, when I examined only H. nubicolens , there was a significant differen ce in the number captured over site ( Fig. 7, Chi Squared test, p<0.05 , df=3 ), though not in naturally regenerating versus reforested areas. Figure 6 : Rodent Distribution Across the Different Site s 0 2 4 6 8 10 12 14 Bajo del Tigre Curi Cancha La Calandria 1 La Calandria 2 Number of Rodents Number of Species Number of Individuals
The effects of reforestation on mammal diversity and abundance Pruett 8 Figure 7 : Rodent Species Distribution Across the Different Site s DISCUSSION Medium and Large Mammals Significantly more large mammals were photographed in areas with natural regen eration than in areas with reforestation , yet there was a no difference in the number of species present between these. I observed m ore species in the areas with natural generation, but as I only observed five species in total, it is difficult to generalize about the relative species diversities. However, the signif icant difference in the number of individual appearances based on the type of site indicates that sites with natural regeneration have higher abundances of mammals than reforested sites. This is likely due to the natural regeneration sites abilities to support a more complex diet and habitat with its more diverse plant s (Oosterhoorn et al., 2000). When I compared the individual appearance data between Curi Cancha and Bajo del Tigre, there was a significant difference between the two sites, with Curi Cancha having significantly more individual appearances. This was likely since it was the only site that was not a fragment and was connected to a larger, continuous forest. Because of this, I excluded it fro m the data and compared Bajo del Tigre to the sites with reforestation, which still showed a mammal abundance was likely increased by its connection to the larger forest, the fact that Bajo del Tigre still had a significantly larger mam mal abundance than the La Calandria sites indicates that this increase was at least partly due to natural regeneration . When individual appearances were broken down into species, N. narica, P. concolor, and D. punctata were shown to be significantly more abundant i n areas of natural regeneration, while 0 2 4 6 8 10 12 14 Heteromys nubicolens Scotinomys teguina Oligoryzomys fulvescens Tylomys watsoni Sigmodon hispidus Number of Rodents Bajo del Tigre Curi Cancha La Calandria 1 La Calandria 2
The effects of reforestation on mammal diversity and abundance Pruett 9 there were no significant differences in the distributions of D. novemcinctus and D. marsupialis between site. The reason for the lack of significant differences was likely because very few D. novemcinctus a nd D. marsupialis were seen at any of the sites, so it was hard to come to conclusions about their relative abundances in such a short time period. The increase in N. narica contradicted the study done by SÃ¡nchez in 2008 , which found coatis to be the only species that had a higher abundance in areas of reforestation. Coatis ( N. narica ) typically survive well in many locations, in part since they can coexist with humans. All the sites were in locations in which coatis would not benefit from humans, as there was no one consistently giving t hem food scraps in these sites, which may account for the differences with the study done by SÃ¡nchez in 2008 . They may have been less present in La Calandria since mostly Inga spp. and Ocotea spp. were planted there, whereas there was more plant diversity in the naturally regenerating sites. Coatis typicall y feed on arthropods and fruits and frequently feed on figs from Ficus spp . (Wainwright 2007). From personal observations, n either of the sites in La Calandria had large nu mbers of Ficus spp. , likely since none were planted here (Hamilton 2017), while the site in Curi Cancha had more of them. Additionally, areas with more diversity in vegetation support significa ntly more diversity in insects (He et al., 2009), which may be the reason for the higher abundance of coatis , as this would give them more types of insects to prey upon. The increase in the diversity of D. punctata is also likely due to dietary factors. Central American a goutis ( D. puncata ) feed on many types of fruits but prefer the fruits of certain types of Arecaceae (Wainwright 2007). However, in La Calandria, there was a significantly higher abundance of Fabaceae and Lauraceae, as these were used in the reforestation that occurred here (Hamilton 2017) . From p ersonal observations, t he re were more Arecaceae and other plant families in the areas with natural regeneration, which may account for why agoutis were less abundant in reforested areas. The increase in P. concolor abundance is likely due to the increased abundance in prey. Curi Cancha, the only site with P. concolor, had significantly more individual appearances of medium sized mammals, which are common prey items for P. concolor . Some of these prey items include agoutis, opossums, and armadillos (Wainwri ght 2007), all of which were observed to be more abundant in Curi Cancha. Throughout the project, there were several issues with the cameras. Several of them did not function or had a corrupt SD card, so they may not have taken pictures for several days of the experiment. However, this was distributed fairly equally across the sites, so there was still a similar number of hours of pictures taken between naturally regenerating sites and sites with reforestation. Another issue with the camera traps was that s ome shifted in position, so even though they were recording pictures, they were aimed above the height that many medium sized species would have passed by. Small Mammals There was no difference in number of rodent species between naturally regenerated areas and reforested areas. Other than H. nubicolens, only one rodent of each species was found, so it is not possible to make assumptions about their dist ributions and abund ance with this little data . There were more individuals caught in reforested sites than there were in sites with natural regeneration, though this was not significant. There was a high abundance of potential rodent food sources in both the reforested and n aturally regenerated sites, which is likely the reason for the lack of diversity and abundance between reforested and naturally regenerated sites.
The effects of reforestation on mammal diversity and abundance Pruett 10 There were no significant differences in the distributions of H. nubicolens between naturally regenerated ar eas and reforested areas. However, there were significant differences in the distribution of H. nubicolens between each site. These differences may be due to the different types of vegetation that grow at each site. La Calandria 2, the site with the most H . nubicolens , had Inga punctata and Ocotea s p p. p lanted. H. nubicolens tend to consume lots of fruit (Wainwright 2007 ), so they may have been attracted to this site due to the fact that the two most commonly planted families (Fabaceae and Lauraceae) produce high quantities of potential food . Additionally, no other species of rodents were found at this site, so they may have had such a high abundance due to the lack of interspecific competition. La Calandria 1, which had fewer H. nubicolens, grew prima rily Ocotea floribunda and Ocotea monteverdensis. The smaller numbers of H. nubicolens at this site may have been due to the differences in vegetation when compared with La Calandria 2 or based on the fact that other species of rodents were found here, so they may encounter competition . This study showed a significant increase in the abundance of mammals in sites that are naturally regenerated as opposed t o reforested . A study done in the Andes examining diversity of medium sized mammals in reforested areas showed a similar trend ( SÃ¡nchez et al., 2008). However, few studies have been done examining the effects of reforestation on mammal diversity and abundance, so more studies are needed to exami ne this across different types of climates and life zones. Studies are needed to examine the potential reasons for the decreases in abundance of medium and large mammals in reforested areas, so that this can be attributed to something specific, like diet. They could also examine how specific plant species affect diversity and abundance of mammals. This would allow those involved in reforestation efforts to learn more about how to reforest in a way favorable for mammals that inhabit the area. Additionally, studies need to be done examining the other factors that may have affected these results. There were many humans and domestic animals ( C. familiaris and F. catus) seen in the cameras along my sites, so studies examining the role these play in mammal divers ity would be very useful. ACKNOWLEDGEMENTS I would like to thank Frank Joyce, Debbie Hamilton , and Mauricio Ramirez for allowing me to use their land at Bajo del Tigre, La Calandria, and Curi Cancha, respectively, for this project. I would also like to thank them for the information they provided me about the natural history of each of the sites in which I was working . I would like to t hank Federico Chinchilla for his support and assistance throughout this project, his knowledge of rodents and traps, and his help with translation. I would like to thank SofÃa Arce Flores for her advice and assistance throughout this project. I would like to thank Emily Parker and Kathy Dao for working with me to set Sherman t raps and take rodent data. I would additionally like to thank Emily Parker for the advice she provided me throughout this paper. LITERATURE CITED Galetti , M . , Giacomini, H C., Bueno, R S., Bernardo, CSS., Marques, R M., Bovendorp, R S., Steffler, CE., Rubim , P., Gobbo, S K., Donatti, C I., Begotti, R A., Meirelles, F., Nobre, RdA., Chiarello, A G., Peres, CA. June 2009. Priority Areas for the Conservat ion of
The effects of reforestation on mammal diversity and abundance Pruett 11 Atlantic Forest Large Mammals. Biological Conservation, vol. 142, no. 6 , pp. 1229 1241. Biological Sciences Database. Hamilton, Debbie. Nov. 2017. Personal Communication. Monteverde, Costa Rica. He, J X . , Zhou, H., Guo, L F., Wang, X G., Zhao, Z G., Zhao, J. Mar. 2009. Effects of Dynamic Variety of Plant Community on In sect Diversity in Litchi Orchard. Guihaia , vol. 29, no. 3 , pp. 337 341. Biological Sciences Database. Joyce, Frank. Dec. 2017. Personal Communication. Monteverde, Costa Rica. Lamb, David., Erskine, Peter D., Parrotta, John A. 9 Dec. 2005. Restoration of Degraded Tropical Forest Landscapes. Science , vol. 210, no. 5754 , pp. 1628 1632. Biological Sciences Database. Martin, Patrick H ., Peter J Bellingham. Sept. 2016. Towards Integrated Ecological Research in T ropical Montane Cloud Forests. Journal of Tr opical Ecology, vol. 32, no. 5 , pp. 345 354. Biological Sciences Database. Mugwedi, Lutendo F ., Rouget, Mathieu., Egoh, Benis., Ramd hani, Syd., Slowtow, Rob., RenterÃa, Jorge L. 2017. An Assessment of a Community Base d, Forest Restoration Programme in Du rban (EThekwini), South Africa. Forests, vol. 8, no. 8, Biological Sciences Database. Oosterhoorn, M., Kappelle, M. 25 Feb. 2000. Vegetation Structure and Composition along an Interior Edge Exterior Gradient in a Co sta Rican Montane Cloud Forest. Forest Ecology and Management , vol. 126, no. 2 , pp. 291 307. Biological Sciences Database. RamÃrez , Mauricio. Dec. 2017. Personal Communication. Monteverde, Costa Rica. SÃ¡nchez, Francisco . , SÃ¡nchez Palomino, Pedro., Cadena, Alberto. 2008. Species Richness and Indices of Abundance of Medium Sized Mammals in the Andean Forest and Regenerations w ith Andean Alder: A Preliminary Analysis. Caldasia, vol. 30, no. 1, pp. 197 208. Biological Sciences Database. SÃ¡nchez Azofeif, G A., Harriss, R C., Skole, D L. 2001. D eforestation in Costa Rica: A Quantitative Analysis Using Remot e Sensing Imagery. Biotropica , pp. 378 384. Biological Sciences Database Wainwright, Mark. 2007. The Mammals of Costa Rica: a Natural History and Field Guide. Cornell University Press, 2007.