xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xlink http:www.w3.org1999xlink xmlns:xsi http:www.w3.org2001XMLSchema-instance
leader 00000nas 2200000Ka 4500
controlfield tag 008 000000c19749999pautr p s 0 0eng d
datafield ind1 8 ind2 024
subfield code a M39-00024
Burdick, Caroline C.
Diversidad de mamferos pequeos y densidades de poblacin en Monteverde, Costa Rica: Un estudio de seguimiento de Rothman (1999)
Small mammal diversity and population densities in Monteverde, Costa Rica : A follow- up study of Rothman (1999)
Widespread deforestation and El Nio events may have an effect on Costa Ricans climate, and overall, Monteverdes climate is becoming drier and less predictable due to short-term global warming impacts. Due to widespread habitat fragmentation and hunting, the population of small and medium-sized mammals has decreased dramatically. This study investigates rodent species richness, abundance, and diversity in three life zones in Monteverde, Costa Rica, to determine whether they have changed since a similar study conducted in 1999. Small mammals were sampled in Tropical Lower Montane Wet Forest, Tropical Premontane Moist Forest, and Tropical Lower Montane Rain Forest in the Monteverde region, Costa Rica, during the end of the dry season in mid-April, early-May. Data were collected using thirty 26 x 10 x 12 cm Sherman Live Traps. In 15 nights of trapping (450 total traps), 47 individuals were captured (with a few recaptures). Overall, six species and two orders were represented: Murid and Heteromyid rodents included 33 P. nudipes, six S. teguina, four H. desmarestianas, one O. albigularis, and one R. rattus. Order Insectivora was represented by two soricid shrews, C. gracilis. The mid-elevation site (see comments about mid-elevation below) had the lowest rodent abundance, and thus differing greatly from the results of Rothmans study. Furthermore, P. nudipes was absent from this site. Changing climate conditions could be partially responsible for the missing P. nudipes and relatively low abundance in the Tropical Lower Montane Wet Forest life zone, but also chance could play a strong role, as could sampling methods.
La deforestacin y los eventos del Nio pueden haber afectado el clima de Costa Rica, y en general provocando un clima ms seco y menos predecible debido a los impactos del calentamiento global a corto plazo. Debido a la fragmentacin del hbitat y la cacera, las poblaciones de mamferos pequeos y medianos han disminuido dramticamente. Este estudio investig la diversidad, la riqueza de especies y la abundancia de roedores en tres zonas de vida en Monteverde, Costa Rica para determinar si han cambiado desde que se hizo un estudio similar en 1999.
Text in English.
Rodents--Variation--Costa Rica--Puntarenas--Monteverde Zone
Fragmented landscapes--Costa Rica--Monteverde Zone
Roedores--Variacin--Costa Rica--Zona de Monteverde
Paisajes fragmentados--Costa Rica--Zona de Monteverde
Diversidad de especies
Tropical Ecology 2008
Habititat fragmentation--Costa Rica--Monteverde Zone
Ecologa Tropical 2008
Fragmentacin del hbitat--Costa Rica--Zona de Monteverde
t Monteverde Institute : Tropical Ecology
1Small Mammal diversity and Population Densities in Monteverde, Costa Rica: A FollowUp Study of Rothman (1999) Caroline C. Burdick Department of Ecology and Evolutionary Biology, Uni versity of Colorado-Boulder ABSTRACT Widespread deforestation and El Nio events may hav e an effect on Costa RicanÂ’s climate, and overall, MonteverdeÂ’s climate is becoming drier and less pre dictable due to short-term global warming impacts. Due to widespread habitat fragmentation and hunting, the p opulation of small and medium-sized mammals has dec reased dramatically. This study investigates rodent speci es richness, abundance, and diversity in three life zones in Monteverde, Costa Rica, to determine whether they h ave changed since a similar study conducted in 1999 Small mammals were sampled in Tropical Lower Montane Wet Forest, Tropical Premontane Moist Forest, and Tropi cal Lower Montane Rain Forest in the Monteverde region, Costa Rica, during the end of the dry season in mi d-April, early-May. Data were collected using thirty 26 x 1 0 x 12 cm Sherman Live Traps. In 15 nights of trap ping (450 total traps), 47 individuals were captured (with a few recaptures). Overall, six species and two orde rs were represented: Murid and Heteromyid rodents included 33 P. nudipes six S. teguina, four H. desmarestianas one O. albigularis and one R. rattus. Order Insectivora was represented by two soricid sh rews, C. gracilis. The midelevation site (see comments about Â“mid-elevation b elow) had the lowest rodent abundance, and thus dif fering greatly from the results of RothmanÂ’s study. Furth ermore, P. nudipes was absent from this site. Changing climate conditions could be partially responsible for the m issing P. nudipes and relatively low abundance in the Tropical Lower Montane Wet Forest life zone, but also chance could play a strong role, as could sampling method s. RESUMEN La deforestacion y los eventos EL Nino pueden haber afectado el clima de Costa Rica, y en Monteverde p orvocando mayor sequia y un clima en general menos predecible Debido a la fragmentacion del habitat y la caceri a, las poblaciones de pequenos y medianos mamiferos han de crecido dramaticamente. Este estudio investigo la d iversidad, riqueza de especies y abundancia de roedores en tre s zonas de vida en Monteverde al final de la epoca seca. En 15 noches de captura (450 trampas en total) se captura ron 47 individuos; seis especies en dos ordenes. El sitio de elevacion intermedia presento la menor abundancia d e roedores, lo que provoco que este resultado fues e muy diferente al de Rothman. P nudipes estuvo ausente de este sitio. Las condiciones clim aticas cambiantes podrian ser parcialmente responsables por las bajas abundancias de roedores. INTRODUCTION Although Costa Rica boasts an incredible mammal div ersity (207 species), few experiments have been conducted on the ecology, distribution, abunda nce, altitudinal zonation, systematic relationships, biogeography and effect of climatic changes on most mammals. Widespread deforestation and El Nio events may have an effect on Costa RicanÂ’s climate. Overall, MonteverdeÂ’s climate is becoming drier and less pre dictable due to short-term global warming impacts (Timm & LaVal 2000). Enquist (2002) sugges ts that in general, high elevation life zones are particularly sensitive to temperature cha nges, while lower elevation life zones tend to be more sensitive to changes in precipitations. Wi th rising global temperatures, geographic ranges increase in elevation, thereby decreasing in size.
2 Due to widespread habitat fragmentation and huntin g, the population of small and medium-sized mammals has decreased dramatically. M any generalist species have adapted to the increase in human disturbance; some have even i ncreased in abundance. However, many specialists have been unable to adapt to habitat fr agmentation and have become extremely rare. This could be because of life zone differences in s pecies abundance and diversity, as they change on an elevational gradient (Timm and LaVal 2000). Furthermore, many changes in the Monteverde area may have affected, either directly or indirectly, the rodent populations in the past ten years, due to climate change, human distur bance and contamination levels, or the introduction of new predators into the area. Very few studies have been done to determine the effect these disturbances have had on the distribut ion and abundances of small rodents in the tropics. Rothman (1999) completed a study in the M onteverde area to examine effects of climate change and elevation on small rodent populations. He compared his capture rate to a previous study conducted by Stephen Anderson (1982) to deter mine if the rodent population had changed. No conclusive evidence was found at that time suppo rting the idea that rodent population varied in abundance or distribution. A survey conducted b y Velez-Zuazo et al (2005) determined that Monteverde has four prominent small rodent species that inhabit the area: Peromyscus nudipes (Cloud Forest Deer Mouse), Heteromys desmarestianas (Forest Spiny Pocket Mouse), Scotonomys teguina (AlstonÂ’s Singing Mouse), and Oryzomys albigularis (Montane Rice Rat). Because of their widespread distribution and opport unistic and omnivorous diet, Peromyscus nudipes (Cloud Forest Deer Mouse) is known as a generalist species, thus tolerating both disturbed and forested areas. Therefore it has bec ome the most abundant small mammal in the area (Rothman 1999; Nowak and Paradiso 1983). Alth ough the ultimate goal is to understand fluxuations in small mammal populations, and thwart possible losses in abundance, basic census data is needed so that monitoring can begin. This study seeks to determine if there has been a change in abundance of P. nudipes since RothmanÂ’s experiment in 1999, and if there i s a difference in species richness, abundance, and dive rsity in small mammals between three life zones in Monteverde, Costa Rica. METHODS AND MATERIALS Study Areas The following three life zones were sampled in this study: Tropical Lower Montane Wet Forest, Tropical Premontane Moist Forest, and Tropical Lowe r Montane Rain Forest. Tropical Premontane Wet Forest is a two-layered, se mideciduous, seasonal forest of medium elevation between 1300 and 1470 m. Canopy t rees are mostly dry-season deciduous, understory trees are evergreen, the shrub layer is dense, and the ground layer is sparse. Woody vines are abundant, but epiphytes are rare (Hartsho rn 1983). The second five nights of trapping occurred in this life zone, on land owned by Frank Joyce. Traps were placed along two separate trails; the first trail was narrow and along a ravi ne, and the other trail was along flat ground with various micro habitats. This area will be referred to as site one from now on. Tropical Lower Montane Wet Forest life zone (site t wo) is primarily restricted to central Costa Rica, is an evergreen forest found at elevati ons of 1470 to 1600, and has two tree strata. This forest consists of canopy trees, a fairly open understory, and a relatively dense shrub layer. The ground layer is well-covered with ferns, vines, and moist, rotting leaves. Epiphytes, such as orchids, bromeliads, and meliads, are common (Harts horn 1983). The first five nights of
3 trapping on the property of John and Martha Campbel l, in what is commonly called the Bullpen. This area was relatively open due to previous defor estation. Traps were placed along the trail in the understory of the forest, along the edge, and o ut in the open in a tree fall area. Also, traps were set in an open area where remnant, emergent tr ees remained. The final life zone sampled, Tropical Lower Montan e Rain Forest, is an evergreen forest above 1600 m. This site will be referred as site t hree. Ericaceae and Melastomataceae are abundant shrubby epiphytes, and large-leaved vines are occasional. Large lianas are uncommon in this area. Canopy trees are mostly 25-30 m tall and understory stratum is often dense with small trees. The shrub layer is very dense, and th e ground layer is well covered with ferns, sedges, and patches of moss (Hartshorn 1983). This was the third trapping site for this study, and contains the continental divide above La Estaci on between 1600 and 1842 m in elevation. Traps were laid along the Sendero Principal trail i n various micro-habitats, consisting of small trees, undergrowth, and roots. Sampling Small mammals were sampled during the end of the 20 08 dry season from the April 15th to May 10th. Data were collected using thirty 26 x 10 x 1 2 cm Sherman Live Traps. Each study site was sampled five times, for a total of 15 nights of trapping. Traps were laid in various microhabitats along trails, most often placed one to fiv e meters apart on alternating sides of the trail. Micro-habitats included the underbrush in the fores t, edge habitat, the base of trees, among roots, along logs, and in tree fall areas. Each trap was carefully placed even with the ground so no tipping would occur if a small mammal entered. A m ixture of creamy peanut butter and banana slices were used as bait, to be consistent with Rot hmanÂ’s (1999) study. The thirty Sherman traps were opened each afternoon between 3 and 4, and che cked between 7:30 a.m. and 8:30 a.m. the following morning. The trap number, micro-habitat, weight, sex, body length, tail length, ear and hind foot length were recorded when a small mam mal was discovered. These measurements were used, with the help of pictures, to identify t he species of each captured individual. A handheld Pesola 100g spring scale was used to weigh the individual, and lengths were taken using a dial caliper. Data Analysis The values of HÂ’ and t were calculated using a Shan non-Weiner index of diversity to determine if there was a difference in diversity, evenness an d species richness between the three sites sampled. To determine if there was a difference in frequency of species per site, a contingency table analysis was conducted. Finally, a Kruskal-W allis nonparametric test was performed to determine if there was a difference in abundance be tween sites. Comparisons in capture rates were then made between these data and that of Rothm an (1999), who performed a study to determine the abundance of small mammals in Monteve rde, Costa Rica.
4 RESULTS In 15 nights of trapping (450 total traps), 47 indi viduals were captured (with a few recaptures). Overall, six species and two orders were represente d: Murid and Heteromyid rodents included 33 P. nudipes six S. teguina, four H. desmarestianas one O. albigularis and one R rattus. Order Insectivora was represented by two soricid sh rews, C. gracilis. In the Bullpen (Tropical Lower Montane Wet Forest), a total of five individu als were captured including two C. gracilis, two S. teguina, and one H. desmarestianas This area yielded the least amount of captured individuals than the following two sites. The land of Frank Joyce (Tropical Premontane Moist Forest) produced 23 total captures, 19 of which wer e P. nudipes three were H. desmarestianas and one that was R. rattus. The final site at the continental divide (Tropical Lower Montane Figure 1. Species Richness of Tropical Premontane Wet Forest (Site 1), Tropical Lower Montane Wet For est (Site 2), and the Tropical Lower Montane Rain Forest (Sit e 3) Monteverde, Costa Rica 0 2 4 6 8 10 12 14 16 18 20 Site 1Site 2Site 3 Location# of Individuals Caught P. nudipes C. gracilis S. teguina H. desmarestianas O. albigularis R. rattus
5 Rain Forest) produced a total of 19 captures. Four teen of the 19 captures were P. nudipes, four were S. teguina, and one was O. albigularis (Fig. 1). When comparing P. nudipes to the other species caught, zero P. nudipes and five individuals of other species were capture d in site one. Site two produced 19 P. nudipes and four other individuals. The final site generat ed 14 P. nudipes and five individuals of other species (Fig. 2). 0 2 4 6 8 10 12 14 16 18 20 Site 1Site 2Site 3 Location# of Individuals Caught Peromyscus nudipes Other species Figure 2. Species Richness in Monteverde, Costa Ric a, comparing P. nudipes with other captured species There was a difference in diversity between site tw o (n2 = 5) and site one (n1 = 23, t = 3.18, p < 0.05), with site two having a higher diversity than Frank JoyceÂ’s land (HÂ’2 = 1.05, HÂ’1 = 0.56). Also, site two had a higher species richnes s than site one (Smarg2 = 1.24, Smarg1 = 0.64) and had a higher evenness (E2 = 0.96, E1 = 0.51). Furthermore, there was a difference in diversity between site two and site three (n3 = 19, t = 2.33, p < 0.05). Site two had a higher diversity than site one (HÂ’2 = 1.05, HÂ’1 = 0.71), a higher species richness (Smarg2 = 1.24, Smarg1 = 0.68), and a higher evenness (E2 = 0.96, E1 = 0.64). Finally, there was no difference between site one and site three, however, site three had a slightly higher diversity than site one (HÂ’1 = 0.56, HÂ’3 = 0.71). A contingency table analysis determined that there was a difference between the frequency of species (x2= 31.4 df = 10, p = 0.0005). In site one, more H. desmarestianas and P. nudipes were caught than anticipated. In site two, there were more C. gracilis and S. teguina captured than expected. Significantly fewer P. nudipes were caught than expected. Finally, in
6 site three, more P. nudipes and S. tequina were captured, while fewer H. desmarestianas were captured than projected (Table 1). These differenc es can be seen in Figure 3. Table 1. Chi-square test showing the difference in frequen cy of species per site (x2 = 31.4, p = 0.0005). Site one yielded more H. desmarestianus P. nudipes, and less S. teguina than predicted. Capture rates of C. gracilis and S. teguina were higher than expected, while the abundance of P. nudipes was less than predicted in site two. Finally, more P.nudipes and S. teguina were captured than anticipated in site three. According to a KruskalWallis nonparametric test, no difference in abunda nce was found between the three sampled sites (H = 0.16, df = 2, n = 18, p = 0.92). When comparing the success rate of RothmanÂ’s total captures with this study, defined by Anderson (1982) as the number of rodents captured d ivided by the total number of traps (# captured/ # traps), it was determined that Rothm an caught a larger percentage of P. nudipes that the current study. In his study, 70 individua ls were caught, 60 of which were P. nudipes, nine were H. desmarestianas and one of O. albigularis Refer to Table 2 to compare percentage of captures between Rothman (1999) and this study ( Burdick 2008). Table 2. Comparison of capture rates between Rothman (1999) and that of Burdick (2008). Study Total Captures P. nudipes C. gracilis S. teguina H. desmarestianas O. albigularis R. rattus Rothman 70 60 0 0 9 1 0 % 86 0 0 13 1 0 Burdick 47 33 2 6 4 1 1 % 70 4 13 9 2 2 x2 O. albigularis C. gracilis H. desmarestianus P. nudipes R. rattus S. teguina Site 1 Observed Expected 0 0.489 0 0.978 3 1.956 18 15.644 1 0.489 0 2.444 Site 2 Observed Expected 0 0.111 2 0.222 1 0.444 0 3.556 0 0.111 2 0.556 Site 3 Observed Expected 1 0.4 0 0.8 0 1.6 14 12.8 0 0.4 3 2
7 DISCUSSION Due to the absence of P. nudipes, there was a higher species richness, diversity, and evenness between t he Tropical Lower Montane Wet Forest life zone and the other two sites. This site was much more even than the other two due to the fact that it was not dominated by P. nudipes Therefore, because site two and three were both dominated by P. nudipes their species richness, diversity, and evenness wer e not different. RothmanÂ’s (1999) capture success of P. nudipes in the three life zones was much greater (86% vs. 70%) than I found. However, more species of rodents and soricids are represented in this sur vey, therefore giving rise to greater species richness. He found only three species ( P. nudipes, S. teguina, and O. albigularis ), while six species were captured in this study. It appears that the abundance of P. nudipes has decreased since 1999. However, this could also be due to chance or a small sample size, especially since the species was completely missing from one site. Clearly the results for this study are different than those of RothmanÂ’s study, where he found a large abundance of P. nudipes in all three of his sampled sites. Here, the highest and lowest elevations (Premontane Wet and Lower Montane Rain Forest) held a disproportionately higher abundance of P. nudipes However, it was absent from the mid-elevation (Lower Montane Wet Forest). Also, these data are not consistent with previous work done by McCain (2004) in which the highest capture frequencies was 1500-1550, the same elevation as the Lower Montane Wet Forest. Therefore, instead of having the lowest success rat e, 0 0.5 1 1.5 2 2.5 3 3.5 4P n u d ip e s C g r a ci li s S t e g u in a H d e s m a r estia n a s O a l bi g u la ri s R r a tt u s Number of Individuals Caught Site 2 observed Site 2 expected 0 2 4 6 8 10 12 14 16P nudi p es C. gracilis S. teg u i n a H. de sma r es t ianas O. al b i gu l a ris R. r attusNumber of Individuals Caught Site 3 observed Site 3 expected 0 2 4 6 8 10 12 14 16 18 20P. nudipes C g ra ci l i s S. teg ui na H d esm ar esti a n a s O. albigularis R. rattusNumber of Individuals Caught Site 1 observed Site 1 expected a) b) c)
8 this site should have been the most abundant and di verse. There are numerous possible reasons that may expla in why the trapping rates differed between this study and RothmanÂ’s, spanning simple m ethodological differences, to chance, and to potential environmental change. Their semi-arbo real nature could account for the lack of success in trapping this generalist species, as tra ps were solely placed on the ground. As for the role of chance, a lot of natural variation in rodent popula tion sizes exists, as they fluctuate from year to year (Janzen 1983). Also, this study provided small samples, therefore may not have enough data to prove representative of the population. Lastly, an important consideration for the Monteverde zone is the role that climate change may play in po pulation dynamics. While this study reveals no obvious factor or factors as reasonable for the lack of P. nudipes from the Tropical Lower Montane Wet Forest life zone, the results here, as well as RothmanÂ’s, could provide baseline data for the type of long-term monitoring program needed to access population changes. ACKNOWLEDGMENTS I would like to thank the following for the help an d support I received during this study: Frank Joyc e and Martha Campbell for the use of their land near Bajo del Ti gre and the Bullpen, respectively. Thanks are also extended to Marvin Hildago, and the employees of the Estacion B iologica for the use of their forest and resources. A big thank you goes to Federico Chinchilla for his knowledge a nd identifying skills of small mammals in the Monte verde area. Extra special thanks go to Tania Chavarra and Kar en Masters for their help in finding sites, support suggestions, and statistical genius. Pablo and Tae gan, I owe you my sanity at times. Also, thanks to Laura Grieneisen and Rachel Skalina for their assistance and company in setting up and breaking down my trap syou two saved me a lot of time and frustration at the divid e. A final huge thank you goes to my parents for a llowing, financing, and supporting my decision to study abro ad in Costa RicaIt was an experience I will never forget. LITERATURE CITED Anderson, S. D. 1982 Comparative population ecolo gy of Peromyscus mexicanus in a Costa Rican wet forest. Ph. D dissertation, University o f Southern California, Los Angeles, CA. Enquist, C. A. F. 2002 Predicted regional impacts of climate change on the geographical distribution and diversity of tropical forests in C osta Rica. Pp. 519-534. Journal of Biogeography. Blackwell Science Ltd. Fleming, T. H. 1977 Response of two species of tr opical heteromyid rodents to reduced food and water availability. Journal of Mammology. 58:1 0-6. Found in Costa Rican Natural History D. H. Janzen (ed.) University of Chicago, C hicago.USA. Fleming, T. H. 1983 Heteromys desmarestianus (Raton Semiespinosa, Spiny Pocket Mouse) pp. 474-475 In: Costa Rican Natural History. D. H. Janzen (ed.). The University of Chicago, Chicago. USA. Hayes M. and R. LaVal. 1989 The Mammals of Monteverde: An annotated check list to the mammals of Monteverde. Tropical Science Center: S an Jose, Costa Rica. Hartshorn, G. S. 1983 Plants. Pp. 118-25 In: Costa Rican Natural History D. H. Janzen (ed.) The University of Chicago, Chicago, USA. Figure 3. Visual representation of chi-square tests (x2= 31.4, p= 0.0005). (a) Observed vs. expected values for Site 1 (b) Observed and expected values for Site 2 (c) Values for Site 3.
9 Holdridge, L. R. 1942 Life Zone Ecology Revised E dition. Tropical Science Center, San Jose, Costa Rica. Janzen, D. H. and D. E. Wilson 1982 pp. 427 in Co sta Rican Natural History. D. H. Janzen (ed.). The University of Chicago, Chicago, USA. McCain, C. M. 2004 The mid-domain effect applied to elevational gradients: species richness of small mammals in Costa Rica. Pp. 19-31. Journa l of Biogeography. Blackwell Publishing Ltd. Nowak, R. M. AND J. L. Paradiso. 1983 WalkerÂ’s Ma mmals of the World. 4th ed. Vol. II. John Hopkins University Press, Baltimore and London Pounds, J. A., M. R. Bustamante, L. A. Coloma, J.A. Consuegra, M.P.L. Fogden, P. N. Foster, E. L. Marca, K. L. Masters, A. Merino-Viteri, R. Pusch endorf, S. R. Ron, G. A. SanchezAzofeifa, C. J. Still, and B. E. Young. 2006 Wide spread amphibian extinctions from epidemic disease driven by global warming. Pp. 161 -167. Nature. V. 439. Nature Publishing Group. ______. J. A. 2000 Amphibians and Reptiles. In Mon teverde: Ecology and Conservation of a Tropical Cloud Forest. Eds. Nadkarni, N. M. and N. T. Wheelwright, Oxford University Press, New York. Rehmeier, R. L. 1996 An examination of vertical s tratification and niche partitioning in a cloud forest rodent community. CIEE, Spring semester. Reid, F. A. 1997 A Field Guide to the Mammals of Central America & Southeast Mexico. pp. 197-198; 206-207; 221; 223-224; 232-233. Oxford Un iversity Press, New York, New York 10016. Rothman, A. W. 1999 Effects of Climate Change an d Elevation on Small Rodent Communities in Monteverde, Costa Rica. CIEE, Spring semester. Timm, R. M., and R. K. LaVal. 2000 Mammals. In M onteverde: Ecology and Conservation of a Tropical Cloud Forest. pp. 233-244. Nadkarni, N M. and N. T. Wheelwright (eds). New York: Oxford University Press. Velez-Zuazo, X., L. Lawson, B. Timm, M. Ferro, E. K uprewicz, and R. Weaver. 2005 comparison of diversity and abundance of mice in tw o types of forest during the dry and wet season. In OTS 2005-3. APPENDIX Natural History of Species Peromyscus nudipes (Cloud Forest Deer Mouse) has a large distribution and is especially abundant in semideciduous, secondary forest, and al ong streams (Reid 1997). It is largely terrestrial but can climb well, and in a study cond ucted by Rehmeier (1996), it was found that they are surprisingly more arboreal than originally thought. Burrows can be found in open areas of the forest floor, as well as in the roots of tre es or under logs (Reid 1997). It is a generalist species, and extremely abundant in Monteverde (Nowa k and Paradiso 1983; Hayes and LaVal 1989) (Fig. 6). Heteromys desmarestianus (Forest Spiny Pocket Mouse), is distributed from M exico through Central America to Northwestern Colombia, a nd ranges between the lowlands and 2400 meters in elevation (Reid 1997). According to Janz en (1983), its usual habitat is montane or tropical rain forest. It is primarily a granivore and uses its external fur-lined cheek pouches to
10 travel with seeds (Janzen 1983). Much like P. nudipes it burrows in open areas on the forest floor and under roots of trees. This species was e specially susceptible to food and water deprivation, which is consistent with previous stud ies (Fleming 1977). Within two or more days of water deprivation, individuals can lose too much weight and die. After a limited diet of only sunflower seeds, as tested by Fleming (1997), indiv iduals can also lose weight quickly and die when about 20% of their weight has been lost. Even after only one night in the trap, I observed that most H. desmarestianas were in very poor health when discovered in the mor ning. Although they have a large distribution, specialize d eating habits may make them vulnerable to changes in seed and water availability (Fig. 9). Scotinomys teguina (AlstonÂ’s Singing Mouse) ranges from 900 to 2900 me ters in elevation, and is common and abundant in cloud fore st, forest edge and grassy clearings. It is generally diurnal and typically most active in the morning. S. teguina is mostly terrestrial and uses runways and well known paths through grass and under logs. It has been found from one to three meters above the ground after feeding on the nectar of Blakea austin-smithii a Melastome. It is mostly insectivorous-80% of its diet consis ts of insects-but will also feed on seeds, nectar and fruits. An interesting characteristic of this species is that both sexes will frequently make an insect-like, trilling call that can last up to 10 s econds (Reid 1997). Due to its insectivorous nature, S. teguina could be negatively affected by a decrease in inse ct diversity due to a change in environment and temperature (Fig. 5). Oryzomys albigularis (Montane Rice Rat) is found throughout the mountai ns of Costa Rica, from 1000-3000 meters in elevation. It is co mmon in mature, evergreen, and highland forests and is often found along streams, on wetter slopes and ridges, and near mountaintops. O. albigularis swims well, but is mainly terrestrial and can be t rapped on the ground near fallen logs and on low, mosscovered branches (Reid 1997). Al though it has a widespread metapopulation, it has a specific habitat among the mountains of Co sta Rica, which may make it vulnerable to changes in habitat and climate. Its metapopulation size is much more limited than P. nudipes so it may be a more specialist species than the deer m ice (Fig. 4). Cryptotis gracilis (Talamancan Small-Eared Shrew) is a very small sori cid mammal in the order Insectivora. It ranges between 1800 and 3400 m in the Central and Talamancan mountains of Costa Rica. Although is fairly common in highland oak forest and paramo, this order is poorly represented in Costa Rica (Reid 199 7) (Fig. 8). Rattus rattus (Roof Rat) has been introduced worldwide, but in t his area it is extremely uncommon (Hayes and Laval 1989). It is found in an d around human habitations-both in cities and rural areas-and less commonly in forest or at forest edge. It is mainly nocturnal and climbs well. Its diet includes grain, fruit, and garbage, and is an extremely prolific breeder. Due to its adaptability, it may kill or displace native specie s (Reid 1997) (Fig. 7).
11 Figure 4. Oryzomys albigularis (F. Muridae) Figure 5. Scotonomys teguina (F. Muridae) Figure 6. Peromyscus nudipes (F. Muridae) Figure 7. Rattus Rattus?? Unknown juvenile. (F. Muridae) Figure 8. Cryptotis gracilis (F. Soricidae) Figure 9. Heteromys desmarestianas (F. Hetermyidae)