|USFLDC Home | Tropical Ecology Collection [Monteverde Institute]||| RSS|
This item is only available as the following downloads:
The mosaic effect: habitat fragmentation and microhabitat partitioning among three wren species in San Luis, Costa Rica Mackenzie Most Department of Zoology, University of California at Santa Barbara ABSTRACT Increasing human presence in previously forested areas leads to the alteration and fragmentation of habitats used by resident species of birds. In this study I examine the effects of a mosaic like arrangement of habitat types on the abundances and distributions of three species of wrens, the Ho use Wren ( Troglodytes aedon ), Plain Wren ( Thryothorus modestus ), and Rufous and white Wren ( Thryothorbus rufalbus ) to determine any resulting differences in abundances and distributions between species. I conducted visual and auditory observations of wren s during timed walks through four habitat types in the San Luis valley of Costa Rica including primary forest, secondary forest, disturbed home and garden areas, and cleared pastures. Wrens were observed to partially overlap in range likely due to factors of proximity and niche partitioning on the basis of overall habitat, microhabitat, and edge characteristics, but not tree height. Overall wren abundance was equal between species but differed between habitats, and abundance was higher in disturbed areas wi th high heterogeneity than in homogeneous disturbed areas. Effects of human disturbance are thus demonstrated to vary based on structural characteristics in terms of species response. INTRODUCTION Worldwide problems of habitat loss and fragmentation are currently posing increasing pressures on biodiversity and species distributions. Human presence in and use of land causes alteration of natural community structures, as exemplified by processes of deforestation and land transformation for homes and agricul tural purposes. Because species rely on particular habitat characteristics to which they are adapted, changes in natural environments can be disastrous for survival (Barbosa et al ., 2010). Examinations of disturbances have often confirmed such reductions in resident avian diversity and abundance (Lee et al ., 2010), indicating threats to future biodiversity associated with continued development. However, species can respond differently to similar pressures (Wilson et al ., 2009). The valley of San Luis, Pun tarenaes, Costa Rica presents a unique and interesting place in which to study habitat fragmentation, being characterized by a mosaic like mix of areas of varying disturbance levels. Intact primary forest exists alongside areas of secondary growth near ri vers and human disturbance, interspersed with farms, gardens, and pastures and divided by dirt roads. This arrangement creates both areas of isolation and corridors for movement among mmon at habitat interfaces and create differences in conditions such as light penetration and soil moisture (Kupfer & Runkle 2003, Chazdon & Fetcher 1984). I chose to examine the distribution of wrens within such an
environment due to their highly vocal a nd visible nature as well as the high dispersal abilities of birds. For these reasons as well as their differences in preferred habitat, I expect wrens to provide a strong indicator of the effects of fragmentation on species interactions including microhab itat selection and niche partitioning in cases of non analog community formation. Previous studies in the Puntarenaes region have examined wren distributions based on elevation gradients and microhabitat differences, as well as overall bird diversity as a function of disturbance. Niche partitioning among wrens was determined to occur between both foraging height in trees (Cronholm 1999) and varying forest disturbance levels (Burke 2004). In this study I extend the reach of these findings to include highe r levels of human disturbance and occupation including home and garden areas as well as cleared pastures, using a local subset of wren species consisting of the House Wren ( Troglodytes aedon ), Plain Wren ( Thryothorus modestus ), and Rufous and white Wren ( T hryothorus rufalbus ). Due to the close proximity of the habitats in focus, I hypothesize that wrens will occur in areas outside their expected ranges and therefore experience overlap. I aim to determine whether such shifts and coexistence occur, and if so whether wrens respond via niche partitioning or competitive exclusion. I will then explore potential implications of wren habitat preferences in the context of a changing landscape to make predictions regarding future populations. METHODS Study Sites I determined four habitat types for comparison within the San Luis Valley, all within the Holdridge Life Zone of premontane moist forest between elevations of 1000 and 1200 meters. Habitat A = Homes and gardens, characterized by high heterogeneity of mic rohabitat including domestic structures, cleared lawns and small fields, mixed trees and crops, integrated forest edge, and gardens of both native and exotic flora. Habitat B = Primary forest, characterized by old growth and closed canopy with low light pe netration, lacking human disturbance. Habitat C = Secondary forest, having undergone disturbance and subsequent recovery or occurring along primary forest edges; characterized by lack of a closed canopy cover, higher levels of undergrowth, and younger flor a including more pioneer and light loving species. Habitat D = Pastures, characterized by open grassy expanse and minimal tree occurrence. The only instances of continuous tree cover in pastures were located in dividing lines or segments one or two trees t hick. Study Organisms The House Wren, Plain Wren, and Rufous and white Wren are all insectivorous members of the Troglodytidae family known to occur in the Puntarenes region of Costa Rica. The known preferred habitat types of the three species differ, however. The House Wren probably once preferred scrubby second growth areas and riversides, but has adapted its range to human habitats such as yards, doorways, and roofs, where they commonly nest. The Plain Wren is known to occupy areas of dense second gr owth or overgrown field areas but to avoid closed tree cover. The Rufous and white Wren prefers moist forest habitat in dry areas but adjusts to more open habitats in wetter areas (Stiles and Skutch 1989).
Sampling Design I located four trails of approx one trail each at two sites per day for a total of eight days, alternating the hour of visitation so as to correct for potential variances due to time of observation. Sites were either vis ited early (between 06:30 and 08:30 GMT) or late morning (between 08:30 and 10:30 GMT). I conducted both visual and auditory observations of wrens, documenting species name, microhabitat type, height, and presence of habitat edge characteristics. Microh abitat types were divided into five categories as follows: Microhabitat A = Clustered trees bordering open space, where open space is defined as cleared road, pasture, or lawn without tree cover, and is adjacent to two or more sides of the tree area. Micr ohabitat B = Forest border, composed of closed tree canopy, at the edges of open or mixed microhabitats. Microhabitat C = Single tree, standing within open space such as pasture or lawn and disconnected from other trees or tree cover. Microhabitat D = Mixe d native trees among crops, including coffee, bananas, and varieties of citrus. Microhabitat E = Ground, shrubs, and fences up to 1 meter in height. Height of observation was estimated visually, and mean height values were calculated for individuals that varied position during observation. To determine whether wrens spatially partition between microhabitats based on edge characteristics, I focused on primary and secondary forests as well as pastures, where edges were most defined. Home and garden habitat was not considered due to its high intrinsic heterogeneity and lack of distinct edges. Edges within focal habitats were defined as areas within 10 meters of a different bordering habitat type. RESULTS I recorded a total of 168 observations of wrens amo ng the four distinguished habitat types. Due to the territoriality of wrens I assume that observations are generally of distinct individuals, although due to the proximity of some sites visited, slight overlap may have occurred. Although I observed no dif ference in overall abundance between the three different species of wren (X 2 = 0.036, d.f. = 2, p > 0.05), there were significant differences in wren abundances (X 2 = 29.2, d.f. = 3, p < 0.05) and species distributions (X 2 = 42.07, d.f. = 6, p < 0.05) betw een habitats (Fig. 1). House Wrens occured most often in home and garden areas as well as in pastures, while occurring less than expected in forested areas. Only one House Wren was observed in a primary forest habitat, and the individual was considered to be in forest edge space. Plain Wren individuals were observed across all habitats, but showed higher abundance in secondary forest. Rufous and white Wren distribution was concentrated in both primary and secondary forests, showing apparent dominance in primary forest with respect to other wren abundance. Total wren abundance was highest in mixed home and garden habitat and lowest in pastures (Fig. 1).
Figure 1: Abundances of three wren species among four habitat types in close proximity within a premontane moist forest zone. Significant differences in species distributions were observed (X 2 = 29.2, d.f. = 3, p < 0.05). The habitat displaying the greatest number of wrens in overlap was mixed homes and gardens. Within this area mean tree heigh t at which observed individuals occurred does not reveal any significant difference between species (ANOVA: F (2,65) = 2.94, p > 0.05; Fig. 2). However, distribution between microhabitats defined by tree type and cover exposes significant trends (G = 25.74 d.f. = 2, p < 0.05). The distribution of House Wrens includes ground and areas below 1 meter, whereas no other wrens were observed in this space. Similarly, Plain Wrens were found in mixed crop trees, where no other wrens occurred. Rufous and white Wren s occurred exclusively within tree cover, both bordering open space and within closed forest borders. However, both House Wrens and Plain Wrens were also seen occurring in these spaces. House Wrens constituted most observations of birds seen in single tre es, although Plain Wrens were also occasionally noted there (Fig. 3). There were also significant differences in occupation of edge space between wren species (G = 21.21, d.f. = 2, p < 0.05; Fig. 4).
Figure 2: Distribution of tree hei ght based observations of three wren species in a heterogeneous disturbed habitat of homes and gardens. No significant difference was found (ANOVA: F (2,65) = 2.94, p > 0.05). House Wren height mean =2.18 1.79, Plain Wren height mean =3.19 1.87, Rufous and white Wren height mean = 3.16 1.15. Figure 3: Percentage distributions of three wren species into microhabitat types. Observations were taken within a heterogeneous disturbed habitat of homes and gardens. Significant niche part itioning was observed (G = 25.74, d.f. = 2, p < 0.05).
Figure 4: Percentage distributions of three wren species into edge and interior habitat. Observations taken within primary and secondary forest as well as pastures, and edges defin ed as space within ten meters of different adjacent habitat. Significant differences in edge usage between species were observed (G = 21.21, d.f. = 2, p < 0.05). DISCUSSION According to the distribution observed in this study, wrens do overlap through out differing habitats of close proximity, but significant trends in abundance and diversity occur both between and within habitats. The difference in wren distribution between undisturbed forest and disturbed habitats means that either overall wren commun ity composition within the San Luis valley must accordingly change with increasing disturbance area, or otherwise wrens must experience crowding within remaining intact and otherwise suitable habitat. The observation of wrens outside their expected ranges indicates either that individuals are foraging in suboptimal habitats (Winker et al ., 1995) or that the new ranges do not constitute a decrease in suitability. Interestingly, the lack of difference in overall abundance between wren species indicates two po ssibilities. The first possibility is that no difference in fitness exists between species, and competitive exclusion is not occurring. The wrens must therefore successfully undergo some form of niche partitioning (Hutchinson 1991). My study at least part ially confirmed this possibility. Although each wren species displayed significant overall preferences, each was also observed throughout all habitat types. The observation of significant microhabitat partitioning between wrens within a highly heterogeneou s habitat indicates that wrens have adapted methods for coexistence. This apparent ability to maintain fitness despite habitat alteration may be
explained by wrens selecting habitat based on highly localized characteristics such as branch configuration ra ther than larger scale community or ecosystem types (Hutto 1985). Although tree height in isolation was not in this case an important determinant of wren distribution, amount and type of tree cover did appear to have significant effect. For example, the s election of local areas of tree cover within the greater context of a disturbed area by Rufous and white Wrens might mimic conditions within intact forest tree cover and therefore provide just as optimal of an environment. An apparent problem with this int erpretation of the observed wren distribution remains in the level of habitat and microhabitat distribution overlap that still occurs. Rufous and white wrens do not occupy any space without House and Plain Wrens. In light of my observation that all wrens appear to be competing and surviving equally, I explain this apparent contradiction with the Theory of Competitive Exclusion (Gause & Witt 1935) by the partitioning of habitats into edge and interior areas. Because edge areas provide different characterist ics than forest interiors such as higher light and wind inputs, (Barbosa et al., 2010), certain species have adapted characteristics to take advantage of these additional niches (Hutto 1985). While Plain Wrens were observed within primary forest, which ten ded to be generally dominated and preferred by Rufous and white Wrens, the occurrence of Plain Wrens was largely restricted to edge areas, whereas Rufous and white Wrens made greater use of interior space. The proximity of preferred habitats to suboptimal ones may also allow optimal areas to act as source pools for the local movement of species between habitat types (Winker et al ., 1995). Interestingly, the observed preference of the Rufous and white Wren for closed primary forest corresponds to the species open areas in wetter habitats (Stiles and Skutch 1989, Burke 2004). Its increased presence in wet primary forest may therefore represent a shift into suboptimal habi tat caused by competition with the Plain Wren for secondary like growth (Winker et al., 1995). The second possibility for explaining the lack of difference observed between wren abundances is that the apparent equilibrium may in fact reflect not coexisten ce but a rather an isolated time state within an ongoing process of competition and eventual exclusion (Sirami et al ., 2009). Further examination of this system over longer time scales is needed to determine whether local wren populations are actually in equilibrium or undergoing change. If habitat transformation continues favoring certain habitat types, the abundance proportions of wrens may shift toward the proportion observed in that habitat; because House Wrens were observed to succeed within disturbe d habitat, their population in San Luis is likely to grow with increasing disturbance. Furthermore, mosaic like division of environments increases the proportion of edge space within habitats, which can alter species compositions by changing characteristi cs and suitability (Kupfer and Runkle 2003). In this study, Plain Wrens appear to benefit by the creation of more edge space. An examination of overall abundance reveals that while areas of disturbance contain different amounts of overall wren presence th an forested areas, this effect is dependent on the type of disturbance. In terms of the species considered here, abundance of individuals actually increases in all home and garden areas in comparison with primary forest, for example, while remaining relat ively unchanged in pastures (although examination of pasture interiors or pastures without edge trees would likely return lower wren abundance, according to trends observed in this study). Observations of increased bird abundance in disturbed areas may app ear to contradict findings of numerous other studies (Lee et al., 2010). However, my study does show a shift in species success and dominance occurring in both types of disturbed habitat, with the House
Wren benefiting. This result agrees with previous st composition toward different or more generalized species in the presence of disturbance (Sirami et al ., 2009). Furthermore, overall wren responses to disturbance appear to depend on the heterogeneity of the resulting habitat. Each habitat observed in this study is comprised of a combination of the microhabitats also described. When viewed as such, the abundance of wrens within a habitat becomes a function of the amount of microhabitats or niches available. Therefore structurally heterogeneous disturbed habitats such as homes and gardens may be much less detrimental to sustaining populations than homogeneous disturbances such as pastures. Despite this observation, the restricted variety of species considered in this study necessitates the consideration that all organisms do not respond similarly to such factors. Nonmigrant species, for example, may select habitat based on differently scaled characteristics than migratory species or species with wider day to day ranges (Hutto 1985) and thus microhabitat partitioning may not be possible for other species, resulting in their exclusion from disturbed habitats. Such complex potential trends are important to consider and study further in the face of further development due to their implications for future biodiversity. ACKNOWLEDGEMENTS I would like to thank Karen and Alan Masters for their assistance in developing my project concept, Pablo Allen for aiding and advising me during the writing process, and Federico for his help identifying different wren species. I would also like to thank the Leiton family and the Finca La Bella project for allowing me to birdwatch on their properties. Finally, I owe this entire opportunity and experience to the support and generosity of my parents, Morgan and Don Most. LITERATURE CITED Burke, A. (2004), Niche partitioning among the Plain Wren ( Thryothorus modestus ), Rufous and white Wren ( T. rufalbus ), and Gray Breasted Wood Wren (Henicorhina leucophyrs ) along an altitudinal gradient in Monteverde, Costa Rica. CIEE Tropical Ecology and Conservation. Monteverde, Costa Rica, 169 176. Chazdon, R.L. and N. Fetcher. (1984), Photosynthetic light environments in a lowland tropical rain forest in Costa Rica. Journal of Ecology, 72: 553 564. Cr onholm, M. (1999), Vertical niche partitioning between Gray breasted Wood Wrens ( Henicorhina leucophrys ) and Ochraceous Wrens ( Troglodytes ochraceus). CIEE Tropical Ecology and Conservation, Monteverde, Costa Rica, 143 147. Gause, G.F. and A.A. Witt. (193 5), Behavior of mixed populations and the problem of natural selection The American Naturalist, 69: 596 609. Hutchinson, G.E. (1991), Population studies: animal ecology and demography Bulletin of Mathematical Biology, 53: 193 213. Habitat Selection in Birds Academic Press, Inc.
Kupfer., J.A., and J.R. Runkle. (2003), Edge mediated effects on stand dynamic processes in forest interiors: a coupled field and simulation approach. Oikos, 101: 135 146 Lee, D., E. Kim., J. Choi, and K. Oh. (2010), The effects of development on forest patch characteristics and bird diversity in Suji, South Korea Landscape and Ecological Engineering, 6: 171 179. Sirami, C., L. Brotons, and J.L. Ma rtin. (2009), Do bird spatial distribution patterns reflect population trends in changing landscapes? Landscape Ecology, 24: 893 906. Stiles, F.G. and A.F. Skutch, (1989), A Guide to the Birds of Costa Rica Cornell University Press, Cornell, New York. Wilson, T.L., E.J. Johnson, and J.A. Bissonette. (2009), Relative importance of habitat area and isolation for bird occurrence patterns in a naturally patchy landscape Landscape Ecology, 24: 351 360. Winker, K., J.H. Rappole, and M.A. Ramos. (1995), The use of movement data as an assay of habitat quality. Oecologia, 101: 211 216.
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 M39-00225
El mosaico afecta la fragmentacin del hbitat y la particin del micro hbitat entre tres especies de soterr, en San Luis, Costa Rica.
The mosaic effect habitat fragmentation and microhabitat partitioning among three wren species in San Luis, Costa Rica
Increasing human presence in previously forested areas leads to the alteration and fragmentation of habitats used by resident species of birds. In this study I examine the effects of a mosaic-like arrangement of habitat types on the abundances and distributions of three species of wrens, the House Wren (Troglodytes aedon), Plain Wren
(Thryothorus modestus), and Rufous-and-white Wren (Thryothorbus rufalbus) to determine any resulting differences in abundances and distributions between species. I conducted visual and auditory observations of wrens
during timed walks through four habitat types in the San Luis valley of Costa Rica including primary forests, secondary forests, disturbed home and garden areas, and cleared pastures. Wrens were observed to partially overlap in range likely due to factors of proximity and niche partitioning on the basis of overall habitat, microhabitat, and edge characteristics, but not tree height. Overall wren abundance was equal between species but differed between
habitats, and abundance was higher in disturbed areas with high heterogeneity than in homogeneous disturbed areas.
Effects of human disturbance are thus demonstrated to vary based on structural characteristics in terms of species
El aumento de la presencia humana en las zonas previamente boscosas conduce a la alteracin y fragmentacin de los hbitats utilizados por las especies de aves residentes. En este estudio examino los efectos de un acuerdo de tipo mosaico de tipos de hbitat en la abundancia y distribucin de las tres especies de soterr, el soterr (Troglodytes aedon), el soterr (Thryothorus modestus), y el soterr (Thryothorbus rufalbus) para determinar las diferencias resultantes de la abundancia y la distribucin entre las especies. Lleve a cabo observaciones visuales y auditivas del soterr durante las caminatas cronometradas a travs de cuatro tipos de hbitat en el valle de San Luis de Costa Rica, incluyendo bosques primarios, bosques secundarios, reas del jardn, casas alterados y potreros despejados. Se observaron a los soterr solapar parcialmente en el rango probable debido a factores de proximidad y nicho de particin sobre la base del hbitat en general, micro hbitats y las caractersticas de borde, pero no la altura del rbol. En general la abundancia del soterr era igual entre las especies pero difiri entre los hbitats, y de la abundancia fue mayor en reas perturbadas con alta heterogeneidad que en las zonas perturbadas homogneas. Efectos de la perturbacin humana se demuestra a variar en funcin de las caractersticas estructurales en cuanto a la respuesta de las especies.
Text in English.
Costa Rica--Puntarenas--Monteverde Zone--San Luis
Costa Rica--Puntarenas--Zona de Monteverde--San Luis
Tropical Ecology Summer 2010
Ecologa Tropical Verano 2010
t Monteverde Institute : Tropical Ecology