- Email: [email protected]
The conservation value of linear forest remnants in central Amazonia
Biological Conservation 91 (1999) 241±247
www.elsevier.com/locate/biocon
The conservation value of linear forest remnants in central Amazonia Marcelo G. de Lima, Claude Gascon*,1 Biological Dynamics of Forest Fragments Project, National Institute for Research in the Amazon (INPA), CP 478, Manaus, AM 69011-970, Brazil
Abstract Riparian forest is protected under federal legislation in Brazil. In the Amazon Basin, numerous streams and rivers provide huge potential for increasing the conservation value of deforested and fragmented landscapes through the protection of linear remnants along watercourses. However, the potential of such remnants to be used as faunal habitat and possibly as movement corridors has never been fully investigated. We surveyed small mammal and litter-frog communities in linear remnants of primary rainforest ranging from 140 to 190 m in width, and in adjacent continuous rainforest, to compare their species richness, composition, and abundance. No signi®cant dierences were found in any aspect of community structure or species abundance. This suggests that linear remnants along watercourses provide suitable habitat for at least some forest vertebrates, a conclusion reinforced by the fact that many frogs and small mammals were found reproducing and moving in the remnants. These results highlight the potential of linear remnants to serve as habitat for small forest vertebrates and suggest they could function as corridors for some species to increase landscape connectivity. # 1999 Elsevier Science Ltd. All rights reserved. Keywords: Amazon rainforest; Corridors; Frogs; Habitat fragmentation; Landscape connectivity; Linear forest remnants; Small mammals
1. Introduction Rates of deforestation in the Amazon Basin have increased alarmingly in recent years (INPE, 1998), leading to rapid fragmentation of forest cover (Skole and Tucker, 1993). Habitat fragmentation dramatically alters forest dynamics (Lovejoy et al. 1986; Laurance et al., 1998a) and microclimatic variables like light, humidity, and wind near forest edges (Kapos, 1989; Murcia, 1995), which in turn alter plant and animal distributions (Lovejoy et al., 1986; Brown and Hutchings, 1997; Didham, 1997; Ferreira and Laurance, 1997; Kapos et al., 1997; Malcolm, 1997; Tocher et al., 1997; Carvalho and Vasconcelos, 1999). In the Amazon, many faunal groups are negatively aected by forest fragmentation (Rylands and Keuroghlian, 1988; Bierregaard and Stouer, 1997; Brown and Hutchings, 1997; Malcolm, 1997; Tocher et al., 1997). For forest fragments, the degree of isolation will determine, in part, the severity of changes in community composition (Laurance, 1991; Gascon et al., 1999). * Corresponding author. 1 Present address: Conservation International, 2501 M Street NW, Suite 200, Washington, DC 20037, USA. E-mail address: [email protected] (C. Gascon)
Increasing or maintaining landscape connectivity can reduce species extinctions and prevent inbreeding depression in isolated fragments (Noss, 1987; Bennett, 1990; Henein and Merriam, 1990; Mee and Carrol, 1994; Laurance and Gascon, 1997). Linear remnants of vegetation, such as riparian corridors, hedgerows, and gallery forests (Forman, 1997), may help maintain some level of connectivity in fragmented forest landscapes. Much debate revolves around the eectiveness of faunal corridors (Simberlo and Cox, 1987; Noss, 1987; Simberlo et al., 1991; Hobbs, 1992). A key criticism against corridors is the relative lack of empirical work demonstrating their eectiveness in increasing connectivity, but several recent studies have addressed this issue. DeMers (1993) observed ants (Pogomymex occidentalis) using roadside ditches to disperse during a drought. Ant-following birds were observed in a 100-ha forest fragment in the Amazon while it was connected to continuous forest by a 2 km-long corridor (Lovejoy et al., 1986), but when the corridor was severed, the birds disappeared. Bentley and Catterall (1997) found no dierences in the density of bird species between linear remnants and continuous forest in southeastern Queensland, suggesting that remnants provided useful bird habitat. Studies in Australia, North America, and Europe have recorded volant and nonvolant small
0006-3207/99/$ - see front matter # 1999 Elsevier Science Ltd. All rights reserved. PII: S0006-3207(99)00084-1
242
M.G. de Lima, C. Gascon / Biological Conservation 91 (1999) 241±247
mammals in linear strips of vegetation of diering sizes and widths, and posed recommendations for corridor design (Bennet, 1990; Lindenmayer and Nix, 1993; La Polla and Barrett, 1995; Ruefenacht and Knight, 1995; Andreassen et al., 1996a,b; Downs et al., 1998; Laurance and Laurance, 1999). Riparian areas in Illinois were also surveyed to assess their use as a dispersal corridor for frogs and lizards (Burbrink et al., 1998). In the central Amazon, long-term studies of small mammals and frogs have revealed that species vary considerably in their responses to fragmentation (Malcolm, 1997; Tocher et al., 1997). Surprisingly, both groups actually increase in species richness after fragmentation, a result mainly attributed to the establishment of opportunistic species associated with modi®ed matrix habitats (cattle pastures and regrowth forest) in fragments (Tocher et al., 1997). Some nominally forestdependent species can move through the disturbed matrix (Malcolm, 1997; Tocher et al., 1997; Gascon et al., 1999), while others are likely to require primary forest for movement. The goal of this study was to assess the conservation value of linear riparian remnants in a deforested Amazonian landscape, by comparing the community composition of small mammals and litter-dwelling frogs between linear remnants and adjoining areas of continuous forest. If the communities in remnants represent a biased or depauperate subset of those in continuous forest, then their eectiveness as habitat and potential movement corridors may be quite limited. 2. Methods 2.1. Study area Field work was carried out within the experimental landscape of the Biological Dynamic of Forest Fragments Project (BDFFP), located 70 km north of Manaus, Brazil (Lovejoy et al., 1986; Bierregaard et al., 1992). Forest in this area is terra ®rma (not seasonally ¯ooded), ca. 50-100 m elevation, and has very high tree species diversity (Rankin-de-Merona et al., 1992). The experimental design consisted of paired sites of linear streamside remnants and adjoining areas in continuous forest along the same stream. Adjoining sites were used as controls to minimize the eects of patchy species distributions. Four pairs of sites were sampled for small mammals and litter frogs (Cidade Powell, Florestal, Ponta Verde, and Area 51; Fig. 1). The remnants, which varied from 15 to 19 years in age, are primary forest retained by ranchers along small streams (<5 m width) when the surrounding forest was clearcut in the early 1980s. Most are now surrounded by regrowth forest (mainly dominated by Cecropia or Vismia spp. ), but all remain linked to continuous forest. The
remnants average 1225 m in length and 168 m in width, and have hilly topography (Table 1). 2.2. Small mammal sampling Small (<2 kg) mammals were sampled by live-trapping March±July and October±December, 1997. Five transects running parallel to the stream were established in each remnant and in the adjoining continuous forest. Forest and remnant transects were >100 m apart. Tomahawk traps of two sizes (ten 1655 inches and ten 2077 inches) were placed in an alternating sequence at 25 m intervals along each transect. Additionally, one Sherman trap was placed 2±3 m from each of the large Tomahawk traps. Traps were baited with peanut butter and bananas and run for eight consecutive nights, resulting in 9600 trap-nights of total eort. Site pairs (remnant-continuous forest) were sampled in consecutive weeks. Traps were checked during early morning and each animal was identi®ed, ear tagged, and released. 2.3. Litter-frog sampling At each site, a 375 m-long transect was used to sample litter frogs using 20 evenly spaced 55 m litter plots (Gascon, 1996). Forest and remnant transects were >100 m apart. A total of 160 plots were sampled. Plastic-mesh fences of 50 cm height were initially placed around the selected plot, and the area within the fence was carefully searched by two people for litter frogs. All frogs were identi®ed and released. Paired sites were sampled consecutively. 2.4. Statistical analysis For both small mammals and frogs, paired t-tests were used to compare species richness and total abundance between linear remnants and continuous forest. Abundances of the most common small mammal (Didelphis marsupialis, Metachirus nudicaudatus, Philander opossum, Proechimys sp., Oryzomys megacephalus) and litter-frog species (Adenomera andreae, Eleutherodactylus fenestratus, Colosthetus stepheni, Dendrophryniscus minutus) were also compared between treatments using paired t-tests. For each faunal group, dierences in species composition between linear remnants and continuous forest were tested using ANOSIM (Clarke, 1993), using the PATN program (Belbin, 1991). This test uses the ratio of the average dissimilarity between groups to the average dissimilarity within groups to determine whether groups of sites with similar species composition exist (Belbin, 1991). In this case, we tested for the existence of distinctive ``remnant'' and ``continuous forest'' groups. An ordination analysis using Detrended Correspondence Analysis (DCA) was performed on each faunal
M.G. de Lima, C. Gascon / Biological Conservation 91 (1999) 241±247
243
Fig. 1. Study area in the central Amazon showing paired sites with linear remnants and adjoining continuous forest. Shaded areas represent deforested land, while unstippled areas are rainforest. Light wavy lines are streams while the bold line is an access road.
Table 1 Characteristics of linear rainforest remnants studied in the central Amazon Remnant
Mean width (m)
Length (m)
Year created
Topography
Dominant surrounding vegetation
Cidade Powell Florestal Ponta Verde Area 51
140 180 160 190
1200 700 1400 1600
1983 1983 1983 1981
Steep sinuous valley Steep valley with some plateaus on each side of stream Lowland habitat adjacent to stream Steep sinuous valley
Cecropia sp. Vismia sp./Cecropia sp. Cecropia sp. Vismia sp.
group to illustrate relationships between sampled localities. Species captured only once were excluded from these analyses. The Sorensen index was used to generate similarity matrices, using the statistical package PCORD (McCune and Metford, 1995). 3. Results 3.1. Small mammal richness and abundance A total of 115 individuals of 14 mammal species were captured (Table 2). At least two Proechimys species
were captured (P. cuvieri and P. cayennensis) but were lumped in the following analysis because of diculties in ®eld identi®cation (Malcolm, 1991). D. marsupialis and M. nudicaudatus were the most common species, occurring at seven of eight sites and representing 43% of all captures. No signi®cant dierence was found in species richness or abundance of the most common species between remnants and continuous forest (Table 2). (Richness: mammals, t=0, P=1.00; frogs, t=1.0, P=0.39) (Abundance of mammals: D. marsupialis t=0, P=1.00; M. nudicaudatus, t=ÿ0.81, P=0.47, P. opossum, t=1.66, P=0.19, Proechimys sp., t=0.58, P=0.60, O.megacephalus, t=ÿ1.32, P=0.28; frogs: A. andreae,
244
M.G. de Lima, C. Gascon / Biological Conservation 91 (1999) 241±247
t=1.46, P=0.24; E. fenestratus, t=1.41, P=0.25; C. stepheni, t=1.16, P=0.33; D. minutus, t=0.50, P=0.65). 3.2. Mammal species composition The ANOSIM failed to identify any group structure based on the species composition (F=0.94, P=0.78), indicating that small mammals communities in forest remnants could not be distinguished from those in continuous forest. The ordination analysis revealed two gradients in species composition, with axis 1 describing a gradient in abundance of P. opossum, and axis 2 re¯ecting variation
in Marmosa murina and Micoureus demerarae (Table 3). No natural grouping of sites according to treatments was apparent (Fig. 2). 3.3. Frog richness and abundance We captured 231 individuals of 12 frog species (Table 2). C. stepheni and A. andreae represented the majority of captures (40 and 33%, respectively) and occurred in all eight sites. E. fenestratus, C. marchesianus, Atelopus pulcher, and D. minutus collectively represented 22% of captures, while Eleutherodactylus zimmermanae, Leptodactylus rhodomystax, L. mystaceus, Bufo typhonius,
Table 2 Abundances (No. of individuals captured) of (A) small mammals and (B) litter frogs in continuous forest and linear forest remnants in the central Amazon Cidade Powell
Florestal
Ponta Verde
Area 51
Remnant
Forest
Remnant
Forest
Remnant
Forest
Remnant
Forest
Didelphidae Didelphis marsupialis Metachirus nudicaudatus Philander opossum Marmosops parvidens Marmosa murina Micoureus demerarae Monodelphis brevicaudata Calouromys philander
± 2 3 ± ± ± ± ±
4 3 1 ± 1 ± ± ±
10 2 ± ± 3 1 2 1
3 6 ± 2 3 2 ± ±
3 1 4 ± ± ± ± ±
7 2 1 1 ± ± ± ±
3 2
2 ±
2 ±
± ±
1 ±
± ±
Echimydae Proechimys sp.
±
2
3
1
1
±
3
2
Muridae Oryzomys megacephalus Oryzomys macconelli Oecomys sp. Nacomys sp.
± ± ± ±
1 ± ± 1
± ± 3 ±
± ± 2 ±
2 ± 1 ±
6 ± ± ±
2 4 ± ±
2 ± ± ±
Dasyproctidae Myoprocta acouchy
±
±
±
±
1
±
±
±
Leptodactylidae Adenomera andreae Eleutherodactylus fenestratus Eleutherodactylus zimmermaneae Leptodactylus rhodomystax Leptodactylus mystaceus
29 5 1 ± ±
17 2 ± ± ±
7 3 ± 1 ±
4 3 ± ± ±
6 ± 1 ± ±
6 ± 1 ± ±
5 1 ± ± 1
4 ± ± ± ±
Dendrobatidae Colosthetus stepheni Colosthetus marchesianus
12 3
10 3
9 ±
5 ±
8 ±
11 ±
29 ±
10 2
Hylidae Hyla minuta
±
±
±
±
±
±
±
Microhylidae Synapturanus sp.
±
±
±
2
±
±
±
±
Bufonidae Bufo typhonius Atelopus pulcher Dendrophryniscus minutus
± 6 11
± 1 1
± ± ±
± ± ±
± ± 1
± ± 5
2 1 ±
± 1 ±
A. Small mammals
B. Litter frogs
1
M.G. de Lima, C. Gascon / Biological Conservation 91 (1999) 241±247
245
Table 3 Correlation coecients for (A) small mammals and (B) litter frogs with the ®rst two axes produced by detrended correspondance analysis Speciesa
Axis 1
Axis 2
ÿ0.558 ÿ0.483 0.794* ÿ0.454 ÿ0.791* ÿ0.712* ÿ0.565 ÿ0.664 0.245 ÿ0.663
0.476 ÿ0.746* ÿ0.241 ÿ0.386 ÿ0.200 ÿ0.428 0.423 0.577 0.340 0.002
0.822* 0.527 0.564 ÿ0.501 0.644 0.638 0.808*
0.008 0.648 ÿ0.584 ÿ0.293 ÿ0.086 ÿ0.094 ÿ0.275
A. Mammals Didelphis marsupialis Metachirus nubicauidata Philander opossum Marmosops parvidens Marmosa murina Micoureus demerarae Monodelphis brevicaudata Proechimys sp. Oryzomys megacephalus Oecomys sp. B. Frogs Adenomera andreae Eleutherodactylus fenestratus Eleutherodactylus zimmermaneae Colosthetus stepheni Colosthatus marchesianus Atelopus pulcher Dendrophryniscus minutus
a Species with an asterix were signi®cantly correlated (P<0.05) with that axis.
a similar species composition, as did the paired sites at Area 51.
Fig. 2. (A) Ordination of linear remnants and continuous forest sites based on composition of small mammal communities. (B) Ordination of the same sites based on composition of frog communities. Site abbreviations: CPR=Cidade Powell linear remnant; CPF=Cidade Powell continuous forest; FR=Florestal linear remnant; FF=Florestal forest; PVR=Ponta Verde linear remnant; PVF=Ponta Verde forest; A51R=Area 51 linear remnant; A51F=Area 51 forest.
Hyla minuta, and Synapturanus sp. occurred at only a single site and were each represented by just 1±2 individuals. Neither species richness nor the abundance of the most common species diered between treatments (Table 2). 3.4. Frog species composition The ANOSIM failed to detect evidence of distinctive groupings of forest and remnant frogs (F=0.91, P=0.97), again indicating that community composition of linear remnants was not systematically dierent from that in continuous forest. The ordination also did not reveal any grouping according to treatments (Table 3; Fig. 2). However, some pairs of adjoining sites did appear to be similar in composition. The remnant and continuous-forest site at Ponta Verde, for example, had
4. Discussion The linear remnants surveyed in this study were not retained for the purpose of protecting wildlife, but rather for watershed conservation. Although considerably wider (140±190 m) than is required by Brazilian legislation (minimum of 60 m for streams of under 10 m in width; Laurance and Gascon, 1997), these remnants were still vulnerable to physical and biotic edge eects, which have been detected up to 200±300 m from forest margins (Brown and Hutchings, 1997; Malcolm, 1997; Laurance and Bierregaard, 1997; Laurance et al., 1998a; Stevens and Husband, 1998). The fact that we detected no consistent dierences in species richness, composition, or abundance for litter frogs and small mammals, however, suggests that edge eects in remnants of moderate width (140±190 m) were tolerated by these taxa. Interestingly, species such as Calouromys philander and Atelopus pulcher, which appear sensitive to habitat disturbance and fragmentation (Malcolm, 1991; Tocher, 1998), were detected in linear remnants in this study. Nevertheless, as a group, small mammals and frogs are less sensitive to fragmentation than many other vertebrate groups, and a number of the species captured in this study can also use the matrix of modi®ed habitats surrounding fragments (Gascon et al.,
246
M.G. de Lima, C. Gascon / Biological Conservation 91 (1999) 241±247
1999). It should also be noted that the linear remnants we studied were directly linked to nearby continuous forest, and thus would be analogous to short (<1 km long) corridors. Our results suggest that linear remnants of moderate width could have important conservation value in the context of a deforested landscape, at least for some smaller (<2 kg) forest vertebrates. The remnants probably function as movement corridors and breeding habitat for some species. For example, we recorded marked individuals of D. marsupialis, P. opossum and M. murina moving between remnants and adjoining continuous forest, while the presence of juvenile M. nudicaudatus, D. marsupialis, and P. opossum in some remnants (de Lima, 1998) suggests that these species were reproducing there. Frogs of several species (C. stepheni, E. fenestratus, and D. minutus) were also heard vocalizing in remnants, which is a good indicator of mating activity. Likewise, in tropical Queensland, some arboreal mammal species were commonly observed with young in linear rainforest remnants (S. G. Laurance, 1996). The design of faunal corridors will depend on the species of interest and local context. For example, Andreassen et al. (1996a,b) showed that voles (Microtus oeconomus) could move quite easily through very narrow corridors (<0.5 m). In tropical regions, however, much wider corridors will be needed, both because many tropical species are forest-interior specialists (e.g. Lovejoy et al., 1986; Bierregaard et al., 1992; Laurance et al., 1998b) and because closed tropical forests are very prone to physical and biotic edge eects (Laurance and Bierregaard, 1997; Carvalho and Vasconcelos, 1999). In these ecosystems, remnants of at least 200±300 m width may be necessary to facilitate movements of some sensitive species (Laurance and Gascon, 1997; Laurance and Laurance, 1999), and even greater widths may be needed for long (>5 km) corridors. In addition to providing forest-interior habitat, wider remnants have greater spatial heterogeneity, oering a broader array of microhabitats which is often correlated with greater species richness (Lindenmayer and Nix, 1993; Mee and Carrol, 1994). There is a great need to improve outdated legislation in Brazil to require the retention of wider (200±300 m) linear remnants along streams and rivers (Laurance and Gascon, 1997). There is also an urgent requirement to increase enforcement of existing legislation, which is often ignored by landowners. Educational initiatives are direly needed at all levels Ð from landowners to resource-managers to policymakers Ð to improve understanding of the key importance of retaining riparian vegetation in deforested landscapes. Given the explosive pace of development in the Amazon (Laurance, 1998), proactive initiatives to manage the deforestation process could go a long way toward maintaining some
level of ecosystem connectivity in human-dominated regions. Acknowledgements We thank O. de S. Pereira and A. M. dos Reis for help with ®eld work. Support was provided by the National Institute for Research in the Amazon (INPA), Smithsonian Institution, Brazilian National Council for Research (CNPq), USAID-Brazil, The MacArthur Foundation, and The Summit Foundation. W.F. Laurance, W.Z. Lidicker, and W. Magnusson provided valuable comments on earlier versions of the paper, and W.F. Laurance helped in ®nal editing. This is publication number 231 of the BDFFP technical series. References Andreassen, H.P., Ims, R.A., Steinset, O.K., 1996a. Discontinuous habitat corridors: eects on male root vole movements. Journal of Applied Ecology 33, 555±560. Andreassen, H.P., Halle, S., Ims, R.A., 1996b. Optimal width of movement corridors for root voles: not too narrow and not too wide. Journal of Applied Ecology 33, 63±70. Belbin, L., 1991. The analysis of pattern in bio-survey data. In: Margules, C.R., Austin, M.P. (Eds.), Nature Conservation: Cost Eective Biological Surveys and Data Analysis. CSIRO, Melbourne, Australia, pp. 176±190. Bennett, A.F., 1990. Habitat corridors and the conservation of small mammals in a fragmented forest environment. Landscape Ecology 4, 109±122. Bentley, J.M., Catterall, C.P., 1997. The use of bushland, corridors, and linear remnants by birds in Southeastern Queensland, Australia. Conservation Biology 11, 1173±1189. Bierregaard, R.O., Stouer, P.C., 1997. Understory birds and dynamic habitat mosaics in Amazonian rainforest. In: Laurance, W.F., Bierregaard, R.O. (Eds.), Tropical Forest Remnants: Ecology, Management, and Conservation of Fragmented Communities. University of Chicago Press, Chicago, pp. 138±155. Bierregaard Jr., R.O., Lovejoy, T.E., Kapos, V., dos Santos, A.A., Hutchings, R.W., 1992. The biological dynamics of tropical rainforest fragments. Bioscience 42, 859±866. Brown Jr., K.S., Hutchings, R.W., 1997. Disturbance, fragmentation, and the dynamics of diversity in Amazonian Forest Butter¯ies. In: Laurance, W.F., Bierregaard, R.O. (Eds.), Tropical Forest Remnants: Ecology, Management, and Conservation of Fragmented Communities. University of Chicago Press, Chicago, pp. 91±110. Burbrink, F.T., Phillips, C.A., Heske, E.J., 1998. A riporian zone in Southern Illinois as a potential dispersal corridor for reptiles and amphibians. Biological Conservation 86, 107±115. Carvalho, K.S., Vasconcelos, H.L., 1999. Forest fragmentation in central Amazonia and its eects on litter-dwelling ants. Biological Conservation. Clarke, K.R., 1993. Non-parametric multivariate analysis of changes community structure. Australian Journal of Ecology 18, 117±143. de Lima, M.G., 1998. ComposicËaÄo da comunidade de pequenos mamõÂferos e sapos de liteira em remanescentes lineares no Projeto DinaÃmica BioloÂgica de Fragmentos Florestais (PDBFF). M.Sc. thesis, National Institute for Research in the Amazon (INPA), Manaus, Brazil. DeMers, M.N., 1993. Roadside ditches as corridors for range expan-
M.G. de Lima, C. Gascon / Biological Conservation 91 (1999) 241±247 sion of the western harvester ant (Pogonomyrmex occidentalis Cresson). Landscape Ecology 8, 93±102. Didham, R.K., 1997. The in¯uence of edge eects and forest fragmentation on leaf-litter invertebrates in central Amazonia. In: Laurance, W.F., Bierregaard, R.O. (Eds.), Tropical Forest Remnants: Ecology, Management, and Conservation of Fragmented Communities. University of Chicago Press, Chicago, pp. 55±70. Downs, N.C., Racey, P.A., Weller, N.C., 1998. Do bats believe in the wildlife corridor hypothesis? Abstracts of the 11th International Bat Research Conference, Universidade de BrasõÂlia, BrasõÂlia, Brazil, p. 25. Ferreira, L.V., Laurance, W.F., 1997. Eects of forest fragmentation on mortality and damage of selected trees in central Amazonia. Conservation Biology 11, 797±801. Forman, R.T.T., 1997. Land Mosiacs. Cambridge University Press, Cambridge, UK. Gascon, C., 1996. Amphibian litter fauna and river barriers in ¯ooded and non-¯ooded Amazonian rain forests. Biotropica 28, 136±140. Gascon, C., Lovejoy, T.E., Bierregaard Jr., R. O., Malcolm, J.R., Stouer, P.C., Vasconcelos, H., Laurance, W.F., Zimmerman, B., Tocher, M. & Borges, S., 1999. Matrix habitat and species persistence in tropical forest remnants. Biological Conservation 91, 231± 239. Henein, K., Merriam, G., 1990. The elements of connectivity where corridor quality is variable. Landscape Ecology 4, 157±170. Hobbs, R.J., 1992. The role of corridors in conservation: Solution or bandwagon? Trends in Ecology and Evolution 7, 389±391. INPE, 1998. De¯orestamento, 1995-1997. Instituto Nacional de Pesquisas Espaciais. San Jose dos Campos, Brazil. Kapos, V., 1989. Eects of isolation on the water status of forest patches in the Brazilian Amazon. Jounal of Tropical Ecology 5, 173± 185. Kapos, V., Wandelli, E., Camargo, J.L., Ganade, G., 1997. Edgerelated changes in environment and plant responses due to forest fragmentation in central Amazonia. In: Laurance, W.F., Bierregaard, R.O. (Eds.), Tropical Forest Remnants: Ecology, Management, and Conservation of Fragmented Communities. University of Chicago Press, Chicago, pp. 33±54. La Polla, V.N., Barrett, G.W., 1995. Eects of corridor width and presence on the population dynamics of the meadow vole (Microtus pennsylvanicus). Landscape Ecology 8, 25±37. Laurance, S.G., 1996. The utilisation of linear rainforest remnants by arboreal marsupials in north Queensland. M.Nat.Res. thesis, University of New England, Armidale, Australia. Laurance, S.G., Laurance, W.F., 1999. Tropical corridors: use of linear rainforest remnants by arboreal mammals. Biological Conservation. Laurance, W.F., 1991. Ecological correlates of extinction proneness in Australian tropical rain forest mammals. Conservation Biology 5, 79±89. Laurance, W.F., 1998. A crisis in the making: Responses of Amazonian forests to land use and climate change. Trends in Ecology and Evolution 11, 411±415. Laurance, W.F., Gascon, C., 1997. How to creatively fragment a landscape. Conservation Biology 11, 577±579. Laurance, W.F., Bierregaard, R.O. (Eds.), 1997. Tropical Forest Remnants: Ecology, Management, and Conservation of Fragmented Communities. University of Chicago Press, Chicago. Laurance, W.F., Ferreira, L.V., Rankin-de Merona, J.M., Laurance, S.G., 1998a. Rain forest fragmentation and the dynamics of Amazonian tree communities. Ecology 69, 2032±2040. Laurance, W.F., Ferreira, L.V., Rankin-de Merona, J.M., Laurance,
247
S.G., Hutchings, R., Lovejoy, T.E., 1998b. Eects of forest fragmentation on recruitment patterns in Amazonian tree communities. Conservation Biology 12, 460±464. Lindenmayer, D.B., Nix, H.A., 1993. Ecological principles for the design of wildlife corridors. Conservation Biology 7, 627±630. Lovejoy, T.E., Bierregaard Jr., R.O., Rylands, A.B., Malcolm, J.R., Quintela, C.E., Harper, L.H., Brown Jr., K.S., Powell, A.H., Powell, G.V.N., Schubart, H.O.R., Hays, M.B., 1986. Edge eects and other eects of isolation on Amazon forest fragments. In: SouleÂ, M.E. (Ed.), Conservation Biology: The Science of Scarcity and Diversity. Sinauer Associates, Sunderland, Massachusetts, pp. 257±285. Malcolm, J.R., 1991. The small mammals of Amazonian forest fragments: Pattern and process. Ph.D. thesis, University of Florida, Gainesville, Florida. Malcolm, J.R., 1997. Biomass and diversity of small mammals in Amazonian forest fragments. In: Laurance, W.F., Bierregaard, R.O. (Eds.), Tropical Forest Remnants: Ecology, Management, and Conservation of Fragmented Communities. University of Chicago Press, Chicago, pp. 207±221. McCune, B., Meord, M.G., 1995. PC-ORD: Multivariate Analysis of Ecological Data, Version 2.0. MjM Software Design, Gleneden Beach, Oregon, USA. Mee, G.K., Carroll, C.R., 1994. What is conservation biology?. In: Mee, G.K., Carroll, C.R. (Eds.), Principles of Conservation Biology. Sinauer Associates, Sunderland, Massachusetts, pp. 3±23. Murcia, C., 1995. Edge eects in fragmented forests: Implications for conservation. Trends in Ecology and Evolution 10, 58±62. Noss, R.F., 1987. Corridors in real landscapes: A reply to Simberlo and Cox. Conservation Biology 1, 159±164. Rankin-de Merona, J.M., Prance, J.M., Hutchings, R.W., Silva, M.F., Rodrigues, M.A., Uehling, M.E., 1992. Preliminary results of a large-scale inventory of upland rain forest in the central Amazon. Acta Amazonica 22, 493±534. Ruefenacht, B., Knight, R.L., 1995. In¯uences of corridor continuity and width on the survival and movement of deermice Peromyscus maniculatus. Biological Conservation 71, 269±274. Rylands, A.B., Keuroghlian, A., 1988. Primate populations in continuous forest and forest fragments in central Amazonia. Acta AmazoÃnica 18, 67±83. Simberlo, D., Cox, J., 1987. Consequences and costs of conservation corridors. Conservation Biology 1, 63±71. Simberlo, D., Farr, J.A., Cox, J., Mehlman, D.W., 1991. Movement corridors: Conservation bargains or poor investments. Conservation Biology 6, 493±503. Skole, D., Tucker, C., 1993. Tropical deforestation and habitat fragmentation in the Amazon: Satellite data from 1978 to 1988. Science 260, 1905±1910. Stevens, S.M., Husband, T.P., 1998. The in¯uence of edge on small mammals: Evidence from Brazilian Atlantic forest fragments. Biological Conservation 85, 1±8. Tocher, M., 1998. A comunidade de anfõÂbios da AmazoÃnia central: DiferencËas na composicËaÄo especõ®ca entre a mata primaÂria e pastagens. In: Gascon, C., Moutinho, P. (Eds.), Floresta AmazoÃnica: DinaÃmica, RegeneracËaÄo e Manejo. National Institute for Research in the Amazon (INPA), Manaus, Brazil, pp. 219±232. Tocher, M.D., Gascon, C., Zimmerman, B.L., 1997. Fragmentation eects on a central Amazonian frog community: a ten year study. In: Laurance, W.F., Bierregaard, R.O. (Eds.), Tropical Forest Remnants: Ecology, Management, and Conservation of Fragmented Communities. University of Chicago Press, Chicago, pp. 124±137.
www.elsevier.com/locate/biocon
The conservation value of linear forest remnants in central Amazonia Marcelo G. de Lima, Claude Gascon*,1 Biological Dynamics of Forest Fragments Project, National Institute for Research in the Amazon (INPA), CP 478, Manaus, AM 69011-970, Brazil
Abstract Riparian forest is protected under federal legislation in Brazil. In the Amazon Basin, numerous streams and rivers provide huge potential for increasing the conservation value of deforested and fragmented landscapes through the protection of linear remnants along watercourses. However, the potential of such remnants to be used as faunal habitat and possibly as movement corridors has never been fully investigated. We surveyed small mammal and litter-frog communities in linear remnants of primary rainforest ranging from 140 to 190 m in width, and in adjacent continuous rainforest, to compare their species richness, composition, and abundance. No signi®cant dierences were found in any aspect of community structure or species abundance. This suggests that linear remnants along watercourses provide suitable habitat for at least some forest vertebrates, a conclusion reinforced by the fact that many frogs and small mammals were found reproducing and moving in the remnants. These results highlight the potential of linear remnants to serve as habitat for small forest vertebrates and suggest they could function as corridors for some species to increase landscape connectivity. # 1999 Elsevier Science Ltd. All rights reserved. Keywords: Amazon rainforest; Corridors; Frogs; Habitat fragmentation; Landscape connectivity; Linear forest remnants; Small mammals
1. Introduction Rates of deforestation in the Amazon Basin have increased alarmingly in recent years (INPE, 1998), leading to rapid fragmentation of forest cover (Skole and Tucker, 1993). Habitat fragmentation dramatically alters forest dynamics (Lovejoy et al. 1986; Laurance et al., 1998a) and microclimatic variables like light, humidity, and wind near forest edges (Kapos, 1989; Murcia, 1995), which in turn alter plant and animal distributions (Lovejoy et al., 1986; Brown and Hutchings, 1997; Didham, 1997; Ferreira and Laurance, 1997; Kapos et al., 1997; Malcolm, 1997; Tocher et al., 1997; Carvalho and Vasconcelos, 1999). In the Amazon, many faunal groups are negatively aected by forest fragmentation (Rylands and Keuroghlian, 1988; Bierregaard and Stouer, 1997; Brown and Hutchings, 1997; Malcolm, 1997; Tocher et al., 1997). For forest fragments, the degree of isolation will determine, in part, the severity of changes in community composition (Laurance, 1991; Gascon et al., 1999). * Corresponding author. 1 Present address: Conservation International, 2501 M Street NW, Suite 200, Washington, DC 20037, USA. E-mail address: [email protected] (C. Gascon)
Increasing or maintaining landscape connectivity can reduce species extinctions and prevent inbreeding depression in isolated fragments (Noss, 1987; Bennett, 1990; Henein and Merriam, 1990; Mee and Carrol, 1994; Laurance and Gascon, 1997). Linear remnants of vegetation, such as riparian corridors, hedgerows, and gallery forests (Forman, 1997), may help maintain some level of connectivity in fragmented forest landscapes. Much debate revolves around the eectiveness of faunal corridors (Simberlo and Cox, 1987; Noss, 1987; Simberlo et al., 1991; Hobbs, 1992). A key criticism against corridors is the relative lack of empirical work demonstrating their eectiveness in increasing connectivity, but several recent studies have addressed this issue. DeMers (1993) observed ants (Pogomymex occidentalis) using roadside ditches to disperse during a drought. Ant-following birds were observed in a 100-ha forest fragment in the Amazon while it was connected to continuous forest by a 2 km-long corridor (Lovejoy et al., 1986), but when the corridor was severed, the birds disappeared. Bentley and Catterall (1997) found no dierences in the density of bird species between linear remnants and continuous forest in southeastern Queensland, suggesting that remnants provided useful bird habitat. Studies in Australia, North America, and Europe have recorded volant and nonvolant small
0006-3207/99/$ - see front matter # 1999 Elsevier Science Ltd. All rights reserved. PII: S0006-3207(99)00084-1
242
M.G. de Lima, C. Gascon / Biological Conservation 91 (1999) 241±247
mammals in linear strips of vegetation of diering sizes and widths, and posed recommendations for corridor design (Bennet, 1990; Lindenmayer and Nix, 1993; La Polla and Barrett, 1995; Ruefenacht and Knight, 1995; Andreassen et al., 1996a,b; Downs et al., 1998; Laurance and Laurance, 1999). Riparian areas in Illinois were also surveyed to assess their use as a dispersal corridor for frogs and lizards (Burbrink et al., 1998). In the central Amazon, long-term studies of small mammals and frogs have revealed that species vary considerably in their responses to fragmentation (Malcolm, 1997; Tocher et al., 1997). Surprisingly, both groups actually increase in species richness after fragmentation, a result mainly attributed to the establishment of opportunistic species associated with modi®ed matrix habitats (cattle pastures and regrowth forest) in fragments (Tocher et al., 1997). Some nominally forestdependent species can move through the disturbed matrix (Malcolm, 1997; Tocher et al., 1997; Gascon et al., 1999), while others are likely to require primary forest for movement. The goal of this study was to assess the conservation value of linear riparian remnants in a deforested Amazonian landscape, by comparing the community composition of small mammals and litter-dwelling frogs between linear remnants and adjoining areas of continuous forest. If the communities in remnants represent a biased or depauperate subset of those in continuous forest, then their eectiveness as habitat and potential movement corridors may be quite limited. 2. Methods 2.1. Study area Field work was carried out within the experimental landscape of the Biological Dynamic of Forest Fragments Project (BDFFP), located 70 km north of Manaus, Brazil (Lovejoy et al., 1986; Bierregaard et al., 1992). Forest in this area is terra ®rma (not seasonally ¯ooded), ca. 50-100 m elevation, and has very high tree species diversity (Rankin-de-Merona et al., 1992). The experimental design consisted of paired sites of linear streamside remnants and adjoining areas in continuous forest along the same stream. Adjoining sites were used as controls to minimize the eects of patchy species distributions. Four pairs of sites were sampled for small mammals and litter frogs (Cidade Powell, Florestal, Ponta Verde, and Area 51; Fig. 1). The remnants, which varied from 15 to 19 years in age, are primary forest retained by ranchers along small streams (<5 m width) when the surrounding forest was clearcut in the early 1980s. Most are now surrounded by regrowth forest (mainly dominated by Cecropia or Vismia spp. ), but all remain linked to continuous forest. The
remnants average 1225 m in length and 168 m in width, and have hilly topography (Table 1). 2.2. Small mammal sampling Small (<2 kg) mammals were sampled by live-trapping March±July and October±December, 1997. Five transects running parallel to the stream were established in each remnant and in the adjoining continuous forest. Forest and remnant transects were >100 m apart. Tomahawk traps of two sizes (ten 1655 inches and ten 2077 inches) were placed in an alternating sequence at 25 m intervals along each transect. Additionally, one Sherman trap was placed 2±3 m from each of the large Tomahawk traps. Traps were baited with peanut butter and bananas and run for eight consecutive nights, resulting in 9600 trap-nights of total eort. Site pairs (remnant-continuous forest) were sampled in consecutive weeks. Traps were checked during early morning and each animal was identi®ed, ear tagged, and released. 2.3. Litter-frog sampling At each site, a 375 m-long transect was used to sample litter frogs using 20 evenly spaced 55 m litter plots (Gascon, 1996). Forest and remnant transects were >100 m apart. A total of 160 plots were sampled. Plastic-mesh fences of 50 cm height were initially placed around the selected plot, and the area within the fence was carefully searched by two people for litter frogs. All frogs were identi®ed and released. Paired sites were sampled consecutively. 2.4. Statistical analysis For both small mammals and frogs, paired t-tests were used to compare species richness and total abundance between linear remnants and continuous forest. Abundances of the most common small mammal (Didelphis marsupialis, Metachirus nudicaudatus, Philander opossum, Proechimys sp., Oryzomys megacephalus) and litter-frog species (Adenomera andreae, Eleutherodactylus fenestratus, Colosthetus stepheni, Dendrophryniscus minutus) were also compared between treatments using paired t-tests. For each faunal group, dierences in species composition between linear remnants and continuous forest were tested using ANOSIM (Clarke, 1993), using the PATN program (Belbin, 1991). This test uses the ratio of the average dissimilarity between groups to the average dissimilarity within groups to determine whether groups of sites with similar species composition exist (Belbin, 1991). In this case, we tested for the existence of distinctive ``remnant'' and ``continuous forest'' groups. An ordination analysis using Detrended Correspondence Analysis (DCA) was performed on each faunal
M.G. de Lima, C. Gascon / Biological Conservation 91 (1999) 241±247
243
Fig. 1. Study area in the central Amazon showing paired sites with linear remnants and adjoining continuous forest. Shaded areas represent deforested land, while unstippled areas are rainforest. Light wavy lines are streams while the bold line is an access road.
Table 1 Characteristics of linear rainforest remnants studied in the central Amazon Remnant
Mean width (m)
Length (m)
Year created
Topography
Dominant surrounding vegetation
Cidade Powell Florestal Ponta Verde Area 51
140 180 160 190
1200 700 1400 1600
1983 1983 1983 1981
Steep sinuous valley Steep valley with some plateaus on each side of stream Lowland habitat adjacent to stream Steep sinuous valley
Cecropia sp. Vismia sp./Cecropia sp. Cecropia sp. Vismia sp.
group to illustrate relationships between sampled localities. Species captured only once were excluded from these analyses. The Sorensen index was used to generate similarity matrices, using the statistical package PCORD (McCune and Metford, 1995). 3. Results 3.1. Small mammal richness and abundance A total of 115 individuals of 14 mammal species were captured (Table 2). At least two Proechimys species
were captured (P. cuvieri and P. cayennensis) but were lumped in the following analysis because of diculties in ®eld identi®cation (Malcolm, 1991). D. marsupialis and M. nudicaudatus were the most common species, occurring at seven of eight sites and representing 43% of all captures. No signi®cant dierence was found in species richness or abundance of the most common species between remnants and continuous forest (Table 2). (Richness: mammals, t=0, P=1.00; frogs, t=1.0, P=0.39) (Abundance of mammals: D. marsupialis t=0, P=1.00; M. nudicaudatus, t=ÿ0.81, P=0.47, P. opossum, t=1.66, P=0.19, Proechimys sp., t=0.58, P=0.60, O.megacephalus, t=ÿ1.32, P=0.28; frogs: A. andreae,
244
M.G. de Lima, C. Gascon / Biological Conservation 91 (1999) 241±247
t=1.46, P=0.24; E. fenestratus, t=1.41, P=0.25; C. stepheni, t=1.16, P=0.33; D. minutus, t=0.50, P=0.65). 3.2. Mammal species composition The ANOSIM failed to identify any group structure based on the species composition (F=0.94, P=0.78), indicating that small mammals communities in forest remnants could not be distinguished from those in continuous forest. The ordination analysis revealed two gradients in species composition, with axis 1 describing a gradient in abundance of P. opossum, and axis 2 re¯ecting variation
in Marmosa murina and Micoureus demerarae (Table 3). No natural grouping of sites according to treatments was apparent (Fig. 2). 3.3. Frog richness and abundance We captured 231 individuals of 12 frog species (Table 2). C. stepheni and A. andreae represented the majority of captures (40 and 33%, respectively) and occurred in all eight sites. E. fenestratus, C. marchesianus, Atelopus pulcher, and D. minutus collectively represented 22% of captures, while Eleutherodactylus zimmermanae, Leptodactylus rhodomystax, L. mystaceus, Bufo typhonius,
Table 2 Abundances (No. of individuals captured) of (A) small mammals and (B) litter frogs in continuous forest and linear forest remnants in the central Amazon Cidade Powell
Florestal
Ponta Verde
Area 51
Remnant
Forest
Remnant
Forest
Remnant
Forest
Remnant
Forest
Didelphidae Didelphis marsupialis Metachirus nudicaudatus Philander opossum Marmosops parvidens Marmosa murina Micoureus demerarae Monodelphis brevicaudata Calouromys philander
± 2 3 ± ± ± ± ±
4 3 1 ± 1 ± ± ±
10 2 ± ± 3 1 2 1
3 6 ± 2 3 2 ± ±
3 1 4 ± ± ± ± ±
7 2 1 1 ± ± ± ±
3 2
2 ±
2 ±
± ±
1 ±
± ±
Echimydae Proechimys sp.
±
2
3
1
1
±
3
2
Muridae Oryzomys megacephalus Oryzomys macconelli Oecomys sp. Nacomys sp.
± ± ± ±
1 ± ± 1
± ± 3 ±
± ± 2 ±
2 ± 1 ±
6 ± ± ±
2 4 ± ±
2 ± ± ±
Dasyproctidae Myoprocta acouchy
±
±
±
±
1
±
±
±
Leptodactylidae Adenomera andreae Eleutherodactylus fenestratus Eleutherodactylus zimmermaneae Leptodactylus rhodomystax Leptodactylus mystaceus
29 5 1 ± ±
17 2 ± ± ±
7 3 ± 1 ±
4 3 ± ± ±
6 ± 1 ± ±
6 ± 1 ± ±
5 1 ± ± 1
4 ± ± ± ±
Dendrobatidae Colosthetus stepheni Colosthetus marchesianus
12 3
10 3
9 ±
5 ±
8 ±
11 ±
29 ±
10 2
Hylidae Hyla minuta
±
±
±
±
±
±
±
Microhylidae Synapturanus sp.
±
±
±
2
±
±
±
±
Bufonidae Bufo typhonius Atelopus pulcher Dendrophryniscus minutus
± 6 11
± 1 1
± ± ±
± ± ±
± ± 1
± ± 5
2 1 ±
± 1 ±
A. Small mammals
B. Litter frogs
1
M.G. de Lima, C. Gascon / Biological Conservation 91 (1999) 241±247
245
Table 3 Correlation coecients for (A) small mammals and (B) litter frogs with the ®rst two axes produced by detrended correspondance analysis Speciesa
Axis 1
Axis 2
ÿ0.558 ÿ0.483 0.794* ÿ0.454 ÿ0.791* ÿ0.712* ÿ0.565 ÿ0.664 0.245 ÿ0.663
0.476 ÿ0.746* ÿ0.241 ÿ0.386 ÿ0.200 ÿ0.428 0.423 0.577 0.340 0.002
0.822* 0.527 0.564 ÿ0.501 0.644 0.638 0.808*
0.008 0.648 ÿ0.584 ÿ0.293 ÿ0.086 ÿ0.094 ÿ0.275
A. Mammals Didelphis marsupialis Metachirus nubicauidata Philander opossum Marmosops parvidens Marmosa murina Micoureus demerarae Monodelphis brevicaudata Proechimys sp. Oryzomys megacephalus Oecomys sp. B. Frogs Adenomera andreae Eleutherodactylus fenestratus Eleutherodactylus zimmermaneae Colosthetus stepheni Colosthatus marchesianus Atelopus pulcher Dendrophryniscus minutus
a Species with an asterix were signi®cantly correlated (P<0.05) with that axis.
a similar species composition, as did the paired sites at Area 51.
Fig. 2. (A) Ordination of linear remnants and continuous forest sites based on composition of small mammal communities. (B) Ordination of the same sites based on composition of frog communities. Site abbreviations: CPR=Cidade Powell linear remnant; CPF=Cidade Powell continuous forest; FR=Florestal linear remnant; FF=Florestal forest; PVR=Ponta Verde linear remnant; PVF=Ponta Verde forest; A51R=Area 51 linear remnant; A51F=Area 51 forest.
Hyla minuta, and Synapturanus sp. occurred at only a single site and were each represented by just 1±2 individuals. Neither species richness nor the abundance of the most common species diered between treatments (Table 2). 3.4. Frog species composition The ANOSIM failed to detect evidence of distinctive groupings of forest and remnant frogs (F=0.91, P=0.97), again indicating that community composition of linear remnants was not systematically dierent from that in continuous forest. The ordination also did not reveal any grouping according to treatments (Table 3; Fig. 2). However, some pairs of adjoining sites did appear to be similar in composition. The remnant and continuous-forest site at Ponta Verde, for example, had
4. Discussion The linear remnants surveyed in this study were not retained for the purpose of protecting wildlife, but rather for watershed conservation. Although considerably wider (140±190 m) than is required by Brazilian legislation (minimum of 60 m for streams of under 10 m in width; Laurance and Gascon, 1997), these remnants were still vulnerable to physical and biotic edge eects, which have been detected up to 200±300 m from forest margins (Brown and Hutchings, 1997; Malcolm, 1997; Laurance and Bierregaard, 1997; Laurance et al., 1998a; Stevens and Husband, 1998). The fact that we detected no consistent dierences in species richness, composition, or abundance for litter frogs and small mammals, however, suggests that edge eects in remnants of moderate width (140±190 m) were tolerated by these taxa. Interestingly, species such as Calouromys philander and Atelopus pulcher, which appear sensitive to habitat disturbance and fragmentation (Malcolm, 1991; Tocher, 1998), were detected in linear remnants in this study. Nevertheless, as a group, small mammals and frogs are less sensitive to fragmentation than many other vertebrate groups, and a number of the species captured in this study can also use the matrix of modi®ed habitats surrounding fragments (Gascon et al.,
246
M.G. de Lima, C. Gascon / Biological Conservation 91 (1999) 241±247
1999). It should also be noted that the linear remnants we studied were directly linked to nearby continuous forest, and thus would be analogous to short (<1 km long) corridors. Our results suggest that linear remnants of moderate width could have important conservation value in the context of a deforested landscape, at least for some smaller (<2 kg) forest vertebrates. The remnants probably function as movement corridors and breeding habitat for some species. For example, we recorded marked individuals of D. marsupialis, P. opossum and M. murina moving between remnants and adjoining continuous forest, while the presence of juvenile M. nudicaudatus, D. marsupialis, and P. opossum in some remnants (de Lima, 1998) suggests that these species were reproducing there. Frogs of several species (C. stepheni, E. fenestratus, and D. minutus) were also heard vocalizing in remnants, which is a good indicator of mating activity. Likewise, in tropical Queensland, some arboreal mammal species were commonly observed with young in linear rainforest remnants (S. G. Laurance, 1996). The design of faunal corridors will depend on the species of interest and local context. For example, Andreassen et al. (1996a,b) showed that voles (Microtus oeconomus) could move quite easily through very narrow corridors (<0.5 m). In tropical regions, however, much wider corridors will be needed, both because many tropical species are forest-interior specialists (e.g. Lovejoy et al., 1986; Bierregaard et al., 1992; Laurance et al., 1998b) and because closed tropical forests are very prone to physical and biotic edge eects (Laurance and Bierregaard, 1997; Carvalho and Vasconcelos, 1999). In these ecosystems, remnants of at least 200±300 m width may be necessary to facilitate movements of some sensitive species (Laurance and Gascon, 1997; Laurance and Laurance, 1999), and even greater widths may be needed for long (>5 km) corridors. In addition to providing forest-interior habitat, wider remnants have greater spatial heterogeneity, oering a broader array of microhabitats which is often correlated with greater species richness (Lindenmayer and Nix, 1993; Mee and Carrol, 1994). There is a great need to improve outdated legislation in Brazil to require the retention of wider (200±300 m) linear remnants along streams and rivers (Laurance and Gascon, 1997). There is also an urgent requirement to increase enforcement of existing legislation, which is often ignored by landowners. Educational initiatives are direly needed at all levels Ð from landowners to resource-managers to policymakers Ð to improve understanding of the key importance of retaining riparian vegetation in deforested landscapes. Given the explosive pace of development in the Amazon (Laurance, 1998), proactive initiatives to manage the deforestation process could go a long way toward maintaining some
level of ecosystem connectivity in human-dominated regions. Acknowledgements We thank O. de S. Pereira and A. M. dos Reis for help with ®eld work. Support was provided by the National Institute for Research in the Amazon (INPA), Smithsonian Institution, Brazilian National Council for Research (CNPq), USAID-Brazil, The MacArthur Foundation, and The Summit Foundation. W.F. Laurance, W.Z. Lidicker, and W. Magnusson provided valuable comments on earlier versions of the paper, and W.F. Laurance helped in ®nal editing. This is publication number 231 of the BDFFP technical series. References Andreassen, H.P., Ims, R.A., Steinset, O.K., 1996a. Discontinuous habitat corridors: eects on male root vole movements. Journal of Applied Ecology 33, 555±560. Andreassen, H.P., Halle, S., Ims, R.A., 1996b. Optimal width of movement corridors for root voles: not too narrow and not too wide. Journal of Applied Ecology 33, 63±70. Belbin, L., 1991. The analysis of pattern in bio-survey data. In: Margules, C.R., Austin, M.P. (Eds.), Nature Conservation: Cost Eective Biological Surveys and Data Analysis. CSIRO, Melbourne, Australia, pp. 176±190. Bennett, A.F., 1990. Habitat corridors and the conservation of small mammals in a fragmented forest environment. Landscape Ecology 4, 109±122. Bentley, J.M., Catterall, C.P., 1997. The use of bushland, corridors, and linear remnants by birds in Southeastern Queensland, Australia. Conservation Biology 11, 1173±1189. Bierregaard, R.O., Stouer, P.C., 1997. Understory birds and dynamic habitat mosaics in Amazonian rainforest. In: Laurance, W.F., Bierregaard, R.O. (Eds.), Tropical Forest Remnants: Ecology, Management, and Conservation of Fragmented Communities. University of Chicago Press, Chicago, pp. 138±155. Bierregaard Jr., R.O., Lovejoy, T.E., Kapos, V., dos Santos, A.A., Hutchings, R.W., 1992. The biological dynamics of tropical rainforest fragments. Bioscience 42, 859±866. Brown Jr., K.S., Hutchings, R.W., 1997. Disturbance, fragmentation, and the dynamics of diversity in Amazonian Forest Butter¯ies. In: Laurance, W.F., Bierregaard, R.O. (Eds.), Tropical Forest Remnants: Ecology, Management, and Conservation of Fragmented Communities. University of Chicago Press, Chicago, pp. 91±110. Burbrink, F.T., Phillips, C.A., Heske, E.J., 1998. A riporian zone in Southern Illinois as a potential dispersal corridor for reptiles and amphibians. Biological Conservation 86, 107±115. Carvalho, K.S., Vasconcelos, H.L., 1999. Forest fragmentation in central Amazonia and its eects on litter-dwelling ants. Biological Conservation. Clarke, K.R., 1993. Non-parametric multivariate analysis of changes community structure. Australian Journal of Ecology 18, 117±143. de Lima, M.G., 1998. ComposicËaÄo da comunidade de pequenos mamõÂferos e sapos de liteira em remanescentes lineares no Projeto DinaÃmica BioloÂgica de Fragmentos Florestais (PDBFF). M.Sc. thesis, National Institute for Research in the Amazon (INPA), Manaus, Brazil. DeMers, M.N., 1993. Roadside ditches as corridors for range expan-
M.G. de Lima, C. Gascon / Biological Conservation 91 (1999) 241±247 sion of the western harvester ant (Pogonomyrmex occidentalis Cresson). Landscape Ecology 8, 93±102. Didham, R.K., 1997. The in¯uence of edge eects and forest fragmentation on leaf-litter invertebrates in central Amazonia. In: Laurance, W.F., Bierregaard, R.O. (Eds.), Tropical Forest Remnants: Ecology, Management, and Conservation of Fragmented Communities. University of Chicago Press, Chicago, pp. 55±70. Downs, N.C., Racey, P.A., Weller, N.C., 1998. Do bats believe in the wildlife corridor hypothesis? Abstracts of the 11th International Bat Research Conference, Universidade de BrasõÂlia, BrasõÂlia, Brazil, p. 25. Ferreira, L.V., Laurance, W.F., 1997. Eects of forest fragmentation on mortality and damage of selected trees in central Amazonia. Conservation Biology 11, 797±801. Forman, R.T.T., 1997. Land Mosiacs. Cambridge University Press, Cambridge, UK. Gascon, C., 1996. Amphibian litter fauna and river barriers in ¯ooded and non-¯ooded Amazonian rain forests. Biotropica 28, 136±140. Gascon, C., Lovejoy, T.E., Bierregaard Jr., R. O., Malcolm, J.R., Stouer, P.C., Vasconcelos, H., Laurance, W.F., Zimmerman, B., Tocher, M. & Borges, S., 1999. Matrix habitat and species persistence in tropical forest remnants. Biological Conservation 91, 231± 239. Henein, K., Merriam, G., 1990. The elements of connectivity where corridor quality is variable. Landscape Ecology 4, 157±170. Hobbs, R.J., 1992. The role of corridors in conservation: Solution or bandwagon? Trends in Ecology and Evolution 7, 389±391. INPE, 1998. De¯orestamento, 1995-1997. Instituto Nacional de Pesquisas Espaciais. San Jose dos Campos, Brazil. Kapos, V., 1989. Eects of isolation on the water status of forest patches in the Brazilian Amazon. Jounal of Tropical Ecology 5, 173± 185. Kapos, V., Wandelli, E., Camargo, J.L., Ganade, G., 1997. Edgerelated changes in environment and plant responses due to forest fragmentation in central Amazonia. In: Laurance, W.F., Bierregaard, R.O. (Eds.), Tropical Forest Remnants: Ecology, Management, and Conservation of Fragmented Communities. University of Chicago Press, Chicago, pp. 33±54. La Polla, V.N., Barrett, G.W., 1995. Eects of corridor width and presence on the population dynamics of the meadow vole (Microtus pennsylvanicus). Landscape Ecology 8, 25±37. Laurance, S.G., 1996. The utilisation of linear rainforest remnants by arboreal marsupials in north Queensland. M.Nat.Res. thesis, University of New England, Armidale, Australia. Laurance, S.G., Laurance, W.F., 1999. Tropical corridors: use of linear rainforest remnants by arboreal mammals. Biological Conservation. Laurance, W.F., 1991. Ecological correlates of extinction proneness in Australian tropical rain forest mammals. Conservation Biology 5, 79±89. Laurance, W.F., 1998. A crisis in the making: Responses of Amazonian forests to land use and climate change. Trends in Ecology and Evolution 11, 411±415. Laurance, W.F., Gascon, C., 1997. How to creatively fragment a landscape. Conservation Biology 11, 577±579. Laurance, W.F., Bierregaard, R.O. (Eds.), 1997. Tropical Forest Remnants: Ecology, Management, and Conservation of Fragmented Communities. University of Chicago Press, Chicago. Laurance, W.F., Ferreira, L.V., Rankin-de Merona, J.M., Laurance, S.G., 1998a. Rain forest fragmentation and the dynamics of Amazonian tree communities. Ecology 69, 2032±2040. Laurance, W.F., Ferreira, L.V., Rankin-de Merona, J.M., Laurance,
247
S.G., Hutchings, R., Lovejoy, T.E., 1998b. Eects of forest fragmentation on recruitment patterns in Amazonian tree communities. Conservation Biology 12, 460±464. Lindenmayer, D.B., Nix, H.A., 1993. Ecological principles for the design of wildlife corridors. Conservation Biology 7, 627±630. Lovejoy, T.E., Bierregaard Jr., R.O., Rylands, A.B., Malcolm, J.R., Quintela, C.E., Harper, L.H., Brown Jr., K.S., Powell, A.H., Powell, G.V.N., Schubart, H.O.R., Hays, M.B., 1986. Edge eects and other eects of isolation on Amazon forest fragments. In: SouleÂ, M.E. (Ed.), Conservation Biology: The Science of Scarcity and Diversity. Sinauer Associates, Sunderland, Massachusetts, pp. 257±285. Malcolm, J.R., 1991. The small mammals of Amazonian forest fragments: Pattern and process. Ph.D. thesis, University of Florida, Gainesville, Florida. Malcolm, J.R., 1997. Biomass and diversity of small mammals in Amazonian forest fragments. In: Laurance, W.F., Bierregaard, R.O. (Eds.), Tropical Forest Remnants: Ecology, Management, and Conservation of Fragmented Communities. University of Chicago Press, Chicago, pp. 207±221. McCune, B., Meord, M.G., 1995. PC-ORD: Multivariate Analysis of Ecological Data, Version 2.0. MjM Software Design, Gleneden Beach, Oregon, USA. Mee, G.K., Carroll, C.R., 1994. What is conservation biology?. In: Mee, G.K., Carroll, C.R. (Eds.), Principles of Conservation Biology. Sinauer Associates, Sunderland, Massachusetts, pp. 3±23. Murcia, C., 1995. Edge eects in fragmented forests: Implications for conservation. Trends in Ecology and Evolution 10, 58±62. Noss, R.F., 1987. Corridors in real landscapes: A reply to Simberlo and Cox. Conservation Biology 1, 159±164. Rankin-de Merona, J.M., Prance, J.M., Hutchings, R.W., Silva, M.F., Rodrigues, M.A., Uehling, M.E., 1992. Preliminary results of a large-scale inventory of upland rain forest in the central Amazon. Acta Amazonica 22, 493±534. Ruefenacht, B., Knight, R.L., 1995. In¯uences of corridor continuity and width on the survival and movement of deermice Peromyscus maniculatus. Biological Conservation 71, 269±274. Rylands, A.B., Keuroghlian, A., 1988. Primate populations in continuous forest and forest fragments in central Amazonia. Acta AmazoÃnica 18, 67±83. Simberlo, D., Cox, J., 1987. Consequences and costs of conservation corridors. Conservation Biology 1, 63±71. Simberlo, D., Farr, J.A., Cox, J., Mehlman, D.W., 1991. Movement corridors: Conservation bargains or poor investments. Conservation Biology 6, 493±503. Skole, D., Tucker, C., 1993. Tropical deforestation and habitat fragmentation in the Amazon: Satellite data from 1978 to 1988. Science 260, 1905±1910. Stevens, S.M., Husband, T.P., 1998. The in¯uence of edge on small mammals: Evidence from Brazilian Atlantic forest fragments. Biological Conservation 85, 1±8. Tocher, M., 1998. A comunidade de anfõÂbios da AmazoÃnia central: DiferencËas na composicËaÄo especõ®ca entre a mata primaÂria e pastagens. In: Gascon, C., Moutinho, P. (Eds.), Floresta AmazoÃnica: DinaÃmica, RegeneracËaÄo e Manejo. National Institute for Research in the Amazon (INPA), Manaus, Brazil, pp. 219±232. Tocher, M.D., Gascon, C., Zimmerman, B.L., 1997. Fragmentation eects on a central Amazonian frog community: a ten year study. In: Laurance, W.F., Bierregaard, R.O. (Eds.), Tropical Forest Remnants: Ecology, Management, and Conservation of Fragmented Communities. University of Chicago Press, Chicago, pp. 124±137.