Rapid survey of the herpetofauna in an area of forest management in eastern Acre , Brazil

Timber exploitation is the most profitable and popular use of forest. To evaluate how this activity affects biodiversity is of crucial importance for conservation. Therefore, we carried out a herpetofauna survey in a forest area under a reduced–impact management, in Sena Madureira municipality, state of Acre (Brazil). We used active searching, both day and night, and pitfall traps. In the pre-exploitation phase the search was conducted basically in forests and forest edges; in the post-exploitation phase we also searched in sites such as secondary roads, skid trails and timber storage yards. We recorded 38 amphibian species and 28 reptile species; 17 snakes, nine lizards and two crocodilians. It is expected that this survey can contribute to programs of environmental monitoring in forest management areas in Amazonia.


Introduction
The Amazon has approximately 60% of the forest remnants of the world (Whitmore 1997).The Amazon Forest provides ecosystem services to the whole world, and plays a vital role in the maintenance of biodiversity, hydrological cycle and carbon sequestration (Fearnside 2008).
Estimates of the deforestation in the Amazon show that from 1961 to 2003, 6% of the area was converted to plantations, 62% to pastures and 32% was deforested (Ramankutty et al. 2007).Currently the deforestation rates in the Brazilian Amazon are measured by the "Instituto Nacional de Pesquisas Espaciais" (INPE), with visual interpretation of Landsat images through the Amazon Deforestation Project (PRODES;INPE 2009).These data are considered scientifically sound, though underestimated (Laurance et al. 2002), since the method does not detect other human activities such as changes that cause loss of forest canopy, such as superficial fires, edge effects, hunting, small scale mining and selective logging (Cochrane et al. 1999;Nepstad et al. 1999).
Timber exploitation is the most popular and profitable use of forest (Azevedo-Ramos et al. 2005) and has a significant effect on the biological functioning of the forest (Malhi et al. 2008).Industrial logging has been dramatically increasing in the Amazon (Laurance et al. 2001), with expansion rates of 12,000-20,000 km2 .year-1 , a value that is very similar to deforestation rates (Malhi et al. 2008).
The direct impacts of this practice include road building and opening of yards and clearings during the operations, which cause tree mortality, erosion and soil compaction, grass invasion and microclimate changes associated with the loss of the canopy cover (Uhl and Abstract: Timber exploitation is the most profitable and popular use of forest.To evaluate how this activity affects biodiversity is of crucial importance for conservation.Therefore, we carried out a herpetofauna survey in a forest area under a reduced-impact management, in Sena Madureira municipality, state of Acre (Brazil).We used active searching, both day and night, and pitfall traps.In the pre-exploitation phase the search was conducted basically in forests and forest edges; in the post-exploitation phase we also searched in sites such as secondary roads, skid trails and timber storage yards.We recorded 38 amphibian species and 28 reptile species; 17 snakes, nine lizards and two crocodilians.It is expected that this survey can contribute to programs of environmental monitoring in forest management areas in Amazonia.The aim of this study was to evaluate the impact of timber extraction on the herpetofauna in a forest in southwestern Amazonia.

Study area
The survey was carried out in São Jorge I Farm (09°26′ 11″ S, 68°37′19″ W), located in Sena Madureira municipality (Figure 1), property of the company Laminados Triunfo Ltda.The area comprises 3,798.23 ha, of which 3,461.85 ha were destined for logging using techniques of reduced impact, logging intensity of 30 m 3 /ha and a cutting cycle of 25 years.The predominant forest type is open forest with palm trees, with some open forest with bamboos.The relief is slightly hilly, the annual average temperature is 25°C, and the annual average rainfall is 2,125 mm (Acre 2006).

Data collection
The study was carried out in two phases.The first sampling period was carried out from May to June 2008, when the herpetofauna was surveyed before cutting, skidding and removal of trees, hereafter called "timber pre-exploitation phase".The second sampling period was carried out eleven months later, in May 2009, when the trees had been removed from the forest, hereafter called "timber post-exploitation phase", using the following sampling methods: 1) Active Search (Vanzolini and Papavero 1967;Lema and Leitão-de-Araujo 1985;Franco et al. 2002).This method consisted of searching in all micro-environments: leaf litter, under fallen logs, holes in the ground and in trees, aquatic environments, and others, during the day and night.We also recorded amphibians and reptiles that were found by chance.
2) Pitfall traps (Heyer et al. 1994).Three trap sequences were built, following the recommendations of Cechin and Martins (2000).These traps consist of containers buried in the ground and interconnected by a guide fence (canvas or plastic screen).Each trap array was composed of five 50L plastic buckets, buried 10 m apart and connected by a canvas fence about 1 m high.The three trap sequences were monitored daily for 10 days.
The same methods were repeated with the same number of researchers at the same locations, but after logging, additional locations were added.
Both sampling phases were conducted over 10 days, with 111 h of search in each phase.In the pre-exploitation phase amphibian and reptiles were recorded in forest environments and in forest edges; in the post-exploitation phase we also searched in human altered locations, such as secondary roads and timber storage yards.Specimens were killed with lidocaine (applied to the abdomen of amphibians or injected into reptiles), fixed in a solution of 10% formalin and preserved in a 70% ethanol solution.All vouchers were deposited in the herpetological collection of "Universidade Federal do Acre" campus "Ciências da Natureza", Rio Branco, Acre State, Brazil (IBAMA permit 12178-2) (Appendix 1).

Results
We recorded 38 species of amphibians and 28 species of reptiles; 17 snakes, nine lizards and two crocodilians (Figures 2 and 3).In the first phase we found 28 anuran species, nine lizard species, 14 snake species and only one crocodilian species.In the second phase we recorded 33 anuran species, four lizards, seven snake species and two crocodilian species.

Discussion
The greatest number of anurans were found in the forest: 30 species in the inner forest, six associated with temporary ponds in the forest and 12 in temporary ponds formed on roads and in timber storage yards.Only nine species were found at forest edges, six in dams, and five in temporary ponds (Table 1).Five species of frogs (Rhaebo guttatus, Rhinella castaneotica, Osteocephalus sp, Chiasmocleis bassleri and Hamptophryne boliviana), were found only in the first phase.However, ten species were found only in the second phase (Table 1).
Roads and timber storage yards surveyed in phase 2 aided the formation of temporary ponds, in which some generalist species that depend on water for reproduction were recorded: three species of the genus Phyllomedusa (P.tomopterna, P. camba and P. bicolor), two of the genus Dendropsophus (D. brevifrons and D. parviceps), three of the genus Hypsiboas (H.fasciatus, H. lanciformis and H. punctatus), two species of Scinax (S. ruber and S. garbei) plus Trachycephalus typhonius, Engystomops freibergi and Rhinella marina.
The increase in the number of sites for reproduction in managed areas may make anuran species more abundant, and therefore also raise the abundance of other animals, such as snakes, lizards and other frogs (Vitt and Caldwell 2001).We recorded the presence of a terrestrial predator (Ceratophrys cornuta), a large frog that feeds on other frogs (Duellman and Lizana 1994).Species considered as generalists, occupying more than three environments (Rhinella marina, Hypsiboas fasciatus, Phyllomedusa bicolor and Trachycephalus typhonius), may be favored and may increase their abundance, thereby affecting other species in the area.
Although some species are able to colonize altered environments in the Amazon, others are affected by the loss of reproductive habitats (e.g.Aichinger 1991; Tocher 1998).In general, human interventions lead to an impoverishment of the structure and diversity of vegetation (Liddle and Scorgie 1980).Such a change alters the habitat of many species, causing the disappearance of specialist species in favor of generalist species (Van Rooy and Stumpel 1995) Decreased diversity of frogs due to formation of forest in pasture areas was observed in a adjacent state, Municipality of Espigão do Oeste in Rondônia State (Bernarde 2007), where Bernarde and Macedo (2008) also reported greater richness and diversity of frogs in forest floor litter.Many animals use the leaf-litter as a microhabitat for reproduction, refuge from dessication, protection from predators, feeding, and social behaviors.We found species from the families Aromobatidae (Allobates sp. and A. femoralis) and Dendrobatidae (Ameerega hahneli and A. trivittata).In the leaf litter, these anurans lay eggs on rolled or folded dry leaves and transport their tadpoles to small water bodies, such as bromeliad pools, leaf axils, fallen fruit capsules or tree holes (Caldwell 1998;Lima et al. 2006).
We also found in the leaf litter and on trees some species of the genus Pristimantis (P.fenestratus, Pristimantis sp. 1, Pristimantis sp. 2).This group exhibits the most specialized form of reproduction, which is independent of aquatic environments, and is characterized by eggs laid on the ground (Heinicke et al. 2007).The absence of a larval phase in these anurans makes these species sensitive to humidity.Therefore, the conservation and survival of the eggs depend on high humidity in the environment, which prevents their desiccation (Duellman 1988).Species of the genus Pristimantis are sensitive to logging and effects of logging, since they exhibit a high susceptibility to environmental variables and may diminish their abundance in exploited areas (Pearman 1997).
Based on the ecological characteristics of Amazon Forest lizards the temperature may play an important role on the determination of the distribution of habitat and microhabitats of each species (Vitt and Zani 1996).
In the present survey we recorded six lizard species in the inner forest and three heliothermic lizard species (Ameiva ameiva, Kentropyx calcarata and Tupinambis teguixin) occupying the forest edges and human altered areas such as secondary roads, timber storage yards and clearings (Table 3).These lizards have a preference for open environments with sparse vegetation cover, and are strongly influenced by environmental degradation, mainly the removal of trees in exploited areas (Vitt et al. 1998;Lima et al. 2001), benefiting from the increase of insolation in these sites (Vitt et al. 1998;Fredericksen et al. 1999).
In other locations near the state of Acre, researchers found decreased numbers of frogs (Bernarde et al. 1999;Bernarde 2007), snakes (Bernarde and Abe 2006) and lizards (Macedo et al. 2008) in pasture areas in Espigão do Oeste, Rondônia State.These authors suggest that terrestrial and arboreal lizards are unable to colonize open habitats after deforestation, corroborating Vitt and Caldwell (2001).
A study carried out in an area of small-scale logging in Porto Walter municipality (Acre) showed that heliothermic lizards benefit from basking on fallen logs.In this study, Vitt and collaborators (1998), suggest that there is an increase in the density of heliothermic lizards resulting from the increase in density of fallen trees in the exploited forest.This increase in heliothermic lizards may affect other lizard populations that inhabit the forest because of competition for resources and predation (Vitt et al. 1998).Similarly, the heliothermic lizards found in this study may exert predation pressure on other animals, including other lizard species.
We cannot compare species abundance and richness between in the two phases (pre and post-exploitation, since the survey was carried out for only two months in 2008 and one month in 2009.Rapid surveys provide only some of the necessary information needed for decisionmaking regarding the impact of environmental changes on the resident species (Avila-Pires et al. 2007).estimated; for the column of the phase they correspond: 1-first sampling period (timber pre-exploitation phase) and 2-second sampling period (timber post-exploitation phase).

Figure 1 .
Figure 1.Map of state of Acre, showing the location of São Jorge I Farm (red dot), located in Sena Madureira municipality.