Seasonality , age structure and reproduction of Leptodactylus ( Lithodytes ) lineatus ( Anura , Leptodactylidae ) in Rondônia state , southwestern Amazon , Brazil

Leptodactylus (Lithodytes) lineatus (Schneider, 1799) is an Amazonian leaf litter frog considered rare or uncommon in several studies on anuran communities. Despite being a widely distributed frog in Amazonian forests, knowledge of the biology and ecology of this species is relatively scarce. This species has been reported to live in association with leaf-cutter ant nests (Atta spp.) during the breeding period. In this paper we present data on the seasonality of this species and some reproductive information gathered at a locality of Rondônia state, northwestern Brazil. Field work was carried out between April 2001 and March 2002, with the use of pitfall traps with drift fences as a survey method. Leptodactylus (L.) lineatus had a higher capture frequency in this locality compared to that of other studies carried out in other Amazonian localities, possibly because this species has secretive habits, such as calling and breeding from nests of leaf-cutting ants, and are difficult to find during visual encounter surveys. The breeding period occurs between October and March. Calling males and egg-bearing females were found between September and February and juvenile recruitment occurred mainly from the end of the rainy season to the beginning of the dry season (February to June). Males and females show sexual dimorphism in SVL, females being significantly larger than males. The number of ovarian eggs per female varies from 110 to 328 and analyses indicate that there is a significant correlation with SVL.

During a research studies conducted on anuran communities in the municipality of Espigão do Oeste, State of Rondônia (Brazil) (see BERNARDE, 2007;BERNARDE & MACEDO, 2008), we captured several individuals of Leptodactylus (L.) lineatus.In this paper we present data on the occurrence of this species (males, females and juveniles) over the course of a year and some reproductive information (calling activity of males, number and diameter of ovarian eggs in females).Herein, we compare the abundance of this species at our study site with that of other Amazonian localities.
Field work was carried out from April 2001 to March 2002 with the use of pitfall traps with drift fences (CAMPBELL & CHRISTMAN, 1982).In this method, plastic barrels (200 L) were buried every 10 m in a straight line and connected by a one-meter high canvas fence.Six trap arrays were built with four barrels each and a 44 m long fence.The trap arrays were positioned in three different habitats, with two trap arrays in each habitat: inside the forest, away from permanent water bodies; inside the forest ca 100 m from a river (about 4 m wide and 0.80 m deep) and in pasture areas, far from permanent water bodies.Each trap array was built at least 1,800 m away from the next one, and for each habitat both trap arrays were separated by a minimum distance of 200 m.Traps remained open all year and were monitored from 3 to 4 times a week.Other data, such as calling activity and individuals observed in the leaf litter were also occasionally recorded.The senior author became familiar with the advertisement call of this species from calling males kept captive for two nights in November 2001.
Specimens captured were killed in 10% alcohol and fixed in 10% formalin.To describe the age structure of this population, specimens were classified as adult males, adult females and juveniles.The individuals were sexed by dissection or by reproductive condition (e.g.presence of ovarian eggs in females and extended vocal sacs or hypertrofied gonads in males).The snout-vent length of the smallest female with mature ovarian eggs was used to set the lower size limit of this category; data on male adult size in the literature (RODRÍGUEZ & DUELLMAN, 1994;LIMA et al., 2006) were used to set the lower size limit of the adult male category.The number and diameter of mature ovarian eggs were measured via the removal of ovaries.The maximum diameter of five eggs was obtained for each female and the mean was calculated.Snout-vent length (SVL) and egg diameter were determined to the nearest 0.1 mm with callipers.The SVL data were used to separate females without ovarian eggs from juveniles.All specimens collected are housed in the "AAG-UFU" herpetological collection at the Museu de Biodiversidade do Cerrado, in the Universidade Federal de Uberlândia (MG), Brazil.
Spearman's test (ZAR, 1984) was used to look for possible correlations between capture frequency and rainfall and between SVL of the egg-bearing females and number of ovarian eggs.A Student's t-test was used to compare the SVL of males and females.The computational analysis was performed in JMP 5.0.1 to PC (SAS Institute).

RESULTS AND DISCUSSION
A total of 112 individuals of Leptodactylus (L.) lineatus were captured in pitfall traps.Of these, 101 were analysed, belonging to the following sex and size classes: 46 juveniles, 32 females and 23 males (Tab.I; Fig. 1).
Less individuals (n=38) were captured during the six driest months (April to September; rainfall = 456mm) in relation to the six rainiest months (n=74; October to March; rainfall = 1,850 mm).A smaller number of individuals was captured between June and August 2001, which were the driest months (Fig. 1).However, no positive correlation was found between the capture frequency and rainfall patterns (rs =0.3184; p=0.3213; n = 12).Adults were more frequent during the months of September to December (Fig. 1) and the egg-bearing females were captured only from September to February.Calling activity of the males was observed from November to January in forested areas.During the whole study period, only in two occasions, a male (not founded) was heard calling from inside a nest of the leaf-cutting ant (Atta sp.) in open areas.The recruitment period of juveniles occurred mainly between February and March (see Figs. 1, 2).
The majority of individuals (n=84) was captured in the forest near a river environment; another 28 individuals were captured in a forest distant from permanent water bodies.No individuals were captured with pitfall traps in pasture areas.
Females larger than 38.1 mm had mature ovarian eggs (Tab.III).The egg number, estimated from ovarian Leptodactylus (L.) lineatus was the fourth most abundant anuran species captured in pitfall traps (BERNARDE & MACEDO, 2008).By comparison, during a snake community study, only four individuals (three juveniles amid the leaf litter in the forest and one adult under a fallen tree trunk in pasture areas) were observed in 960 hours of visual encounter surveys (BERNARDE & ABE, 2006).Our abundance data for L. (L.) lineatus are greater than those from other studies conducted elsewhere in the Amazon (Tab.IV).
The capture frequency of L. (L.) lineatus was higher at Espigão do Oeste (RO) than in many other Amazonian localities (see table IV for comparisons) possibly due to the use of pitfall traps to assess anuran community diversity in our study.The species' secretive habits, such as calling from and bredding within nests of the leaf cutting ant (SCHLÜTER & REGÖS, 1981;LAMAR & WILD, 1995), make it difficult to detect individuals during visual encounter surveys at all times of day.The lower observed abundance of this frog in anuran communities may be a result of the biases of the most common methods employed to inventory communities (visual encounter surveys, plots), which are not necessarily designed to detect this species.Without the use of pitfall traps, we would have only recorded a total of four individuals.Pitfall traps are an efficient technique to detect some species that may not be detected by other methods very often (CECHIN & MARTINS, 2000).The results of our study indicate that this is the case for the leaf litter frog L. (L.) lineatus.
Only one individual was found in a pasture habitat by occasional survey.Despite having been found in pasture habitats (TOCHER, 1998) to be mostly a forest (primary and secondary forest) dweller (AICHINGER, 1991;RODRÍGUEZ & DUELLMAN, 1994;CALDWELL & ARAÚJO, 2005).In the municipality of Espigão d'Oeste, the capture frequency for this species was greater with proximity to the river, possibly because it provided the most appropriate environment for breeding, especially for tadpole development (see RODRÍGUEZ & DUELLMAN, 1994;LAMAR & WILD, 1995).Egg-bearing females and calling males were found in the rainy season (September to February), and juvenile recruitment occurred mainly during the end of rainy season to beginning of the dry season (February to June).Although some species of anurans can breed throughout the year (DUELLMAN, 1978), in unstable environments (e.g. that have dry spells) the reproduction of some species is limited to the rainy season (AICHINGER, 1987;MOREIRA & LIMA, 1991).The seasonal recruitment pattern observed in L. (L.) lineatus is similar to that observed in other species that depend on standing water for larval developement in Central Amazon (MOREIRA & LIMA, 1991).
The ovarian egg numbers recorded for females of this species are 195 and 230 (DUELLMAN, 1978) and an average of 250 (LIMA et al., 2006).In this study we recorded a mean of 243 eggs, with ranges from 110 to 328 eggs (n = 17).There was a significant correlation between number of ovarian eggs and female size, a pattern which is also observed in other species of the genus Leptodactylus (MARTINS, 1988;PRADO et al., 2000;GIARETTA & KOKUBUM, 2004).
Leptodactylus lineatus is one of the most abundant leaf litter frogs in the forests of Espigão do Oeste, Rondônia, Southwestern Amazon (Brazil).The number of individuals detected in other studies may be a reflection of the methods used (visual encounter surveys and plots), and not necessarily a lower density.The secretive habits of this frog (males call within nests of leaf cutting ants) do not facilitate visual detection.
Breeding occurs during the rainy season.Calling males and egg-bearing females were recorded from September to February, and juvenile recruitment occurs mainly from the end of rainy season to the beginning of the dry season (February to June).
Males and females show size-related secondary sexual dimorphism, with females being significantly larger than males.
The number of ovarian eggs per female varies from 110 to 328 and there is a significant correlation between number of eggs and body size.

Table IV .
Relative abundance of Leptodactylus (L.) lineatus published in studies on herpetofauna communities carried out in different Amazonian localities.Abundance = Number of individuals detected or assessments on relative abundance by the respective authors.VES = Visual Encounter Surveys.