This study describes a tropical rainforest snake assemblage in Central Amazonia, based on resource use (microhabitat and food), time of activity, species richness and abundance, morphology, behavoir, and reproduction. The data base was obtained between 1991 and
1994 in several localities around Manaus, mostly at Reserva Ducke (RFAD), a 100 km2
“terra firme” rainforest tract, 25 km north of Manaus, State of Amazonas, northern Brazil. The climate in this region is caracterized by 2100 mm of anual rainfall, with a dry season extending from July to September and a rainy one from November to May, temperature ranged between 18 and 37oC, and mean humidity about 85%.
The main method used was visual search along forest trails. Part of this search was made regularly (about 90-100 man-hours each month, during 18 months) and considered “time constrained search”; this method resulted in 274 snake findings and provided capture rates. The remaining findings (N = 234) were considered occasional and included those by others (N = 98). Most sampling was made at night. Almost all snakes were
marked by ventral scale clipping. In some additional areas around Manaus, visual searches and pitfall traps were used, resulting in 177 snake findings. Only a few snakes were collected at Reserva Ducke.
A total of 508 snake findings, of 50 species, were made at Reserva Ducke. During time constrained search (a total of about 1600 man-hours), the rate of snake findings was 0.064 and 0.217 snakes per man-hour, during the day and at night, respectively. Although time constrained search provides comparable rates of snake findings, only 31 out of 50 species were found using this method; the remaining were found occasionally. The apparently
most abundant species at Reserva Ducke, were Xenoxybelis argenteus, Bothrops atrox, Imantodes cenchoa, and Dipsas sp. The three former species also seemed to be the most abundant in other studies in Amazonia.
Twenty nine species were found by day and 41 at night. The proportions of species found in each microhabitat was similar during the day and at night. A comparison of the patterns of habitat use found at Reserva Ducke and other localities in Amazonia indicated that “exchanges” (or sum and subtractions) of species using different microhabitats, within each major colubrid lineage, may explain most of the differences found among these studies.
Snake activity at Reserva Ducke seemed to be influenced by the amount of rainfall; activity was lower in the dry season and higher during the rainy season, probably as a response to the apparently low availabilty of certain prey during the dry season. No relationship was found between moonlight and snake activity; the number of active snakes found was similar in dark and clear nights.
The most consumed prey types by the snakes of Reserva Ducke were lizards (eaten frequently or occasionally by 60% of the species with known diet, N = 48), frogs (42%), mammals (23%), birds (23%), and snakes (19%). Nine species feed on invertebrates (six on earthworms) and only one on arthropods. These results reflect, mostly, the history of colonization of the region by different snake lineages, and are also contrary to the hypothesis of differential prey abundance as a major factor determining the patterns observed in neotropical snake assemblages (for instance, insects are very abundant at Reserva Ducke, although consumed by only one snake species).Morphological analyses were based in three measurements: body length, tail length,
and weight. An analysis of maximum length distribution within the major colubrid lineages (colubrines, South American xenodontines, and Central American xenodontines), that occur at RFAD, showed that features related to these lineages (thus, historical) are responsible for most of the general pattern observed for colubrids and for the entire assemblage. The relationship between body and tail length showed that, in general, arboreal species have longer tails than terrestrials, that have longer tails than fossorials, in agreement with the idea that there is a strong effect of habitat use on tail length in snakes. Finally, an analysis of weight-length relationships showed that, in general, arboreals tend
to be lighter than terrestrials, that tend to be lighter than aquatics, confirming the effects of habitat use in snake body form. These tendencies became more evident in the analyses where colubrids were separated in major lineages. An additional analysis on color and
color patterns confirmed the effect of defence in snake color patterns.
A cluster analysis based on data on habitat use, time of activity, diet, and size (length and weight) split the assemblage into guilds where high overlaps in form and resource use are evident; in several cases these guilds were made of closely related species, indicating the presence of constraints inherent to each lineage sampled.
Although data on reproduction is scarce for most species, there are snakes at Reserva Ducke in which births occur only during the rainy season and in others occur throughout the year. A general anaysis of the presence of juveniles in the populations sampled indicated a strong tendency to seasonal breeding by the snakes of Reserva Ducke, contradicting most speculations on the patterns of juvenile recruitment in Amazonian snakes. The seasonality in reproduction, as in activity, may be related to the probably low availability of certain prey types during the dry season.
A general analysis of the results indicate that most patterns found at Reserva Ducke may well be explained by historical factors as previously predicted by J. E. Cadle and H. W. Greene in a review of the role of history on the organization of neotropical snake assemblages. Concomitantly, a critical review of the arguments favoring the hypothesis that consider competition as a major structuring force in amazonian snake assemblages indicate that these arguments tend to be irrelevant before several evidences are found in natural assemblages, especially alterations in the reproductive success in the species thought to be competing. In conclusion, it is suggested, based on a series of arguments, that the co-occurrence os 50 snake species at Reserva Ducke may be due to the combination of the following: (1) resource abundance and/or low snake densities would allow the coexistence of a relatively large number of snake species; (2) thus, the
populations would be regulated mainly by predation and/or other biotic and abiotic factors to a level where densities were not high enough to result in resource deployment (and, perhaps, competition) (some studies on Amazonian snake assemblages converged to these speculations while others, to completely conflicting ones). Concomitantly, the patterns found in the assemblage of Reserva Ducke may be a natural result of the history of colonization of the region by the various snake lineages that constitute this assemblage.